Reductive activation of mitomycins A and C by vitamin C

Article abstract of DOI:10.1016/j.bioorg.2013.03.002

The anticancer drug mitomycin C produces cytotoxic effects after being converted to a highly reactive bis-electrophile by a reductive activation, a reaction that a number of 1-electron or 2-electron oxidoreductase enzymes can perform in cells. Several reports in the literature indicate that ascorbic acid can modulate the cytotoxic effects of mitomycin C, either potentiating or inhibiting its effects. As ascorbic acid is a reducing agent that is known to be able to reduce quinones, it could be possible that the observed modulatory effects are a consequence of a direct redox reduction between mitomycin C and ascorbate. To determine if this is the case, the reaction between mitomycin C and ascorbate was studied using UV/Vis spectroscopy and LC/MS. We also studied the reaction of ascorbate with mitomycin A, a highly toxic member of the mitomycin family with a higher redox potential than mitomycin C. We found that ascorbate is capable to reduce mitomycin A efficiently, but it reduces mitomycin C rather inefficiently. The mechanisms of activation have been elucidated based on the kinetics of the reduction and on the analysis of the mitosene derivatives formed after the reaction. We found that the activation occurs by the interplay of three different mechanisms that contribute differently, depending on the pH of the reaction. As the reduction of mitomycin C by ascorbate is rather inefficiently at physiologically relevant pH values we conclude that the modulatory effect of ascorbate on the cytotoxicity of mitomycin C is not the result of a direct redox reaction and therefore this modulation must be the consequence of other biochemical mechanisms.

Full text of DOI:10.1016/j.bioorg.2013.03.002

Bioorganic Chemistry 48 (2013) 1–7
Contents lists available at SciVerse ScienceDirect
Bioorganic Chemistry
journal homepage: www.elsevier.com/locate/bioorg
Reductive activation of mitomycins A and C by vitamin C
⇑
Manuel M. Paz
Departamento de Química Orgánica, Facultade de Química, Universidade de Santiago de Compostela, 15782 Santiago de Compostela, Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 24 September 2012
Available online 6 April 2013
The anticancer drug mitomycin C produces cytotoxic effects after being converted to a highly reactive
bis-electrophile by a reductive activation, a reaction that a number of 1-electron or 2-electron oxidore-
ductase enzymes can perform in cells. Several reports in the literature indicate that ascorbic acid can
modulate the cytotoxic effects of mitomycin C, either potentiating or inhibiting its effects. As ascorbic
acid is a reducing agent that is known to be able to reduce quinones, it could be possible that the
observed modulatory effects are a consequence of a direct redox reduction between mitomycin C and
ascorbate. To determine if this is the case, the reaction between mitomycin C and ascorbate was studied
using UV/Vis spectroscopy and LC/MS. We also studied the reaction of ascorbate with mitomycin A, a
highly toxic member of the mitomycin family with a higher redox potential than mitomycin C. We found
that ascorbate is capable to reduce mitomycin A efficiently, but it reduces mitomycin C rather ineffi-
ciently. The mechanisms of activation have been elucidated based on the kinetics of the reduction and
on the analysis of the mitosene derivatives formed after the reaction. We found that the activation occurs
by the interplay of three different mechanisms that contribute differently, depending on the pH of the
reaction. As the reduction of mitomycin C by ascorbate is rather inefficiently at physiologically relevant
pH values we conclude that the modulatory effect of ascorbate on the cytotoxicity of mitomycin C is not
the result of a direct redox reaction and therefore this modulation must be the consequence of other bio-
chemical mechanisms.
Keywords:
Anticancer drugs
Mitomycin C
Bioreductive drugs
Quinones
Ascorbic acid
Ó 2013 Elsevier Inc. All rights reserved.
1. Introduction
[7,9]. In addition to DNA alkylation and cross-linking, other biologi-
cal modes of action for MMC have been postulated, including redox
The mitomycins are a family of highly cytotoxic antibiotics pro-
duced by Streptomyces cultures. Mitomycin C (MMC, 1) is used clin-
ically in chemotherapy for several types of cancer [1] and also as an
antifibrotic adjuvant to a number of surgical procedures [2,3]. An-
other member of the mitomycin family is mitomycin A (MMA, 2),
which differs from MMC in the substituent at the 7-position (7-
methoxy instead of 7-amino). MMA is several orders of magnitude
more cytotoxic than MMC towards cancer cells, and this cytotoxicity
correlates with its higher redox potential [4,5]. MMC is considered
the prototype bioreductive drug: it is inert towards nucleophiles
in its original structure, but it is converted to an extremely reactive
bis-electrophile after a cascade of reactions initiated by reduction of
the quinone ring (Fig. 1) [6]. The reductive activation of mitomycin C
generates the reactive intermediate 3a, containing two electrophilic
positions -at C1 and C10- that alkylates DNA and other biological
molecules [7,8]. The formation of an interstrand cross-link adduct
with DNA is consistently considered the cause of its cytotoxic effects
cycling [10,11], inhibition of ribosomal RNA [12], and inhibition of
thioredoxin reductase [13].
A number of 1-electron and 2-electron oxidoreductases have
been shown to activate MMC [14], two of which seem to play a
key role in cancer cells: DT-diaphorase under normal oxygen condi-
tions [15] and cytochrome P450 reductase under hypoxic conditions
[16]. Small-molecule biological reducing agents such as glutathione
or dihydrolipoic acid can also activate MMA [5,17], but they do not
reduce MMC [18,19] or do so rather inefficiently [20].
Vitamin C (ascorbic acid, 2, AscH, Fig. 2) is a well-known phys-
iological reducing agent [21]. Ascorbate is a cofactor of a number of
oxidase enzymes, playing a key role in a number of hydroxylation
reactions [22]. Other biologically relevant roles of AscH include its
free radical scavenger activity [23] and the inhibition of the forma-
tion of carcinogenic N-nitrosamines by the reaction of nitrite with
amines [24]. Quinones are also a substrate for reduction by AscH:
several reports have shown that quinones can be reduced by vita-
min C to generate the semiquinone or the hydroquinone forms
[25–27]. In this context, the formation or reactive oxygen species
by redox cycling of quinones using pharmacological concentrations
of ascorbate has been proposed as a novel anticancer therapy [28].
Abbreviations: AscH, ascorbic acid; MMA, mitomycin A; MMC, mitomycin C.
⇑
Fax: +34 881814468.
E-mail address: manuel.paz@usc.es
0045-2068/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.bioorg.2013.03.002

2
M.M. Paz / Bioorganic Chemistry 48 (2013) 1–7
were prepared in H2O, and their concentrations were determined
spectrophotometrically using e367 (H2O) = 21,840 cmꢁ1 Mꢁ1 for
MMC and e320 (H2O) = 10,400 cmꢁ1 Mꢁ1 for MMA.
2.2. Kinetics of the reactions between MMC and ascorbate
Reactions were started by adding an aqueous ascorbic acid or
sodium ascorbate solution in the appropriate buffer to a buffered
aqueous solution of MMC in a UV/Vis cuvette thermostated at
37 °C by
0.50 mL. The absorbance at 360 nm was recorded at 1-s intervals.
The molar fraction of MMC at time t ( MMC(t)) was calculated from
a circulating water bath, using a total volume of
v
the absorbance readings using the following equation:
vMMCðtÞ ¼ ðAt ꢁ A Þ=ðA0 ꢁ A Þ
1
1
where A0 is the absorbance of the solution at 360 nm before the
addition of ascorbate (corrected for the final volume), At is the
absorbance of the solution at a time t after the addition of ascorbate
and A1 is the absorbance of the solution after all the mitomycin C is
consumed.
Fig. 1. Mechanism for the reductive activation of mitomycins. Nu1H and Nu2H
represent two nucleophilic compounds.
2.3. Kinetics of the reactions between MMA and ascorbate
Reactions were performed analogously to those of MMC, except
that the decrease in the absorbance at 550 nm was recorded.
2.4. Kinetics of the reaction of MMC with ascorbate: dependence on the
concentration of ascorbate
Fig. 2. Oxidation of ascorbic acid to dehydroascorbic acid. The structure of the
cyclic hemiacetal monohydrate of dehydroascorbic acid that is present in aqueous
solutions is shown [63].
The reactions were performed according to the general protocol
described above in 500 mM carbonate/bicarbonate buffer pH 10.5.
The final concentrations in the reaction mixture were 50
and 12–100 mM ascorbate.
lM MMC
A substantial number of studies have shown that AscH can
modulate the biological effects of MMC, either enhancing or inhib-
iting its effects [29–39]. Considering that quinones can be reduced
by AscH, one could hypothesize that the observed modulation of
the cytotoxic effects of MMC by ascorbate are the result of a direct
redox reaction between MMC and ascorbate. If this reduction oc-
curs, then it would result in either the activation of MMC to cyto-
toxic alkylating intermediates or in detoxification by conversion of
MMC to non-toxic metabolites. To test the likelihood of this
hypothesis we studied the reactions between MMC and ascorbate
by examining the kinetics of the reduction of MMC to mitosenes
and by analyzing the identity of mitosene derivatives formed after
reaction of MMC and ascorbate. We also included in our studies the
highly toxic parent compound MMA, a more easily reducible mem-
ber of the mitomycin family of antibiotics.
2.5. Kinetics of the reaction of MMC with ascorbate: pH-dependence
The reactions were performed according to the general protocol
described above. The final concentrations in the reaction mixture
were 50 lM MMC, 50 mM sodium ascorbate and 300 mM buffer.
Buffers were AcONa/AcOH (for pH 4–5) NaH2PO4/Na2HPO4
(pH 6–9) and NaHCO3/Na2CO3 (pH 10–11). Tris buffer at pH 9.0
inhibited the reaction.
2.6. Kinetics of the reaction of MMA with ascorbate: dependence on the
concentration of ascorbate
The reactions were performed according to the general protocol
described above in 100 mM phosphate buffer pH 6.7. The final con-
centrations in the reaction mixture were 30
10 mM ascorbate.
lM MMA and 2–
2. Materials and methods
2.7. Kinetics of the reaction of MMA with ascorbate: pH-dependence
2.1. General procedures
The reactions were performed according to the general protocol
described above. The final concentrations in the reaction mixture
were 30 lM MMA, 2 mM sodium ascorbate and100 mM buffer.
Buffers were AcONa/AcOH (for pH 4–5) NaH2PO4/Na2HPO4 (pH 6–
8.5) and NaHCO3/Na2CO3 (pH 9.5–10.5).
Mitomycin C was a gift from Kyowa Hakko Kogyo Co. Ltd. Mito-
mycin A was prepared as reported [40]. Ascorbic acid and sodium
ascorbate were from Sigma–Aldrich. HPLC was performed with a
Hewlett Packard Series 1100 diode array system. LCMS was per-
formed with a Hewlett Packard Series 1100 diode array HPLC sys-
tem connected to a Hewlett Packard Series 1100 MSD mass
spectrometer.
A
Macherey Nagel (Nucleodur C18,
5
l
m,
2.8. LC/MS analysis of the reactions of mitomycins with ascorbate
4.6 ꢀ 125 mm) column was used. The elution system was 5–30%
B in 20 min, 30–60% B in 9 min. (A = 10 mM ammonium acetate,
pH 5.5; B = acetonitrile, concentrations expressed in v/v). UV/Vis
spectra and kinetic studies were performed with a CARY 300
UV–visible spectrometer. Stock solutions of mitomycins A and C
A solution of MMC or MMA in water (100
tion) was treated with a solution of ascorbate (15
solution in the corresponding buffer). The buffers used were
AcOH/AcONa for pH 4 and pH 5, NaH2PO4/Na2HPO4 for pH
l
L of a 3 mM solu-
l
L of a 10 mM

M.M. Paz / Bioorganic Chemistry 48 (2013) 1–7
3
Fig. 3. Structures of the mitosane and mitosene core.
6–pH 8 and NaHCO3/Na2CO3 for pH 9.The resulting solutions were
maintained at room temperature for 2 h and analyzed directly by
LCMS.
2.9. Spectral data
À
Á
8: EI-MS (ES+) m/z: 260 M ꢁ OCONHꢁ2 , 343 (M+Na+), 359
(M+K+). UV/Vis (kmax, nm): 252, 310. 9: EI-MS (ES+) m/z: 418
À
Á
M ꢁ OCONHꢁ2 , 501 (M+Na+), 517 (M+K+). UV/Vis (kmax, nm):
À
Á
250, 312. 10: EI-MS (ES+) m/z: 302 M ꢁ OCONHꢁ2 , 385 (M+Na+),
401 (M+K+). UV/Vis (kmax, nm): 252, 310. 11: EI-MS (ES+) m/z:
À
Á
275 M ꢁ OCONH2ꢁ , 358 (M+Na+), 374 (M+K+). UV/Vis (kmax
,
,
À
Á
nm): 234, 287, 345. 12: EI-MS (ES+) m/z: 433 M ꢁ OCONHꢁ2
516 (M+Na+), 532 (M+K+). UV/Vis (kmax, nm): 234, 286, 352. 13:
À
Á
EI-MS (ES+) m/z: 317 M ꢁ OCONH2ꢁ , 400 (M+Na+), 416 (M+K+).
UV/Vis (kmax, nm): 236, 284, 342. 4a: EI-MS (ES+) m/z: 244
À
Á
M ꢁ OCONHꢁ2 , 327 (M+Na+), 343 (M+K+). UV/Vis (kmax, nm):
À
Á
250, 312 360 (sh). 14: EI-MS (ES+) m/z: 420 M ꢁ OCONH2ꢁ , 442
(M+Na+), 458 (M+K+). UV/Vis (kmax, nm): 244, 316, 360 (sh) 4b:
À
Á
EI-MS (ES+) m/z: 259 M ꢁ OCONH2ꢁ , 342 (M+Na+), 358 (M+K+).
UV/Vis (kmax, nm): 230, 287, 350. 15: EI-MS (ES+) m/z: 435
(M+H+), 457 (M+Na+), 473 (M+K+). UV/Vis (kmax, nm): 231, 288,
À
Á
360. 16: EI-MS (ES+) m/z: 286 M ꢁ OCONHꢁ2 , 347 (M+H+), 369
(M+Na+), 385 (M+K+). UV/Vis (kmax, nm): 252, 310.
Fig. 4. (a) UV/Vis assay of the reduction of 50
lM mitomycin C by 50 mM sodium
ascorbate in 300 mM phosphate buffer, pH 5.0 at 37 °C. (b) UV/Vis assay of the
reduction of 30 M mitomycin A by 2 mM sodium ascorbate in 200 mM phosphate
buffer, pH 7.5 at 37 °C. Reaction times (min) are indicated on top of each spectrum.
l
2.10. Comparison with authentic samples
Samples of 1-hydroxy-7-aminomitosenes (8) were prepared by
acid hydrolysis of MMC in water according to the procedure de-
scribed by Taylor and Remers [41]. Samples of 1-hydroxy-7-meth-
oxymitosenes (11) were prepared similarly by acid hydrolysis of
MMA. 2-amino-7-methoxymitosene (4b) was prepared by reduc-
tive activation of MMA with DTT [17]. 2,7-DAM (4a) was prepared
by catalytic hydrogenation of MMC in methanol [42].
(kmax 550 nm) to the 7-methoxymitosene chromophore (kmax
350 nm) (Fig. 4B). The reduction of MMA can also be monitored
by the decrease of an absorbance maximum at 317 nm [17], but
this wavelength could not be used in this case due to interferences
caused by the presence of ascorbate/dehydroascorbate. MMA was
reduced in 30 min using 2 mM ascorbate at pH 7.5 indicating, as
expected, a much higher reactivity for MMA compared to MMC.
The rate of reaction showed a marked pH-dependence for both
mitomycins (Fig. 5). In the case of MMC the rate decreased when
the pH was increased from pH 4 to pH 5, and then increased the
range pH 5–11 with increasing pH values (Fig. 5a and c). For
MMA the rate of reduction increased with increasing pH values
in the studied pH range (Fig. 5b and d). The rate of reduction of
both mitomycins was proportional to the concentration of added
ascorbate (Fig. S1, Supporting information).
The formation of mitosene compounds formed after the activa-
tion of both mitomycins by AscH was analyzed using LC/MS
(Fig. 6). In addition to the known metabolites 8, 10, 11, 13 [41],
4a [46], 4b [17] and 16 [47] (Fig. 7), several metabolites that incor-
porate ascorbate as nucleophile were identified based in their UV
and ESI-MS.1 For MMC, two of them were assigned as the isomeric
C1 ascorbate adducts 9 (Fig. 7): their UV spectra is characteristic of
1-substituted mitosenes, and the MS shows molecular ions
3. Results
The reduction of the quinone ring in mitomycins A or C triggers
the elimination of methanol, resulting in the irreversible conver-
sion of the mitosane chromophore to the mitosene chromophore
(Fig. 3), two structures with significantly different absorbance
spectra [43]. This irreversible structural change allows to study
of the reductive activation of mitomycins using UV/Vis spectros-
copy [17,20,44,45]. The first experiments to determine if MMC
could be reduced by AscH were performed by tracking the conver-
sion of the 7-aminomitosane chromophore (kmax 360 nm) to the 7-
aminomitosene chromophore (kmax 315 nm). The time-dependent
UV/Vis spectra of a reaction containing MMC and a large excess
of AscH (50 mM) at pH 5.0 clearly showed the gradual decrease
of the mitosane chromophore and the appearance of the mitosene
chromophore (Fig. 4A), indicating the formation of mitosene deriv-
atives at the expense of MMC. The reaction reached completion in
about 90 min, as judged by constant UV/Vis spectra. The reductive
activation of MMA by ascorbate was analyzed similarly by moni-
toring the conversion of the 7-methoxymitosane chromophore
1
The characterization of the novel adducts is based solely in their UV and ESI-MS,
and no further attempts were made to isolate and rigorously characterize the
ascorbate-mitomycin adducts due to limited availability of mitomycins and their low
proportion in a complex mixture of mitosenes.

4
M.M. Paz / Bioorganic Chemistry 48 (2013) 1–7
Fig. 6. HPLC analysis of the reaction of MMC and MMA with ascorbate in phosphate
buffer (pH 6), showing the formation of C1-reduced mitosenes as major products (a)
Chromatogram of a reaction of MMC (0.5 mM) with ascorbate (20 mM) at 20 °C for
18 h (k = 315 nm). (b) Chromatogram of a reaction of MMA (0.5 mM) with ascorbate
(5 mM) at 20 °C for 3 h (k = 280 nm).
Fig. 5. pH-dependence of the reaction of mitomycins A and C with ascorbate. (a)
Reaction of 50 mM MMC with 50 mM ascorbate at 37 °C. (b) Reaction of 30 mM
MMA with 2 mM ascorbate. (c) and (d) Half-time for the reduction of MMC and
MMA by ascorbate respectively, determined from the data shown in (a) and (b). See
materials and methods for details.
at pH 7 we could identify three peaks in the HPLC trace that showed
an MS corresponding to structure 9 (Fig. S3, Supporting information),
indicating that ascorbate is alkylated at two sites, probably C-2 and
O-3.
The identity of the mitosene compounds formed and their rela-
tive proportions depended markedly on the pH of the reaction
(Fig. 8). At pH 6, C1-reduced mitosenes were the most abundant
species observed in the chromatogram. When the pH was either in-
creased or decreased from pH 6, the relative amounts of C1-re-
duced mitosenes diminished in favor of mitosenes derived from
monofunctional activation.
corresponding to Na+ and K+ adducts of 9, and a major peak corre-
sponding to the loss of the carbamoyloxy group (Supporting infor-
mation, Figs. S2A and S2B). We propose structure 14 (Fig. 7), the
adduct at C10 of 2,7-diaminomitosene 4a, for the other ascorbate-
MMC adduct observed in the HPLC chromatograms: the UV spectra
shows the characteristic bands of the mitosene chromophore and
the MS shows as major peaks the molecular ions for the H+, Na+
and K+ adducts of structure 14 (Supporting information, Figs. S2G
and S2H). For MMA, the analogous structures 12 and 15 (Fig. 7)
are proposed using a similar rationale. The precise alkylation site
of ascorbate in these adducts is uncertain. It is known that the alkyl-
ation of AscH provides a mixture of 3-O-alkyl, 2-O-alkyl, 2,3-di-O-al-
kyl and/or 2-C-alkyl derivatives, and the predominant product
depends on the specific reaction conditions [48–50]. The alkylation
of AscH with quinone methides derived from 4-hydroxybenzyl alco-
hols in water proceeded predominantly through the C-2 position
[51]. At pH 10.5 or higher ascorbate reacts predominantly through
the O-2 position [52]. In our case, ascorbate adducts are observed
at pH 8 or below, and therefore the most likely positions for alkyl-
ation are C-2 or O-3. In some reactions between MMC and ascorbate
4. Discussion
4.1. On the mechanisms of activation of mitomycins by ascorbate
Both the pH-dependence of the rate of reduction of mitomycins
and the pH-dependent distribution of mitosene compounds can be
rationalized considering that the three known mechanisms for the
activation of mitomycins operate in the reactions of MMC and
MMA with ascorbate. Before discussing our results on the reduc-
tion of mitomycins by ascorbate we present in the following para-
graphs a brief review of the mechanisms for the activation of
mitomycins (Fig. 9): the acidic (non-reductive) activation [45],
the autocatalytic reductive activation mechanism [44] and the

M.M. Paz / Bioorganic Chemistry 48 (2013) 1–7
5
Fig. 7. Structure of the mitosene derivatives detected after the reactions of MMC or
MMA with AscH.
Fig. 9. Mechanisms for the reductive activation of mitomycins
A
and
C
by
ascorbate.
group intact (e.g. 8–13) [41,45]. Addition of nucleophiles to 18 oc-
curs by a SN1 mechanism to form two isomers at C1.
Fig. 8. Distribution of mitosene compounds detected by HPLC after the reaction of
MMC (a) or MMA (b) with ascorbate at different pH values. White bars represent
compounds derived from monofunctional activation (acid activation of autocata-
lytic reductive activation). Grey bars represent mitosenes derived from bifunctional
activation.
The autocatalytic mechanism and the bifunctional activation
mechanism share in common the first two mechanistic steps: a
reduction of the quinone ring to form hydroquinone 17, followed
by elimination of methanol to give 3 [53]. The leucoaziridinomito-
sene 3 constitutes the branch point for the competing autocatalytic
activation mechanism and the bifunctional activation mechanism.
In the autocatalytic path, 3 reduces mitomycin to form aziridin-
omitosene 18 and hydroquinone 17, that eliminates methanol to
form leucoaziridinomitosene 3, triggering the autocatalytic reac-
tion. The reduction of mitomycins by 3 is driven by the large redox
bifunctional reductive activation mechanism [53]. The acidic acti-
vation occurs by an acid-catalyzed elimination of methanol in
MMC or MMA to form 18 without the need of a reduction [45].
The mitosenes formed by this mechanism are exclusively C-1
substituted mitosenes retaining the C-10 carbamoyloxy functional

6
M.M. Paz / Bioorganic Chemistry 48 (2013) 1–7
potential difference between mitosenes and mitosanes [44]. The
autocatalytic reduction of mitomycins is favored by slow reducing
agents and by alkaline pH values, and it is kinetically characterized
by a sigmoid reaction curve [20,44]. The metabolites formed by
this pathway are C-1 substituted mitosenes, analogous to those
formed in the acidic activation [20,44]. The acidic activation only
occurs significantly at pH values below 5, therefore the formation
of mitosenes containing the C10 carbamoyloxy group in reactions
at pH above 5 indicates that the autocatalytic mechanism is
operating.
The bifunctional activation pathway for the activation of mito-
mycins involves an intramolecular aziridine-opening reaction in
leucoaziridinomitosene 3 to form 19 (Fig. 9). The vinylogous qui-
none methide 19 constitutes another branch point in the mecha-
nism of activation of mitomycins, as a consequence of its dual
reactivity: The C1 carbon in 19 can function as an electrophile
due to its vinylogous quinone methide structure, or as a nucleo-
phile as a consequence of its vinylogous enol nature. The reaction
of 19 as an electrophile (path B in Fig. 9) results eventually in the
addition of two nucleophiles at C1 and C10 to give mitosenes with
structures 5a (from MMC) and 5b (from MMA) (Fig. 1). The forma-
tion of interstrand cross-links in the reaction of reductively acti-
vated mitomycins with duplex DNA is the result of this type of
reactivity. Alternatively, 19 can react as a vinylogous enol by tau-
tomerizing to the corresponding keto form (path A in Fig. 9), which
results in the formation of C1-reduced mitosenes 4a (from MMC)
or 4b (from MMA). 2,7-diaminomitosene 4a is the major in vivo
metabolite of MMC [54], and it functions as a DNA alkylating agent,
forming adducts at C10 [55]. The formation of covalent DNA ad-
ducts by 4a requires an additional reductive activation step that
converts the C10 carbon into an electrophilic position, as shown
in Fig. 9.
The data obtained in the pH-dependence experiments shown in
Figs. 5 and 8 illustrates the interplay of the three mechanisms re-
viewed above in the reduction of mitomycins with ascorbate. The
observed pH-dependent distribution of mitosene derivatives
(Fig. 8) serves to determine the contribution of the three mecha-
nisms at a given pH value. Both MMC and MMA showed a similar
pH-dependent distribution of products, reflecting similar overall
reactivity for both mitomycins. At pH 4 both mitomycins react pre-
dominantly by the acidic activation mechanism. At pH 5, MMC re-
acted mostly by acidic activation, while the bifunctional activation
mechanism starts to contribute significantly to the activation of
MMA. The bifunctional activation becomes the major pathway
for both mitomycins only at pH 6. From this point forward, its con-
tribution starts to decrease in favor of the autocatalytic pathway, a
mechanism that becomes the only one operating at pH 9 both for
MMC and MMA. The data at from pH 6 to pH 9 explains how the
pH determines the partition of 19 between the monofunctional
activation pathway (conversion to 4) and the bifunctional activa-
tion pathway (conversion to 20). The distribution of mitosenes
shown in Fig. 8 can be traced back in the activation mechanisms
(Fig. 9) to the partition of 19 between 4 and 20 (the acidic activa-
tion of MMC that also produces adducts 8–13 is not relevant at
pH P 6). A singular feature of the HPLC traces of the reactions of
mitomycins and ascorbate is that adducts with structures 5a (for
MMC) or 5b (for MMA) were not observed, as it is usually the case
in the reactions of MMC under bifunctional activation conditions
[56–58].
tion of the three operating mechanisms depending on the reaction
pH. Leucoaziridinomitosene 3, the species whose accumulation
triggers the autocatalytic mechanism (Fig. 9) is more stable with
increasing pH, as its conversion to 19 is triggered by protonation
of the aziridine ring. As a consequence, the autocatalytic mecha-
nism becomes increasingly preponderant at higher pH values at
the expense of the bifunctional activation mechanism. The other
feature of these experiments is that the overall rate of reduction in-
creases with increasing pH values (with the exception of the acidic
activation of MMC at pH 4). This pH-influence correlates well with
the pH-dependence reported for the reduction of other quinones
with AscH. The rate of reduction of p-benzoquinone with AscH in-
creased when the pH was increased from pH 2 to pH 5 [27]. Mukai
et al. reported that the reduction of tocopheroxyl radical with vita-
min C was markedly pH-dependent, with a maximum rate at pH 8,
suggesting that ascorbate monoanion is the actual reducing species
[59]. In the reduction of mitomycins with ascorbate, the rate of
reduction continues to increase after pH 8 (Fig. 5), but this effect
should not be attributed to a faster reduction of the quinone ring
by ascorbate. Instead, the faster reaction rates at alkaline pH are
most likely the result of an increasing contribution of the autocat-
alytic mechanism (white bars in Fig. 8) as a consequence of the
greater stability of 3 at higher pH values.
4.2. Biological significance
A number of studies have shown that the biological effects
caused by MMC exposure can be altered by AscH. For example,
Marshall and Rauth reported that the presence of AscH increased
the aerobic but not the hypoxic cytotoxicity of MMC towards Chi-
nese hamster ovary cells and a repair-deficient mutant [30]. Other
studies showed a slight inhibition by AscH of the clastogenic ef-
fects of MMC in mice [34] and a potentiation of the antitumor
activity of MMC by ascorbate [33]. Several reports from Getoff
et al. showed that the activity of MMC was modified by AscH under
irradiation [35–39]. In some cases differing results were obtained
depending on the model studied: when monitoring the effects of
AscH on MMC-induced sister chromatid exchanges (SCEs), Krish-
naja and Sharma found that pretreatment with ascorbate enhanced
MMC-induced SCEs in human lymphocytes in vitro [29], while
other researchers found that AscH ascorbic inhibited SCEs induced
by MMC in bone marrow and spleen cells in mice [31,32].
The work discussed here tried to ascertain if a direct redox reac-
tion between MMC and ascorbate could be responsible for the re-
ported effects of ascorbate on MMC activity. If such an reduction
takes place it could result in either the activation of MMC to
cytotoxic alkylating species or, alternatively, in detoxification by
conversion of MMC to the non-cytotoxic derivative 2,7-diamin-
omitosene metabolite 4a [60]. Our results indicate that, at physio-
logically relevant pH values, the reduction of MMC requires
concentrations of ascorbate larger than 50 mM, much greater than
those present in cells (micromolar-low millimolar range) [61].
Also, at pH 5–7 the reduction takes place at a very slow rate
(Fig. 5a and c), with 60–80% of MMC still unreacted after 2.5 h.
At more acidic or more basic pH values the reaction progresses fas-
ter, but these reactions take place by mechanisms that are not rel-
evant physiologically: an acidic activation mechanism that
requires pH < 5 or an autocatalytic reaction that requires large con-
centrations of MMC.
Two features are evident from the experiments studying the
influence of pH on the rate of reduction of mitomycins (Fig. 5).
One of them is that the reaction curves take different shapes
depending on the pH: The curves for MMC at pH 4 and the curves
for MMA at pH values from 5.5 to 7.5 follow a quasi-exponential
decay, while reactions at higher pH values displayed a sigmoid
curve for both mitomycins. This result reflects a changing contribu-
In contrast, the highly cytotoxic drug MMA is reduced effi-
ciently using concentrations of ascorbate in the low mM range.
The differential reactivity of MMA and MMC with ascorbate
concords well with the higher redox potential of MMA
compared to MMC (E of MMA and MMC are ꢁ0.19 and ꢁ0.40 V,
½
respectively) [4]. The efficient reduction of MMA by ascorbate
was not unexpected, as the ability of ascorbate to reduce other

M.M. Paz / Bioorganic Chemistry 48 (2013) 1–7
7
[15] R.M. Phillips, A.M. Burger, P.M. Loadman, C.M. Jarrett, D.J. Swaine, H.H. Fiebig,
Cancer Res. 60 (2000) 6384–6390.
[16] Y. Wang, J.P. Gray, V. Mishin, D.E. Heck, D.L. Laskin, J.D. Laskin, Mol. Cancer
Ther. 9 (2010) 1852–1863.
[17] M.M. Paz, M. Tomasz, Org. Lett. 3 (2001) 2789–2792.
[18] M. Sharma, Q.Y. He, M. Tomasz, Chem. Res. Toxicol. 7 (1994) 401–407.
[19] M. Sharma, M. Tomasz, Chem. Res. Toxicol. 7 (1994) 390–400.
[20] M.M. Paz, Chem. Res. Toxicol. 22 (2009) 1663–1668.
[21] O. Arrigoni, M.D. Tullio, Biochim. Et Biophys. Acta (BBA) – General 1569 (2002)
1–9.
[22] J. Mandl, A. Szarka, G. Bánhegyi, Br. J. Pharmacol. 157 (2009) 1097–1110.
[23] R.C. Rose, A.M. Bode, FASEB J. : Off. Publ. Federation Am. Soc. Exp. Biol. 7 (1993)
1135–1142.
methoxyquinones has been reported earlier [26]. This efficient
reduction suggests that the direct reaction of ascorbate with
MMA may play a role in the high cytotoxicity of this drug.
Our results agree with the results reported by Marshall and
Rauth in their investigation on the effects of ascorbate in the cyto-
toxicity of MMC towards CHO cells, where they found evidence
that ascorbate was not acting by directly reducing MMC [30]. They
proposed that the increased toxicity of MMC in the presence of
ascorbate that they observed might be due to an increased produc-
tion of bioreductive enzymes or of compounds capable of modulat-
ing the reduction process. An additional mechanism by which
ascorbate may potentiate the effects of MMC is by inhibiting mul-
tidrug resistance glycoprotein (MDR1) [62].
[24] S.S. Mirvish, L. Wallcave, M. Eagen, P. Shubik, Science 177 (1972) 65–68.
[25] V.A. Roginsky, T.K. Barsukova, H.B. Stegmann, Chem. – Biol. Interact. 121
(1999) 177–197.
[26] R. Pethig, P.R. Gascoyne, J.A. McLaughlin, A. Szent-Györgyi, Proc. Natl. Acad. Sci.
USA 80 (1983) 129–132.
[27] N.S. Isaacs, R. van Eldik, J. Chem. Soc. Perkin Trans. 2 (1997) 1465–1468.
[28] J. Verrax, R. Beck, N. Dejeans, C. Glorieux, B. Sid, R.C. Pedrosa, et al., Anti-Cancer
Agents Med. Chem. 11 (2011) 213–221.
5. Conclusions
[29] A.P. Krishnaja, N.K. Sharma, Teratogen. Carcin. Mut. Suppl. 1 (2003) 99–112.
[30] R.S. Marshall, A.M. Rauth, Cancer Res. 46 (1986) 2709–2713.
[31] G. Krishna, J. Nath, T. Ong, Cancer Res. 46 (1986) 2670–2674.
[32] G. Rivas-Olmedo, S.D. Barriga-Arceo, E. Madrigal-Bujaidar, J. Toxicol. Environ.
Health Part A 35 (1992) 107–113.
[33] K. Nakano, S. Fujimoto, H. Tokita, In Vivo 2 (1988) 247–252.
[34] K. Vijayalaxmi, R. Venu, Mutat. Res. – Genet. Toxicol. 438 (1999) 47–51.
[35] C. Kammerer, N. Getoff, In Vivo 18 (2010) 795–798.
[36] L. Fuga, M. Kragl, N. Getoff, Anticancer Res. 24 (2010) 4031–4034.
[37] B. Svoboda, N. Getoff, Anticancer Res. 22 (2010) 949–952.
[38] C. Kammerer, I. Czermak, N. Getoff, R. Kodym, Anticancer Res. 19 (2010) 5319–
5321.
In summary, our results indicate that the modulatory activity
on the effects of MMC by AscH is unlikely to be the result of a direct
redox reaction between MMC and ascorbate and, therefore, alter-
native biochemical mechanisms must be responsible for this mod-
ulation. In contrast, a reductive activation by ascorbate may
contribute to the potent cytotoxic activity of MMA.
Acknowledgment
[39] E. Heinrich, N. Getoff, Z. Naturforsch., C: J. Biosci. 58 (2003) 244–248.
[40] D.M. Vyas, D. Benigni, R.A. Partyka, T.W. Doyle, J. Org. Chem. 51 (1986) 4307–
4309.
We are grateful to Kyowa Hakko Kogyo Co. Ltd. for a gift of
mitomycin C.
[41] W.G. Taylor, W.A. Remers, J. Med. Chem. 18 (1975) 307–311.
[42] E. Champeil, M.M. Paz, S. Ladwa, C.C. Clement, A. Zatorski, M. Tomasz, J. Am.
Chem. Soc. 130 (2008) 9556–9565.
Appendix A. Supplementary material
[43] J.S. Webb, D.B. Cosulich, J.H. Mowat, J.B. Patrick, R.W. Broschard, W.E. Meyer,
et al., J. Am. Chem. Soc. 84 (1962) 3185–3187.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bioorg.2013.03.
002.
[44] D.M. Peterson, J. Fisher, Biochemistry 25 (1986) 4077–4084.
[45] R.A. McClelland, K. Lam, J. Am. Chem. Soc. 107 (1985) 5182–5186.
[46] M. Tomasz, R. Lipman, Biochemistry 20 (1981) 5056–5061.
[47] M.M. Paz, Bioorg. Med. Chem. Lett. 20 (2010) 31–34.
[48] K.G.A. Jackson, J.K.N. Jones, Can. J. Chem. 43 (1965) 450–457.
[49] G. Fodor, Tetrahedron 39 (1983) 2137–2145.
References
[50] U. Beifuss, Tetrahedron 56 (2000) 357–361.
[1] M.M. Paz, Antitumour antibiotics, in: S. Missailidis (Ed.), Anticancer
Therapeutics, John Wiley and Sons, Chichester, UK, 2008, pp. 112–115.
[2] R. Shah, S. Wilson, Curr. Opin. Ophthal. 21 (2010) 269–273.
[3] D. Warner, S.E. Brietzke, Otolaryng. Head Neck. 138 (2008) 700–709.
[4] S.S. Pan, H. Gonzalez, Mol. Pharmacol. 37 (1990) 966–970.
[5] M.M. Paz, A. Das, Y. Palom, Q.Y. He, M. Tomasz, J. Med. Chem. 44 (2001) 2834–
2842.
[6] B.M. Hoey, J. Butler, A.J. Swallow, Biochemistry 27 (1988) 2608–2614.
[7] M. Tomasz, Y. Palom, Pharmacol. Ther. 76 (1997) 73–87.
[8] M. Tomasz, Chem. Biol. 2 (1995) 575–579.
[9] Y. Palom, G. Suresh Kumar, L.-Q. Tang, M.M. Paz, S.M. Musser, S. Rockwell,
et al., Chem. Res. Toxicol. 15 (2002) 1398–1406.
[10] S. Napetschnig, H. Sies, Biochem. Pharmacol. 36 (1987) 1617–1621.
[11] T. Komiyama, T. Kikuchi, Y. Sugiura, Biochem. Pharmacol. 31 (1982) 3651–
3656.
[51] A.J. Pow, R.K. Belter, J. Org. Chem. 53 (1988) 1535–1540.
[52] P.W. Lu, D.W. Lillard, P.A. Seib, K.J. Kramer, Y.T. Liang, J. Agric. Food Chem. 32
(1984) 21–28.
[53] G. Suresh Kumar, R. Lipman, J. Cummings, M. Tomasz, Biochemistry 36 (1997)
14128–14136.
[54] L. Chirrey, J. Cummings, G.W. Halbert, J.F. Smyth, Cancer Chemoth. Pharm. 35
(1995) 318–322.
[55] G. Subramaniam, M.M. Paz, G. Suresh Kumar, A. Das, Y. Palom, C.C. Clement,
et al., Biochemistry 40 (2001) 10473–10484.
[56] P. Schiltz, H. Kohn, J. Am. Chem. Soc. 115 (1993) 10497–10509.
[57] Y.P. Hong, H. Kohn, J. Am. Chem. Soc. 113 (1991) 4634–4644.
[58] M.M. Paz, Chem. Res. Toxicol. 23 (2010) 1384–1392.
[59] K. Mukai, M. Nishimura, S. Kikuchi, J. Biol. Chem. 266 (1991) 274–278.
[60] C.D. Utzat, C.C. Clement, L.A. Ramos, A. Das, M. Tomasz, A.K. Basu, Chem. Res.
Toxicol. 18 (2005) 213–223.
[61] P. Bergsten, G. Amitai, J. Kehrl, K.R. Dhariwal, H.G. Klein, M. Levine, J. Biol.
Chem. 265 (1990) 2584–2587.
[62] E.M. El-Masry, M.B. Abou-Donia, Life Sci. 73 (2003) 981–991.
[63] R.C. Kerber, J. Chem. Educ. 85 (2008) 1237–1242.
[12] R.G. Snodgrass, A.C. Collier, A.E. Coon, C.A. Pritsos, J. Biol. Chem. 285 (2010)
19068–19075.
[13] M.M. Paz, X. Zhang, J. Lu, A. Holmgren, Chem. Res. Toxicol. 25 (2012) 1502–
1511.
[14] J. Cummings, V.J. Spanswick, M. Tomasz, J.F. Smyth, Biochem. Pharmacol. 56
(1998) 405–414.