An Essential Difference between the Flavonoids
MonoHER and Quercetin in Their Interplay with the
Endogenous Antioxidant Network
Hilde Jacobs1*, Mohamed Moalin1,2, Aalt Bast1, Wim J. F. van der Vijgh1, Guido R. M. M. Haenen1
1Department of Pharmacology and Toxicology, NUTRIM School for Nutrition, Toxicology and Metabolism, Maastricht University Medical Centre (MUMC+), Maastricht, The
Netherlands, 2Faculty of Life Sciences, Hogeschool Zuyd, Heerlen, The Netherlands
Antioxidants can scavenge highly reactive radicals. As a result the antioxidants are converted into oxidation products that
might cause damage to vital cellular components. To prevent this damage, the human body possesses an intricate network
of antioxidants that pass over the reactivity from one antioxidant to another in a controlled way. The aim of the present
study was to investigate how the semi-synthetic flavonoid 7-mono-O-(b-hydroxyethyl)-rutoside (monoHER), a potential
protective agent against doxorubicin-induced cardiotoxicity, fits into this antioxidant network. This position was compared
with that of the well-known flavonoid quercetin. The present study shows that the oxidation products of both monoHER
and quercetin are reactive towards thiol groups of both GSH and proteins. However, in human blood plasma, oxidized
quercetin easily reacts with protein thiols, whereas oxidized monoHER does not react with plasma protein thiols. Our results
indicate that this can be explained by the presence of ascorbate in plasma; ascorbate is able to reduce oxidized monoHER to
the parent compound monoHER before oxidized monoHER can react with thiols. This is a major difference with oxidized
quercetin that preferentially reacts with thiols rather than ascorbate. The difference in selectivity between monoHER and
quercetin originates from an intrinsic difference in the chemical nature of their oxidation products, which was corroborated
by molecular quantum chemical calculations. These findings point towards an essential difference between structurally
closely related flavonoids in their interplay with the endogenous antioxidant network. The advantage of monoHER is that it
can safely channel the reactivity of radicals into the antioxidant network where the reactivity is completely neutralized.
Citation: Jacobs H, Moalin M, Bast A, van der Vijgh WJF, Haenen GRMM (2010) An Essential Difference between the Flavonoids MonoHER and Quercetin in Their
Interplay with the Endogenous Antioxidant Network. PLoS ONE 5(11): e13880. doi:10.1371/journal.pone.0013880
Editor: Anna Kristina Croft, University of Wales Bangor, United Kingdom
Received June 28, 2010; Accepted October 18, 2010; Published November 8, 2010
Copyright: ? 2010 Jacobs et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The authors have no support or funding to report.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: email@example.com
The human body is endowed with a wide range of antioxidants
to protect cells from damage induced by free radicals and other
reactive species. Glutathione (GSH) is one of the most important
endogenous hydrophilic antioxidants . It is synthesized in many
different cell types from its constituting amino acids glutamic acid,
cysteine and glycine, and is therefore not required in the human
diet . The actual antioxidant property of GSH is attributable to
the thiol group that is present in its cysteine moiety. As an effective
nucleophile, GSH also plays an important role in the protection
against electrophilic compounds .
Like GSH, ascorbic acid (vitamin C) is also an important
hydrophilic antioxidant. In contrast to GSH, ascorbic acid cannot
be synthesized by humans and, as a consequence, is required in
the human diet . It directly scavenges O2N2and
various other radicals.
Phenolic antioxidants comprise a-tocopherol (the most active
form of vitamin E) and flavonoids. Like ascorbic acid and most
other vitamins, a-tocopherol has to be obtained exclusively from
the diet. It is the major lipid-soluble lipoprotein antioxidant . a-
Tocopherol is localized in biomembranes and functions as an
efficient inhibitor of lipid peroxidation. Flavonoids, on the other
hand, are not essential nutrients but they form an integral part of
the human diet as they are found in fruits, vegetables, nuts and
plant-derived beverages such as tea and wine [5,6]. They have a
wide range of biological activities [7,8,9], but are most commonly
known for their antioxidant activity. Quercetin is one of the most
frequently studied dietary flavonoids [5,10]. It can scavenge highly
reactive species, an activity that is implicated in its health benefits
The flavonoid of interest to us, that closely resembles the
chemical structure of quercetin, is the semi-synthetic flavonoid 7-
mono-O-(b-hydroxyethyl)-rutoside (monoHER). MonoHER is the
most powerful antioxidant constituent of the registered drug
VenorutonH , which is used in the treatment of chronic venous
insufficiency . In vitro screening has shown that monoHER is
the most potent protector against cardiotoxicity induced by the
anticancer agent doxorubicin within a series of flavonoids .
Preclinical experiments have confirmed that monoHER is indeed a
potential protective agent against doxorubicin-induced cardiotox-
icity [15,16]. Because of these promising results, clinical trials are
being performed to study the protection of intravenously admin-
istered monoHER against doxorubicin-induced cardiotoxicity in
cancer patients [17,18]. The antioxidant activity of monoHER is
supposed to be involved in its protection. Because of its excellent
radical scavenging properties monoHER can effectively protect the
heart against free radicals produced by doxorubicin.
PLoS ONE | www.plosone.org1November 2010 | Volume 5 | Issue 11 | e13880
During the scavenging of highly reactive species, antioxidants
donate an electron or a hydrogen atom to the radical involved,
thereby converting the radical into a relatively stable non-radical.
In this way the reactivity of the radical is annihilated. However, in
this reaction the antioxidant itself is converted into an oxidation
product that takes over part of the reactivity of the radical. This
oxidized antioxidant might cause damage to vital cellular
components . For example, when a-tocopherol scavenges free
radicals it is oxidized to produce the corresponding tocopheroxyl
radicals . These radicals can recombine with other radicals,
such as peroxyl radicals, thereby neutralizing them . However,
when these tocopheroxyl radicals cannot be eliminated, lipid
peroxidation is aggravated, a phenomenon referred to as
tocopherol-mediated peroxidation [21,22].
To prevent damage by reactive oxidation products of
antioxidants, the human body has a refined network of
antioxidants that pass over the reactivity from one antioxidant to
another in a controlled way, thereby gradually diminishing the
reactivity of the radical and recycling the antioxidants. In this way,
it has been shown that ascorbate can regenerate a-tocopherol from
tocopheroxyl radicals, thereby preventing tocopherol-mediated
peroxidation [4,20,23]. This illustrates that antioxidants act in
synergy to annihilate radicals. Besides preventing damage induced
by harmful oxidation products, regeneration is important because
it restores the antioxidant network.
The regeneration of a-tocopherol by ascorbate is well
documented, however, not much is known on the regeneration
of flavonoids. When quercetin protects against free radicals, thiol-
reactive oxidation products of quercetin are formed that can cause
damage to vital cellular components, a phenomenon known as the
quercetin paradox . Recently it was found that the oxidation
product of monoHER is also reactive towards thiols . This
might have implications for the applicability of monoHER.
However, as mentioned above, antioxidants do not act in isolation
to protect against oxidative damage.
The aim of the present study was to determine how monoHER
fits into the antioxidant network and to get insight in the
regeneration of flavonoids. Particularly, the reactivity of oxidized
monoHER towards thiols and ascorbate was investigated. In
addition a comparison with quercetin was made.
Materials and Methods
For the study spare, anonymised human blood plasma obtained
from the Academic Hospital Maastricht was used according to the
procedure approved by the medical ethical review board of the
7-mono-O-(b-hydroxyethyl)-rutoside (monoHER) was kindly
provided by Novartis Consumer Health (Nyon, Switzerland).
Stock solutions of the drug were freshly prepared in a methanol/
25 mM phosphate buffer (pH 3.33) mixture (4/1, v/v). Quercetin
was purchased from Sigma (St. Louis, MO, USA) and stock
solutions were freshly prepared in methanol. Bovine serum
albumin (BSA), reduced glutathione (GSH), hydrogen peroxide
(H2O2), horseradish peroxidase (HRP), L-ascorbic acid (vitamin C)
and 5,59-dithiobis-(2-nitrobenzoic acid) (DTNB) were also pur-
chased from Sigma (St. Louis, MO, USA). Trifluoroacetic acid
(TFA) was acquired from Sigma-Aldrich (Steinheim, Germany).
Acetonitrile HPLC grade and methanol were obtained from
Biosolve (Valkenswaard, The Netherlands). 29-GSH-monoHER
was synthesized as described previously .
Oxidation of monoHER
MonoHER was oxidized as described before . Shortly,
50 mM monoHER was incubated for 5 minutes at 37uC together
with 1.6 nM HRP and 33 mM H2O2in a 145 mM phosphate
buffer (pH 7.4). The GSH-monoHER adduct was formed by
oxidizing 50 mM monoHER in the presence of 40 mM GSH. To
investigate the influence of ascorbate on the oxidation of
monoHER and on the formation of the GSH-monoHER adduct,
ascorbate (final concentration of 40 mM, unless noted otherwise)
was added to the incubation mixtures. The reactions were
monitored spectrophotometrically and by HPLC. MonoHER
consumption was determined at 355 nm, ascorbate consumption
at 270 nm.
Spectrophotometric analysis was performed with a Varian
Carry 50 spectrophotometer (Varian, Mulgrave, VIC, Australia).
All absorption spectra were recorded from 220 to 500 nm with a
scan speed of 960 nm/min, using quartz cuvettes. The UV/Vis
scans were started 30, 150 and 300 seconds after the addition of
High-performance liquid chromatography analysis
High-performance liquid chromatography (HPLC) was per-
formed using a HP 1100 series HPLC system (Agilent Technol-
ogies, Palo Alto, CA, USA). Analytical separations were achieved
using a Supelcosil LC 318 column (5 mm, 25 cm 6 4.6 mm)
(Supelco, Bellefonte, PA, USA). The mobile phase consisted of
water containing 0.1% (v/v) TFA with linear gradients of 5%
acetonitrile at t=0 to 20% acetonitrile at 5 min followed by an
increase to 30% acetonitrile at 10 min. Finally 90% acetonitrile
was used from 18 min onward for 5 min. The column was
reequilibrated with 5% acetonitrile for 5 min. A flow rate of 2 ml/
min and an injection volume of 20 ml were used. Detection was
carried out with a diode array detector (DAD). The chromato-
grams presented are based on detection at 355 nm (absorption
maximum of monoHER).
Measurement of thiol reactivity
To determine the thiol reactivity of oxidized monoHER and
quercetin, free SH-groups were measured using the DTNB [5, 59-
dithiobis-(2-nitrobenzoic acid)] assay. The incubation mixtures
contained 50 mM monoHER, 1.6 nM HRP, 33 mM H2O2and
40 mM GSH (or 400 mM BSA) in a 145 mM phosphate buffer
(pH 7.4). When the oxidation was performed in the presence of
ascorbate, the incubation mixture additionally contained 40 mM of
ascorbate. After 0 or 5 minutes of incubation at 37uC, thiol
content was measured by adding DTNB (final concentration of
0.6 mM) to the incubation mixtures. The formation of TNB was
measured spectrophotometrically at 412 nm. Similar experiments
with 50 mM monoHER were performed in human blood plasma
to determine the reactivity towards plasma protein thiols. Identical
experimental conditions were used to determine thiol reactivity of
Molecular quantum chemical calculations
Molecular quantum chemical calculations (ab initio level) were
performed with the software program Spartan ’06 (Wavefunction,
Irvine, CA, USA) to corroborate the experimental results. The
Møller Plesset, RI-MP2 with the 6-31G* basis set was used to
calculate the relative abundance of the tautomers of oxidized
quercetin. The Hartree-Fock method with the 3-21G basis set was
used for calculating the equilibrium geometry and the energies of
The Antioxidant Network
PLoS ONE | www.plosone.org2 November 2010 | Volume 5 | Issue 11 | e13880
the lowest unoccupied molecular orbital (LUMO) of the quercetin
quinone methide and a simplified monoHER quinone (the rutin
group at C3-O and the ethoxygroup at C7-O were replaced by
methyl groups) and the highest occupied molecular orbital
(HOMO) of GSH and ascorbate, unless depicted otherwise. In
addition, a LUMO map for the monoHER quinone and the
quercetin quinone methide were generated to get a visual on the
All experiments were performed, at least, in triplicate. Data are
expressed as mean 6 SD or as a typical example. Statistical
analysis was performed using student’s t-test. P values #0.05 were
considered statistically significant.
GSH reacts with oxidized monoHER to form 29-GSH-
UV and HPLC analysis (Fig. 1A and Fig. 2A) show that
oxidation of 50 mM monoHER by HRP/H2O2 leads to the
consumption of monoHER at a rate of 5.560.4 mM/min (Fig. 3).
In the presence of 40 mM GSH, all the oxidized monoHER is
recovered as 29-GSH-monoHER at a rate of 5.560.3 mM/min
(Fig. 3). This is concluded from the appearance of the
characteristic UV spectrum of 29-GSH-monoHER (Fig. 1B) and
HPLC analysis of the incubation mixture (Fig. 2B). In the HPLC
chromatogram a second peak emerges, eluting at a position
identical to that of the synthesized 29-GSH-monoHER adduct.
These data demonstrate that the monoHER quinone is formed as
a primary oxidation product and that this oxidation product forms
an adduct with GSH.
Ascorbate reduces oxidized monoHER to monoHER
As shown by UV and HPLC analysis, addition of 40 mM
ascorbate to the incubation mixture containing monoHER and
HRP/H2O2 prevents monoHER consumption (Fig. 1C and
Fig. 2C). At the same time the ascorbate concentration decreases,
as seen in the spectrum as a decrease of the absorption at 270 nm.
When monoHER is omitted from the incubation mixture, there is
no detectable ascorbate consumption. These findings suggest that
monoHER is regenerated from its oxidation product by ascorbate.
The average rate of ascorbate consumption in the presence of
monoHER (4.360.3 mM/min) is 23% less than monoHER
consumption in the absence of ascorbate (5.560.4 mM/min)
(Fig. 3). Addition of more ascorbate (final concentration of
100 mM) to the incubation mixture reduces the ascorbate
consumption to 1.560.1 mM/min, 72% less than monoHER
consumption without ascorbate. This indicates that the enzyme
HRP is also partially inhibited by ascorbate, as has been shown
previously . The extent of inhibition depends on the ascorbate
Competition between GSH and ascorbate for oxidized
To investigate the competition between GSH and ascorbate,
monoHER was oxidized in the presence of both compounds
(Fig. 1D and Fig. 2D). Comparison of the rate of 29-GSH-
monoHER formation (1.360.1 mM/min) and the rate of ascor-
bate consumption (3.060.3 mM/min) (Fig. 3) indicates that
oxidized monoHER reacts two to three times faster with ascorbate
than with GSH. These results are in contrast with those found for
quercetin. In a comparable competition experiment with quercetin
it was found that oxidized quercetin predominantly reacts with
Figure 1. Spectrophotometrical analyses. Spectrophotometrical analysis of the incubation mixture containing (A) 50 mM monoHER, 1.6 nM
horseradish peroxidase (HRP) and 33 mM H2O2. The same experiment was carried out in (B) the presence of 40 mM GSH, (C) 40 mM ascorbate and (D)
both 40 mM GSH and 40 mM ascorbate. The UV/Vis scans were recorded 30, 150 and 300 seconds after the addition of HRP. A typical example is
The Antioxidant Network
PLoS ONE | www.plosone.org3 November 2010 | Volume 5 | Issue 11 | e13880
As shown in Fig. 4A, both oxidized monoHER and oxidized
quercetin, produced in situ by HRP/H2O2-mediated oxidation,
decrease the thiol content of the incubation mixture containing
40 mM GSH to approximately 50% in 5 minutes. The presence of
ascorbate (40 mM) does not affect the consumption of thiols by
oxidized quercetin (Fig. 4B). In contrast, ascorbate significantly
decreases the thiol consumption induced by oxidized monoHER
(Fig. 4B). This confirms that oxidized monoHER preferentially
reacts with ascorbate, whereas oxidized quercetin preferentially
reacts with GSH.
Competition between protein thiols and ascorbate for
oxidized monoHER and oxidized quercetin
Next, monoHER and quercetin were oxidized in human blood
plasma. In human blood plasma GSH is practically absent and
ascorbate concentrations are 40–60 mM . The generation of
oxidized quercetin decreases the thiol content of plasma, i.e.
protein thiols, by approximately 40% (Fig. 4C). In contrast, the
generation of oxidized monoHER has no effect on the thiol
content of human blood plasma (Fig. 4C). An additional
experiment shows that oxidized monoHER is able to react with
the thiol group of albumin, which is the most abundant plasma
protein (Fig. 4D). Ascorbate is able to prevent the reaction of
oxidized monoHER with albumin (Fig. 4D).
These results point towards an essential difference between
monoHER and quercetin, i.e. oxidized monoHER rather reacts
with ascorbate than with protein thiols, while oxidized quercetin
preferentially reacts with protein thiols.
Explanation of the difference in reactivity between
oxidized monoHER and oxidized quercetin
Molecular quantum chemical calculations show that the
tautomer depicted in Fig. 5A, a quinone methide, represents
Figure 2. HPLC analyses. HPLC analysis of the incubation mixture containing (A) 50 mM monoHER, 1.6 nM horseradish peroxidase (HRP) and
33 mM H2O2. The same experiment was carried out in (B) the presence of 40 mM GSH, (C) 40 mM ascorbate and (D) both 40 mM GSH and 40 mM
ascorbate. The different incubation mixtures were injected on the HPLC system 5 minutes after the addition of HRP. A typical example is shown. The
retention time of monoHER is 6.7 min and that of 29-GSH-monoHER is 5.4 min. The initial peak height of monoHER before oxidation was 88 mAU,
corresponding to a concentration of 50 mM. After 5 min of oxidation the monoHER concentrations in the incubation mixtures A, B, C and D were
22.5 mM, 22.5 mM, 50 mM and 43.5 mM, respectively.
Figure 3. MonoHER and ascorbate consumption rates. Con-
sumption of monoHER and ascorbate in the incubation mixtures
containing 50 mM monoHER, 1.6 nM HRP and 33 mM H2O2 in the
presence of either 40 mM GSH, 40 mM ascorbate or both 40 mM GSH and
40 mM ascorbate. The incubation time was 5 minutes. All measurements
were carried out in triplicate and data are expressed as mean 6 SD.
The Antioxidant Network
PLoS ONE | www.plosone.org4November 2010 | Volume 5 | Issue 11 | e13880
more than 99% of oxidized quercetin. In oxidized monoHER only
the ortho-quinone, illustrated in Fig. 5B, can be formed.
Generation of a LUMO map of oxidized monoHER and oxidized
quercetin shows that the LUMO of oxidized monoHER is
restricted to the B ring and part of the C ring, while the LUMO of
oxidized quercetin is delocalized over all the phenolic rings (Fig. 5).
The LUMO of the monoHER ortho-quinone and the quercetin
quinone methide are 42.6 kJ/mol and 0.0605 kJ/mol, respective-
ly, showing that oxidized monoHER is a harder electrophile than
oxidized quercetin. The HOMO of ascorbate and GSH are
2394 kJ/mol and 21.35 kJ/mol, respectively, showing that
ascorbate is a harder nucleophile than GSH. According to
Pearson’s HSAB concept , hard electrophiles react faster
and form stronger bonds with hard nucleophiles, explaining the
preferential reaction of oxidized monoHER with ascorbate over
Paradoxically, free radical scavenging antioxidants are chemi-
cally converted into potentially harmful oxidation products when
they protect against free radicals . These oxidation products
Figure 4. Reactivity of oxidized monoHER and oxidized quercetin towards thiols. Thiol content of the incubation mixture containing
50 mM monoHER or quercetin, 1.6 nM HRP and 33 mM H2O2in the presence of either (A) 40 mM GSH, (B) both 40 mM GSH and 40 mM ascorbate, (C)
human blood plasma or (D) 400 mM albumin (BSA) (with or without 40 mM ascorbate). The thiol content of the different incubation mixtures was
measured 5 minutes after the addition of HRP. All measurements were carried out in triplicate and data are expressed as mean 6 SD. *P,0.05
compared to control.
Figure 5. LUMO delocalization maps. LUMO delocalization map of (A) oxidized quercetin and (B) oxidized monoHER. In oxidized monoHER, the
rutin group at C3-O and the ethoxygroup at C7-O were replaced by methyl groups and for quercetin, the most abundant tautomer (.99%) was used.
The LUMO of oxidized monoHER is localized in the B ring and part of the C ring, while the LUMO of oxidized quercetin is distributed over all the
phenolic rings (rings A, C and B).
The Antioxidant Network
PLoS ONE | www.plosone.org5 November 2010 | Volume 5 | Issue 11 | e13880
usually retain a part of the reactivity of the species they have
scavenged, and might therefore cause damage to vital cellular
targets [19,30]. To protect cells against this damage the human
body has a refined network in which the reactivity is transferred
from one antioxidant to another, thereby gradually diminishing
the reactivity [1,29].
The aim of the present study was to investigate how monoHER
fits into this endogenous antioxidant network. The interaction of
monoHER with the network was compared with that of quercetin,
which chemically closely resembles monoHER. The results of this
study show that oxidized monoHER is reduced by ascorbate to
recycle the parent compound monoHER, while oxidized mono-
HER reacts with GSH to form a GSH-conjugate. The reactions of
oxidized quercetin with ascorbate and GSH are similar to those of
oxidized monoHER . However, as shown in the present study,
a major difference is that oxidized quercetin preferentially reacts
with thiols, whereas oxidized monoHER preferentially reacts with
ascorbate. This is an essential difference in the interplay of both
flavonoids with antioxidants of the endogenous antioxidant
The different position of monoHER and quercetin in the
network has to originate from an intrinsic difference in the
chemical nature of their oxidation products. Quantum chemical
calculations revealed that of the four possible tautomeric forms of
oxidized quercetin, the tautomer shown in Fig. 5A, has an
abundance of more than 99%. In this tautomer the distance
between the electron deficient carbonyl centers is maximal, which
is energetically favorable and explains its high abundance. The
high abundance of this specific tautomer is corroborated by the
formation of adducts in the A ring, i.e. 6-GSH-quercetin and 8-
GSH-quercetin, in the reaction of GSH with oxidized quercetin
In monoHER a rutinose is attached to the 3-OH group of the C
ring and a hydroxyethyl group is attached to the hydroxyl group
oxygen at position 7 of the A ring. These substitutions preclude the
formation of quinone methide tautomers in oxidized monoHER.
Therefore, only the ortho-quinone can be formed (Fig. 5B). In this
ortho-quinone two carbonyls are adjacent, which is energetically
unfavorable compared to the larger distance between these groups
in the preferential tautomer of oxidized quercetin. The presence of
an ortho-quinone in the B ring is corroborated by the formation of
an adduct in this ring, i.e. 29-GSH-monoHER, in the reaction of
oxidized monoHER with GSH .
Apparently, the oxidation products of monoHER and quercetin
are energetically different. The LUMO of oxidized monoHER is
primarily concentrated in the B ring and therefore relatively high,
Figure 6. Chemical reaction of oxidized monoHER with ascorbate. (A) Chemical structure of oxidized monoHER (left) and ascorbate (right).
The active part of ascorbate and the active part of oxidized monoHER are indicated by the red ellipses. (B) Suggested route for the reaction of
oxidized monoHER with ascorbate. Only the active parts are shown to illustrate the suggested mechanism more clearly. (1) The active part of
ascorbate (top) approaches the active part of oxidized monoHER (bottom) due to a p-p interaction and a hydrogen bond. The p-electrons of
ascorbate are used to create a new bond. The C3 of ascorbate will most likely attack the C39 of the monoHER quinone because it is more electron
deficient than the C49 according to Spartan ‘06. (2) After the attack, a transition state, with an sp3 bond between ascorbate and oxidized monoHER, is
suggested to be formed. (3) This intermediate rapidly decomposes into monoHER and oxidized ascorbate. The driving force of this reaction is the
restoration of the highly conjugated p-system of monoHER.
The Antioxidant Network
PLoS ONE | www.plosone.org6November 2010 | Volume 5 | Issue 11 | e13880
The Antioxidant Network
PLoS ONE | www.plosone.org7 November 2010 | Volume 5 | Issue 11 | e13880
while that of oxidized quercetin is spread over the whole molecule
(Fig. 5). This is reflected by a LUMO of oxidized quercetin
(0.0605 kJ/mol) that is substantially lower than that of oxidized
monoHER (42.6 kJ/mol). Pearson’s HSAB concept assigns the
terms ‘hard’ or ‘soft’ to chemical species to explain or predict the
outcome of a chemical reaction . ‘Hard’ applies to
electrophiles (the reactants that accept binding electrons) that
have LUMO of high energy or nucleophiles (the reactants that
donate binding electrons) with a low HOMO energy. ‘Soft’, on the
other hand, applies to electrophiles with a low LUMO value or
nucleophiles with a high HOMO value. According to the HSAB
concept, hard electrophiles react faster and form stronger bonds
with hard nucleophiles, whereas soft electrophiles react faster and
from stronger bonds with soft nucleophiles.
Based on their LUMO values, oxidized quercetin is a softer
electrophile than oxidized monoHER. The reaction of GSH with
both oxidized monoHER and quercetin is a Michael addition in
which GSH acts as a nucleophile. The reaction with ascorbate is a
redox reaction in which ascorbate finally donates two electrons to
the oxidized products. GSH is a relatively soft nucleophile
(HOMO value of 21.35 kJ/mol) compared to ascorbate (HOMO
value of 2394 kJ/mol). This can explain the preferential reaction
of the soft electrophile, oxidized quercetin, with thiols over
ascorbate. Oxidized monoHER, on the other hand, is a harder
electrophile than oxidized quercetin explaining its preference for
the harder nucleophile ascorbate over GSH. Moreover, as
depicted in Fig. 6A, the active part of ascorbate can approach
the active part of oxidized monoHER by a hydrogen bond and a
p-p interaction between ascorbate and the ortho-quinone. The
reaction between oxidized monoHER and ascorbate is presented
step by step in Fig. 6B.
Based on our findings, the following concept is proposed.
Flavonoids easily pick up the reactivity of radicals due to their
superior scavenging activity. This reactivity is directed in different
ways by the two flavonoids studied (Fig. 7). Quercetin directs this
reactivity towards thiols. Conjugation of oxidized quercetin with
GSH is primarily a cellular defense mechanism to alleviate the
harmful consequences of the reactive quinone metabolite .
However, this will reduce GSH levels and thus weakens the
endogenous antioxidant network. Moreover, in e.g. blood plasma,
where GSH is practically absent, or when GSH has been depleted,
oxidized quercetin will react with protein thiols. This causes
toxicity such as increased membrane permeability  or
impaired functioning of enzymes that contain a critical thiol-
group [34,35]. In contrast to quercetin, monoHER preferentially
directs its acquired reactivity towards ascorbate. In human blood
plasma, oxidized monoHER, contrary to oxidized quercetin, does
not react with plasma protein thiols. Ascorbate present in plasma
reduces oxidized monoHER to the parent compound and prevents
that oxidized monoHER reacts with thiols. The oxidized ascorbate
formed in this recycling of monoHER can also be regenerated in
the network, e.g. by dehydroascorbate reductase that uses NADH
as cofactor. In this way, the reactivity is completely neutralized
and the antioxidant network is restored. Thus, the advantage of
monoHER is that it can function as a catalyst that safely channels
the reactivity of radicals into the endogenous antioxidant network.
This advantage might have been involved in the superior effect of
monoHER over other structurally related flavonoids  in our
screening procedure for protection against doxorubicin-induced
cardiotoxicity. To conclude, our study demonstrates that struc-
turally related flavonoids, belonging to the same subgroup and
displaying a comparable radical scavenging activity, can have a
different impact on health.
Conceived and designed the experiments: HJ AB WJvdV GRRMH.
Performed the experiments: HJ. Analyzed the data: HJ MM AB WJvdV
GRRMH. Contributed reagents/materials/analysis tools: HJ MM. Wrote
the paper: HJ MM GRRMH.
1. Meister A (1994) Glutathione-ascorbic acid antioxidant system in animals. J Biol
Chem 269: 9397–9400.
2. Yu BP (1994) Cellular defenses against damage from reactive oxygen species.
Physiol Rev 74: 139–162.
3. Meister A (1994) Glutathione, ascorbate, and cellular protection. Cancer Res 54:
4. Niki E (1987) Interaction of ascorbate and alpha-tocopherol. Ann N Y Acad Sci
5. Hertog MG, Hollman PC, Katan MB, Kromhout D (1993) Intake of potentially
anticarcinogenic flavonoids and their determinants in adults in The Netherlands.
Nutr Cancer 20: 21–29.
6. Kuhnau J (1976) The flavonoids. A class of semi-essential food components: their
role in human nutrition. World Rev Nutr Diet 24: 117–191.
7. Liu HL, Jiang WB, Xie MX (2010) Flavonoids: recent advances as anticancer
drugs. Recent Pat Anticancer Drug Discov 5: 152–164.
8. Rathee P, Chaudhary H, Rathee S, Rathee D, Kumar V, et al. (2009)
Mechanism of action of flavonoids as anti-inflammatory agents: a review.
Inflamm Allergy Drug Targets 8: 229–235.
9. Landis-Piwowar KR, Dou QP (2008) Polyphenols: biological activities,
molecular targets, and the effect of methylation. Curr Mol Pharmacol 1:
10. Formica JV, Regelson W (1995) Review of the biology of Quercetin and related
bioflavonoids. Food Chem Toxicol 33: 1061–1080.
11. Boots AW, Haenen GR, Bast A (2008) Health effects of quercetin: from
antioxidant to nutraceutical. Eur J Pharmacol 585: 325–337.
12. Amic D, Davidovic-Amic D, Beslo D, Rastija V, Lucic B, et al. (2007) SAR and
QSAR of the antioxidant activity of flavonoids. Curr Med Chem 14: 827–
13. Haenen GR, Jansen FP, Bast A (1993) The antioxidant properties of five O-(-b-
Hydroxyethyl)-Rutosides of the flavonoid mixture Venoruton. Phlebology
14. Petruzzellis V, Troccoli T, Candiani C, Guarisco R, Lospalluti M, et al. (2002)
Oxerutins (Venoruton): efficacy in chronic venous insufficiency–a double-blind,
randomized, controlled study. Angiology 53: 257–263.
15. van Acker FA, van Acker SA, Kramer K, Haenen GR, Bast A, et al. (2000) 7-
monohydroxyethylrutoside protects against chronic doxorubicin-induced car-
Figure 7. Interplay of monoHER and quercetin with the endogenous antioxidant network. (A) Schematic representation of the
endogenous antioxidant network. Free radicals are scavenged by antioxidants in the network, such as GSH and ascorbate. In this way, free radicals are
neutralized. Free radicals that are not neutralized can damage e.g. proteins, lipids and DNA. (B) The flavonoid quercetin is an excellent radical
scavenger. During the scavenging of free radicals quercetin becomes oxidized. After oxidation of quercetin, four tautomeric forms of the oxidation
product can be formed. In the figure the tautomer which has an abundance of more than 99% is shown. When ascorbate and GSH are present in the
same concentration, oxidized quercetin reacts much faster with GSH than with ascorbate, thereby forming 6-GSH-quercetin and 8-GSH-quercetin.
Because of its high reactivity towards thiols, oxidized quercetin is also prone to react with protein thiols, as was seen in human blood plasma. This
reaction of oxidized quercetin is not prevented by ascorbate and can lead to toxicity. (C) The oxidation product formed out of monoHER is an ortho-
quinone. Ascorbate recycles this oxidation product to the parent compound monoHER, while GSH forms a conjugate with oxidized monoHER, i.e. 29-
GSH-monoHER. When both compounds are present in the same concentration, oxidized monoHER reacts rather with ascorbate (73%) than with GSH
(27%). The oxidized ascorbate formed in this recycling can be regenerated in the network, e.g. by dehydroascorbate reductase (DHAR) that uses
NADH as cofactor. Thus, the advantage of monoHER is that it can safely channel the non-specific reactivity of radicals toward ascorbate, which can be
regenerated in the antioxidant network.
The Antioxidant Network
PLoS ONE | www.plosone.org8November 2010 | Volume 5 | Issue 11 | e13880
diotoxicity when administered only once per week. Clin Cancer Res 6: Download full-text
16. van Acker SA, Kramer K, Grimbergen JA, van den Berg DJ, van der Vijgh WJ,
et al. (1995) Monohydroxyethylrutoside as protector against chronic doxorubi-
cin-induced cardiotoxicity. Br J Pharmacol 115: 1260–1264.
17. Willems AM, Bruynzeel AM, Kedde MA, van Groeningen CJ, Bast A, et al.
(2006) A phase I study of monohydroxyethylrutoside in healthy volunteers.
Cancer Chemother Pharmacol 57: 678–684.
18. Bruynzeel AM, Niessen HW, Bronzwaer JG, van der Hoeven JJ, Berkhof J, et al.
(2007) The effect of monohydroxyethylrutoside on doxorubicin-induced
cardiotoxicity in patients treated for metastatic cancer in a phase II study.
Br J Cancer 97: 1084–1089.
19. Bast A, Haenen GR (2002) The toxicity of antioxidants and their metabolites.
Environmental Toxicology and Pharmacology 11: 251–258.
20. Mukai K, Nishimura M, Kikuchi S (1991) Stopped-flow investigation of the
reaction of vitamin C with tocopheroxyl radical in aqueous triton X-100 micellar
solutions. The structure-activity relationship of the regeneration reaction of
tocopherol by vitamin C. J Biol Chem 266: 274–278.
21. Upston JM, Terentis AC, Stocker R (1999) Tocopherol-mediated peroxidation
of lipoproteins: implications for vitamin E as a potential antiatherogenic
supplement. Faseb J 13: 977–994.
22. Bowry VW, Ingold KU, Stocker R (1992) Vitamin E in human low-density
lipoprotein. When and how this antioxidant becomes a pro-oxidant. Biochem J
288 (Pt 2): 341–344.
23. Packer JE, Slater TF, Willson RL (1979) Direct observation of a free radical
interaction between vitamin E and vitamin C. Nature 278: 737–738.
24. Boots AW, Li H, Schins RP, Duffin R, Heemskerk JW, et al. (2007) The
quercetin paradox. Toxicol Appl Pharmacol 222: 89–96.
25. Jacobs H, van der Vijgh WJ, Koek GH, Draaisma GJ, Moalin M, et al. (2009)
Characterization of the glutathione conjugate of the semisynthetic flavonoid
monoHER. Free Radic Biol Med 46: 1567–1573.
26. Boots AW, Kubben N, Haenen GR, Bast A (2003) Oxidized quercetin reacts
with thiols rather than with ascorbate: implication for quercetin supplementa-
tion. Biochem Biophys Res Commun 308: 560–565.
27. Brubacher D, Moser U, Jordan P (2000) Vitamin C concentrations in plasma as
a function of intake: a meta-analysis. Int J Vitam Nutr Res 70: 226–237.
28. Pearson (1963) Hard and Soft Acids and Bases. J Am Chem Soc 85: 3533–3539.
29. Bast A, Haenen GR, Doelman CJ (1991) Oxidants and antioxidants: state of the
art. Am J Med 91: 2S–13S.
30. Haenen GR, Bast A (2002) The use of vitamin supplements in self-medication.
Therapie 57: 119–122.
31. Galati G, Moridani MY, Chan TS, O’Brien PJ (2001) Peroxidative metabolism
of apigenin and naringenin versus luteolin and quercetin: glutathione oxidation
and conjugation. Free Radic Biol Med 30: 370–382.
32. O’Brien PJ (1991) Molecular mechanisms of quinone cytotoxicity. Chem Biol
Interact 80: 1–41.
33. Yen GC, Duh PD, Tsai HL, Huang SL (2003) Pro-oxidative properties of
flavonoids in human lymphocytes. Biosci Biotechnol Biochem 67: 1215–1222.
34. Moore GA, Weis M, Orrenius S, O’Brien PJ (1988) Role of sulfhydryl groups in
benzoquinone-induced Ca2+ release by rat liver mitochondria. Arch Biochem
Biophys 267: 539–550.
35. Kalyanaraman B, Premovic PI, Sealy RC (1987) Semiquinone anion radicals
from addition of amino acids, peptides, and proteins to quinones derived from
oxidation of catechols and catecholamines. An ESR spin stabilization study.
J Biol Chem 262: 11080–11087.
36. van Acker SA, Towart R, Husken BC, De Jong J, van der Vijgh WJ, et al. (1993)
The protective effect of Venoruton and its main constituents on acute
doxorubicin-induced cardiotoxicity. Phlebology Suppl. 1: 31–32.
The Antioxidant Network
PLoS ONE | www.plosone.org9 November 2010 | Volume 5 | Issue 11 | e13880