An unexpected new pathway for nitroxide radical production via more reactve nitrogen-centered amidyl radical intermediate during detoxification of the carcinogenic halogenated quinones by N-alkyl hydroxamic acids

Article abstract of DOI:10.1016/j.freeradbiomed.2019.07.009

We found previously that nitroxide radical of desferrioxamine (DFO^?) could be produced from the interaction between the classic iron chelating agent desferrioxamine (DFO, an N-alkyl trihydroxamic acid) and tetrachlorohydroquinone (TCHQ), one of the carconogenic quinoind metabolites of the widely used wood preservative pentachlorophenol. However, the underlying molecular mechanism remains unclear. Here N-methylacetohydroxamic acid (N-MeAHA) was synthesized and used as a simple model compound of DFO for further mechanistic study. As expected, direct ESR studies showed that nitroxide radical of N-MeAHA (Ac-(CH₃)NO^?) can be produced from N-MeAHA/TCHQ. Interestingly and unexpectedly, when TCHQ was substituted by its oxidation product tetrachloro-1,4-benzoquinone (TCBQ), although Ac-(CH₃)NO^? could also be produced, no concurrent formation of tetrachlorosemiquinone radical (TCSQ^?) and TCHQ was detected, suggesting that Ac-(CH₃)NO^? did not result from direct oxidation of N-MeAHA by TCSQ^? or TCBQ as proposed previously. To our surprise, a new nitrogen-centered amidyl radical was found to be generated from N-MeAHA/TCBQ, which was observed by ESR with the spin-trapping agents and further unequivacally identified as Ac-(CH₃)N^? by HPLC-MS. The final product of amidyl radical was isolated and identified as its corresponding amine. Analogous radical homolysis mechanism was observed with other halogenated quinoid compounds and N-alkyl hydroxamic acids including DFO. Interestingly, amidyl radicals were found to induce both DNA strand breaks and DNA adduct formation, suggesting that N-alkyl hydroxamic acids may exert their potential side-toxic effects via forming the reactive amidyl radical species. This study represents the first report of an unexpected new pathway for nitroxide radical production via hydrogen abstration reaction of a more reactive amidyl radical intermediate during the detoxification of the carcinogenic polyhalogenated quinones by N-alkyl hydroxamic acids, which provides more direct experimental evidence to better explain not only our previous finding that excess DFO can provide effective but only partial protection against TCHQ (or TCBQ)-induced biological damage, and also the potential side-toxic effects induced by DFO and other N-alkyl hydroxamic acid drugs.

Full text of DOI:10.1016/j.freeradbiomed.2019.07.009

Accepted Manuscript
An unexpected new pathway for nitroxide radical production via more reactve
nitrogen-centered amidyl radical intermediate during detoxification of the carcinogenic
halogenated quinones by N-alkyl hydroxamic acids
Dan Xu, Chun-Hua Huang, Li Qin, Lin-Na Xie, Li Mao, Jie Shao, Balaraman
Kalyanaraman, Ben-Zhan Zhu
PII:
S0891-5849(19)30452-6
DOI:
https://doi.org/10.1016/j.freeradbiomed.2019.07.009
FRB 14339
Reference:
To appear in: Free Radical Biology and Medicine
Received Date: 23 March 2019
Revised Date: 10 July 2019
Accepted Date: 10 July 2019
Please cite this article as: D. Xu, C.-H. Huang, L. Qin, L.-N. Xie, L. Mao, J. Shao, B. Kalyanaraman,
B.-Z. Zhu, An unexpected new pathway for nitroxide radical production via more reactve nitrogen-
centered amidyl radical intermediate during detoxification of the carcinogenic halogenated quinones
by N-alkyl hydroxamic acids, Free Radical Biology and Medicine (2019), doi: https://doi.org/10.1016/
j.freeradbiomed.2019.07.009.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
An Unexpected New Pathway for Nitroxide Radical Production via More
Reactve Nitrogen-centered Amidyl Radical Intermediate during Detoxification of
the Carcinogenic Halogenated Quinones by N-alkyl Hydroxamic Acids
Dan Xu1,2, Chun-Hua Huang1,2,*, Li Qin1,2, Lin-Na Xie1,2, Li Mao1,2, Jie Shao1,2,
Balaraman Kalyanaraman3 and Ben-Zhan Zhu1,2,4,
*
1State Key Laboratory of Environmental Chemistry and Ecotoxicology, Research
Center for Eco-Environmental Sciences, The Chinese Academy of Sciences, Beijing
100085, P.R. China
2University of Chinese Academy of Sciences, Beijing 100049, P. R. China
3Department of Biophysics, Medical College of Wisconsin, Milwaukee, WI 53226,
USA
4Linus Pauling Institute, Oregon State University, Corvallis, OR 97331, USA
*To whom correspondence should be addressed. Tel: +86 10 62849030; Fax: +86 10
62923563; Email: bzhu@rcees.ac.cn; bjsyhch@126.com
1

ACCEPTED MANUSCRIPT
Abstract: We found previously that nitroxide radical of desferrioxamine (DFO•)
could be produced from the interaction between the classic iron chelating agent
desferrioxamine (DFO, an N-alkyl trihydroxamic acid) and tetrachlorohydroquinone
(TCHQ), one of the carconogenic quinoind metabolites of the widely used wood
preservative pentachlorophenol. However, the underlying molecular mechanism
remains unclear. Here N-methylacetohydroxamic acid (N-MeAHA) was synthesized
and used as a simple model compound of DFO for further mechanistic study. As
expected, direct ESR studies showed that nitroxide radical of N-MeAHA
(Ac-(CH3)NO•) can be produced from N-MeAHA/TCHQ. Interestingly and
unexpectedly, when TCHQ was substituted by its oxidation product
tetrachloro-1,4-benzoquinone (TCBQ), although Ac-(CH3)NO• could also be
produced, no concurrent formation of tetrachlorosemiquinone radical (TCSQ•) and
TCHQ was detected, suggesting that Ac-(CH3)NO• did not result from direct
oxidation of N-MeAHA by TCSQ• or TCBQ as proposed previously. To our surprise,
a new nitrogen-centered amidyl radical was found to be generated from
N-MeAHA/TCBQ, which was observed by ESR with the spin-trapping agents and
further unequivacally identified as Ac-(CH3)N• by HPLC-MS. The final product of
amidyl radical was isolated and identified as its corresponding amine. Analogous
radical homolysis mechanism was observed with other halogenated quinoid
compounds and N-alkyl hydroxamic acids including DFO. Interestingly, amidyl
radicals were found to induce both DNA strand breaks and DNA adduct formation,
suggesting that N-alkyl hydroxamic acids may exert their potential side-toxic effects
via forming the reavtive amidyl radical species. This study represents the first report
of an unexpected new pathway for nitroxide radical production via hydrogen
abstration reaction of a more reactive amidyl radical intermediate during the
detoxification of the carcinogenic polyhalogenated quinones by N-alkyl hydroxamic
acids, which provides more direct experimental evidence to better explain not only
our previous finding that excess DFO can provide effective but only partial protection
against TCHQ (or TCBQ)-induced biological damage, and also the potential
side-toxic effects induced by DFO and other N-alkyl hydroxamic acid drugs.
Keywords: N-centered amidyl radical; Nitroxide radical; Hydrogen-abstraction;
Tetrachloroquinoid compounds; N-alkyl hydroxamic acids
2

ACCEPTED MANUSCRIPT
Research Highlights
►
►
Nitroxide radical can be induced by N-alkyl hydroxamic acids/chloroquinones
Nitroxide radical was not from direct oxidation of hydroxamic acid by
chloroquinone
Nitroxide radical produced via H-abstraction from hydroxamic acid by amidyl
radical
The more reactive N-centered amidyl radicals can induce DNA damage
►
►
Graphic Abstract
3

ACCEPTED MANUSCRIPT
Introduction
Pentachlorophenol (PCP) has been a major industrial and agricultural biocide that
used primarily as a wood preservative. In addition, PCP was used to prevent snail
fever in China and other developing countries [1]. In 2016, PCP was classified as a
group 1 carcinogen by IARC since it can cause non-Hodgkin lymphoma in humans
[2]. Though its production and consumption are now restricted, the worldwide usage
and relative stability made PCP a ubiquitous environmental pollutant, which was
considered by UNEP as one of the new persistent organic pollutants (POPs) [3, 4].
Tetrachlorohydroquinone (TCHQ) and tetrachloro-1,4-benzoquinone (TCBQ) are
toxic metabolites or end products produced by PCP [5], which can create a variety of
hazardous effects in vivo, including acute hepatotoxicity, nephrotoxicity, and
carcinogenesis [5-7]. TCBQ itself has been widely used as a fungicide for treatment
of seeds and foliage, and as an oxidizing or dehydrating agent in organic synthesis.
Hydroxamic acids (HAs) have attracted considerable interest because of their
virous clinical applications including inhibition of a variety of enzymes such as
histone deacetylase, ureases, peroxidases and matrix metalloproteinases [8-12].
Desferrioxamine (DFO; Desferal; deferoxamine) is a linear trihydroxamic acid
siderophore that forms a stable octahedral coordination complex with Fe(III) (logβ =
31). DFO has been currently used clinically as the major iron chelator for the
treatment of iron overload [13, 14] and recently also found to be a potential anticancer
agent [15]. Apart from its ability to bind iron, DFO was regarded as an antioxidant
•-
through scavenging free radicals, including superoxide anion radical (O2 ), hydroxyl
radical (•OH) and peroxyl radical (ROO•) [16, 17]. Besides, DFO was also quite
susceptible to oxidation by one-electron oxidants like chelated copper [18],
peroxidases in the presence of hydrogen peroxide (H2O2) [19-21] and peroxynitrite
[22], and DFO-nitroxide radical (DFO•) has been detected in most of these cases.
Similar pathway also applied for the oxidation of other hydroxamic acids to their
corresponding nitroxide radicals [23-25].
Redox cycling of quinoid compounds is a well-known phenomenon. The cyclic
(auto) oxidation of TCHQ (Scheme S1) with the formation of intermediary TCSQ•
•-
•
and ultimate TCBQ can lead to formation of O2 , H2O2 and subsequent OH via
•
metal-mediated Haber-Weiss reaction [6, 26-28]. However, we found that OH (or
alkoxyl radical) and carbon-centered quinone ketoxy radical could be produced by
TCBQ in the presence of H2O2 or organic hydroperoxides through a
4

ACCEPTED MANUSCRIPT
metal-independent nucleophilic substitution followed by homolytical decomposition
mechanism, which can lead to DNA strand breakage, base oxidations, cyto- and
genotoxicity [5, 29-35].
Interestingly, we found that DFO and other HAs could markedly inhibit
TCHQ/TCBQ-induced hydroxyl/alkoxyl radical formation, DNA damage in vitro, and
cyto- and genotoxicity in human fibroblasts independent of their iron-chelating
properties [36-38]. We further found that TCSQ• formation during aotooxidation of
TCHQ can also be inhibited effectively by DFO and other HAs, with formation of the
less reactive DFO• [36, 37]. DFO and other HAs was found to remarkable accelerate
the hydrolysis of TCBQ to the much less reactive and almost nontoxic
2,5-dichloro-3,6-dihydroxy-1,4-benzoquonine (DDBQ) [36-38], which may partly
explain their detoxication effects. However, the molecular mechanism for reaction
between DFO with TCBQ and, in particular, the exact formation pathway for DFO•
remained to be a puzzle and difficult to elucidate due to the complex structure of DFO
(which is a N-alkyl trihydroxymic acid with a molecular weight at 657). In the present
study, N-methylacetohydroxamic acid (N-MeAHA), which possesses the key structure
of DFO (Fig. 1), was synthesized and used as a simpler model compound of DFO for
detailed mechanisistic study [39]. We will address the following questions: (i) Can
N-MeAHA protect the TCHQ-induced DNA damages? and can the corresponding
nitroxide radicals be produced from N-MeAHA with TCHQ or TCBQ? (ii) If so,
whether other free radical intermediates such as nitrogen-centered amidyl radical can
also be produced? (iii) What’s the underlying reaction mechanism for the formation of
nitroxide radicals? (iv) Is this a general molecular mechansim for all chlorinated
quinones and N-alkyl trihydroxymic acids? (v) What are the potential biological and
environmental implications?
N-MeBHA
N-CyBHA
N-MeAHA
N-BnBHA
DFO
Fig. 1. Chemical structures of N-alkylhydroxamic acids used in our study
Materials and Methods
Materials Plasmid pBR322 DNA was purchased from New England Biolabs. DFO,
5

ACCEPTED MANUSCRIPT
TCHQ, TCBQ, tetrafluoro-1,4-benzoquinone (TFBQ), tetrabromo-1,4-benzoquinone
(TBrBQ), 2,3,5-trichloro-1,4-benzoquinone (TrCBQ), 2,3-dichloro-1,4-benzoquinone
(2,3-DCBQ),
2,5-dichloro-1,4-benzoquinone
(2,5-DCBQ,
DCBQ),
2-chloro-1,4-benzoquinone (2-CBQ), N-methylacetamide (N-MeAA), deferoxamine
(DFO), cupric sulfate (CuSO4) and bathocuproine disulfonate (BCS),
N-t-Butyl-α-phenylnitrone (PBN), bovine serum albumin (BSA), catalase from
bovine liver (CAT), superoxide dismutase (SOD) from bovine erythrocytes were
purchased from Sigma-Aldrich. The nucleoside dG was purchased from TCI
(Shanghai, China). 5,5-Dimethyl-1-pyrroline-N-oxide (DMPO) was purchased from
Dojindo Molecular Technologies, Inc. (Japan). Trichlorohydroxy-1,4-benzoquinone
(TrCBQ-OH) was synthesized according to our previously published methods [29].
The chemicals were used as received without further purification. The chemicals were
used as received without further purification. All sodium phosphate buffer (PB)
solutions (100 mM or 10 mM, pH 7.4) or ammonium acetate （NH4Ac） buffer (100
mM, pH 7.4) were treated with Chelex overnight to remove adventitious metals.
Synthesis of N-methyl acetohydroxamic acid (N-MeAHA) N-MeAHA was
synthesized and purified according to the literature method [39]. Briefly, acetyl
chloride (7.85 g) was added drop-wise with stirring and ice-cooling to a mixture of
sodium carbonate (10.6 g) and N-methyl hydroxylamine hydrochloride (8.35 g) in
methanol (60 mL). After it had been stirred for 30 min, the mixture was filtered and
evaporated and the residue distilled under reduced pressure. The residue was purified
by silica gel chromatography with dichloromethane:tetrahydrofuran = 8:2 to get the
1
desired product. The H NMR spectra was recorded on a Bruker ARX 400
spectrometer using CDCl3 as solvent. 1H NMR δ = 2.07 (s, 3H), 3.26 (s, 3H), 8.97 (s,
1H). Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR-MS,
Bruker SolariX-15T) result showed that the synthesized N-MeAHA has the same
molecular mass (M + Na = 112.036852) and formula as the theoretical value (M + Na
= 112.036900, C3H7NNaO2).
Besides,
N-methylbenzohydroxamic
acid
(N-MeBHA),
N-cyclohexyl
benzohydroxamic acid (N-CyBHA) and N-benzylbenzohydroxamic acid (N-BnBHA)
were also synthesized and purified according to the above methods [39], except that
6

ACCEPTED MANUSCRIPT
the corresponding added reaction mixtures were acetyl chloride/N-benzyl
hydroxylamine hydrochloride, benzoyl chloride/N-cyclohexyl hydroxylamine and
benzoyl chloride/N-benzylhydroxylamine hydrochloride for the synthesis of
N-MeBHA, N-CyBHA and N-BnBHA, respectively.
Isolation and purification of 1:1 Adduct A by semi-preparative HPLC Adduct A
was prepared by reaction of 2,5-DCBQ (10 mM) and N-MeAHA (10 mM), and then
the reaction solution (5 mL) was injected into a semi-preparative HPLC column
(Waters XTERRA Prep C18, 19 × 150 mm, 5 µm) and isolated by an Agilent 1200
HPLC apparatus (Agilent Tech. Corp) equipped with a UV detector. The mobile
phase consisted of water/acetonitrile (60/40) in which the aqueous component
contained 50 mM acetic acid. The flow rate was 20 mL/min and the fractions with
retention time 4.3 min for Adduct A was collected manually. The collected fraction
was extracted with ethyl acetate, and then taken to vacuum evaporation to remove
ethyl acetate and further dryness under vacuum overnight. The 1H NMR spectra were
1
recorded on a Bruker ARX 400 spectrometer using CDCl3 as solvent. H NMR δ =
13
2.07 (s, 3H), 3.24 (s, 3H), 6.38 (s, 1H). C NMR δ = 20.62, 35.68, 109.55, 131.91,
145.61, 156.89, 174.11, 178.28, 178.98. Fourier transform ion cyclotron resonance
mass spectrometry (FT-ICR-MS, Bruker SolariX-15T) result showed that the Adduct
A has the same molecular mass (M + Na = 252.003407) and formula as the theoretical
value (M + Na = 252.003407, C9H8ClNNaO4).
DNA strand breakage Plasmid DNA agarose gel electrophoresis was used to
investigate DNA strand breakage induced by TCHQ and the protection effects by
N-MeAHA and DFO.
UV-Visible spectral analysis The autoxidation of TCHQ both in the presence and in
the absence of DFO or N-MeAHA were monitored in a UV-visible spectrophotometer
(Beckman DU-800), using PB (100 mM, pH 7.4) at room temperature. The generation
of TCSQ• was monitored at 455 nm.
Electron spin resonance (ESR) spin-trapping studies ESR spectra were recorded in
a Bruker EMX-plus-10/12 spectrometer at 9.8 GHz. Typical spectrometer parameters
were as follows: scan range, 100 G; field set, 3439 G; time constant, 328 ms; scan
7

ACCEPTED MANUSCRIPT
time, 29 s; modulation amplitude, 1.0 G; modulation frequency, 100 kHz; receiver
gain, 1.00×103; and microwave power, 20 mW. After getting the correct number of
interacting nuclei, spectra simulation and spectral parameters’ optimization were done
by Bruker Winsim Simulation Software.
Quantification of products produced by reaction of N-MeAHA and TCBQ
Yields of TrCBQ-OH and DDBQ by N-MeAHA and TCBQ were quantified using
authentic standards by HPLC (Agilent 1290 Infinity). Yield of N-MeAA was
accurately quantified using HPLC-ESI-TSQ-MS/MS (Thermo Scientific TSQ
Quantum Access MAX LC-MS) in SRM mode by the MS fragmentation of m/z
43.3→74.1 using authentic standards. The DMPO-N(CH3) adduct, hydrogen
abstraction products of DFO• and the dG adduct with Ac-(CH3)N• were identified by
HPLC-ESI-Q-TOF-MS (Agilent Q-TOF 6540). The Waters XTerra C18 (4.6 × 250
mm, 5 µm) was used as the analytical column. The mobile phase consisted of
water/acetonitrile (60/40) in which the aqueous component contained 50 mM acetic
acid. The flow rate was 1.0 mL/min. The eluate from the HPLC column was directly
introduced into an electrospray ionization mass spectrometer with flow-splitting. The
mass spectrometer was operated in the positive-ion mode, and collision energy was
performed at 10 eV. The fragmentor voltage was 90 V, nitrogen was used as nebulizer
gas, and the desolvation gas (nitrogen gas) was heated to 200 °C and delivered at a
flow rate of 9.0 L/min. The capillary voltage was set at 3500 V. The injection volume
was 10 µL for the reaction mixture. The data were collected by Agilent Q-TOF 6540
and Thermo Xcalibur Data Acquisition Workstation, respectively.
Results and Discussion
N-MeAHA can significantly inhibit the production of TCSQ• during TCHQ
autooxidation with the formation of N-MeAHA nitroxide radical (Ac-(CH3)NO•)
We found that both DFO and N-MeAHA could provide remarkable, but not full
protection against TCHQ- or TCBQ/H2O2-induced DNA strand breakage in isolated
plasmid DNA (pBR322) even when N-MeAHA was in excess (Fig. 2; Fig. S1a).
8

ACCEPTED MANUSCRIPT
Interestingly and unexpectedly, we observed that, the combination of N-MeAHA with
TCBQ (in the absence of H2O2) could synergistically induce obvious DNA
single-strand breaks，while neither of them alone has no effect (Fig. S1a). Based on
these results, we speculated that some quite reactive radical species might be
produced from the interaction between N-MeAHA and TCBQ, which may be
responsible for N-MeAHA/TCBQ-induced DNA damage.
Fig. 2. Protection against TCHQ-induced DNA strand breaks by DFO and N-MeAHA. The
reaction mixtures contained 5 µg/mL supercoiled circular pBR322 DNA (Form I), with
TCHQ (0.5 mM, lane 3 and 10) and different concentrations of N-MeAHA (4-7) and DFO
(11-14). All reactions were conducted in Chelex-treated phosphate buffer (100 mM, pH 7.4)
at 37◦C for 1h. The addition of chemicals was indicated as in the chart. Different DNA forms:
Form I, closed circular supercoiled DNA; Form II, relaxed open circle DNA; and Form III,
linear DNA. Each experiment was repeated at least twice.
Since TCSQ• has significant UV absorption at 453 nm, the generation and decay
of TCSQ• can be monitored at 453 nm by UV/Vis method. The concentration of
TCSQ• was found to reach maximal at approximately 15 min during the autooxidation
of TCHQ (Fig. S2a-b). Under the same conditions, when N-MeAHA or DFO was
added into the system, both the maximal concentration and the life span of the TCSQ•
were markedly reduced in concentration-dependent manner.
Interestingly, we found that when the formation of TCSQ• was inhibited by
N-MeAHA, a distinct twelve-line radical signal was produced concurrently, which can
be observed by direct ESR (Fig. 3a-b). Based on our previous study on DFO/TCHQ
reaction system, we speculated that this free radical signal should be the nitroxide
radical of N-MeAHA, i.e., Ac-(CH3)NO•. It has been reported that nitroxide radicals
9

ACCEPTED MANUSCRIPT
can be produced by one-electron oxidation of hydroxamic acids either chemically [23,
40, 41] or enzymatically [42]. Therefore, the typical one-electron oxidation reaction
of N-MeAHA with Cu(II)-bathocuproine disulfonate (BCS) complex was used to
further confirm the formation of Ac-(CH3)NO• (Fig. 3c). As expected, the same
twelve-line Ac-(CH3)NO• ESR signal was also observed from N-MeAHA with
Cu(II)(BCS)2. Further simulation studies showed that the hyperfine coupling constants
of Ac-(CH3)NO• were as follows: aN = 7.74 G, aH1 = aH2 = aH3 = 8.9 G (Fig. 3d). It
should be noted that since N-MeAHA contains only one hydroxamic acid group, its
inhibitory effect on TCSQ• was slightly weaker than the trihydroxamate DFO at the
same concentration.
Based on the above results that N-MeAHA showed similar effects with DFO on
both the inhibitory effect on TCSQ• and the generation of a nitroxide radical during its
reaction with TCHQ, we believe that N-MeAHA can be used as a suitable model
compound of DFO for further elucidating the underlying mechanism for the formation
of nitroxide radical.
(a) TCHQ 6 min
TCSQ·
/50
(b) N-MeAHA/TCHQ
6 min
6 min 30s
(c) N-MeAHA/Cu(BCS)2
(d) Simulation
3480
3490
3500
3510
3520
3530
3540
(G)
Fig. 3. ESR spectra of TCSQ• and the Ac-(CH3)NO• produced by incubation of TCHQ with
N-MeAHA. The central single line signal in the spectra (a) and (b) was identified as TCSQ•
(g=2.0058) and the twelve line signal in the spectra (b) and (c) was identified as Ac-(CH3)NO•.
The spectra were obtained in Chelex-treated phosphate buffer (100 mM, pH 7.4) at room
temperature. Samples contained (a) 1 mM TCHQ; (b) 5 mM N-MeAHA and 1 mM TCHQ; (c)
1 mM N-MeAHA and 1 mM Cu(BCS)2, respectively. (d): ESR spectra Simulation of
Ac-(CH3)NO•. Hyperfine splitting constant for Ac-(CH3)NO•: aN = 7.74 G, aH1 = aH2 = aH3
=
10

ACCEPTED MANUSCRIPT
8.9 G. Each experiment was repeated at least twice.
Ac-(CH3)NO• generated from N-MeAHA/TCBQ was found to be not due to the
one-electron oxidation of N-MeAHA by TCSQ• or TCBQ, i.e., it is not due to direct
scavenging of TCSQ• as previously proposed
As described above, hydroxamic acids are easily converted to the nitroxide
radical form when subjected to oxidation. Therefore, it is interesting to know whether
Ac-(CH3)NO• produced from N-MeAHA/TCHQ system was due to direct
one-electron oxidation of N-MeAHA by TCSQ• (i.e., direct scavenging of TCSQ•) as
previously proposed [36, 37]); If so, the corresponding reduction form of TCSQ•
should be TCHQ. However, we found that as Ac-(CH3)NO• was produced, both
TCSQ• (Fig. S2b) and TCHQ (Fig. S3) were consumed markedly, and TCHQ was
transformed
to
its
hydroxylation
products
including
and
the
initial
final
trichlorohydroxy-1,4-benzoquinone
(TrCBQ-OH)
2,5-dichloro-3,6-dihydroxy-1,4-benzoquinone (DDBQ). Then we wondered whether
N-MeAHA can be oxidized by TCBQ (a stronger oxidizing agent than TCSQ•) or by
other halogenated quinones (XQs) such as 2,5-dichlorobenzoquinone (DCBQ), to
generate Ac-(CH3)NO•. As expected, we found that this is indeed the case (Fig. S4). If
Ac-(CH3)NO• could be generated from direct one-electron oxidation of N-MeAHA by
TCBQ, then its reduction intermediate TCSQ• or product TCHQ, should also be
produced. However, neither TCSQ• nor TCHQ was produced during the reaction of
N-MeAHA with TCBQ as monitored by UV-vis method (at 453 nm for TCSQ•) (Fig.
S2c), or by HPLC-MS analysis for TCHQ. In addition, HPLC-MS results showed that
TrCBQ-OH was formed during the initial stage of reaction between N-MeAHA and
TCBQ (Fig. S5a), which then disappeared quickly as the reaction proceeds. Based on
our earlier findings on the reactions between TCBQ with benzohydroxamic acid
(BHA) [38] or hydroperoxides [29-31, 43] (H2O2 and tert-butylhydroperoxide), we
hypothesized that TrCBQ-OH should further react with excess N-MeAHA to produce
Ac-(CH3)NO• and final products, DDBQ, which was found to be true (Fig. S5b).
Therefore, based on the above results, we speculated that Ac-(CH3)NO• produced
from N-MeAHA/TCHQ or N-MeAHA/TCBQ should not arise from direct
one-electron oxidation by TCSQ• or TCBQ, i.e., it is not due to direct scavenging of
TCSQ• as previously proposed [36, 37], there should be other new pathway for its
11

ACCEPTED MANUSCRIPT
formation.
An unusual new nitrogen-centered radical was produced from the reaction between
TCBQ (DCBQ) and N-MeAHA as detected by ESR spin-trapping method
In order to investigate whether other free radical intermediates were produced
from TCBQ/N-MeAHA or DCBQ/N-MeAHA, the ESR spin-trapping method was
employed with 5,5-dimethyl-1-pyrroline N-oxide (DMPO) as the classic trapping
agent [44-47]. New ESR signals were observed once the reactions were initiated in
both TCBQ/N-MeAHA and DCBQ/N-MeAHA reaction systems (Fig. 4). The new
ESR signal was a unique 18-line signal, and further simulation studies (Fig. 4c)
suggested that it should be a nitrogen-centered radical adduct with DMPO, with the
following hyperfine coupling constants: aH = 21.62 G, aN1 = 14.86 G, aN2 = 2.37 G. To
further identify the nitrogen-centered radical, HLPC-MS method was employed.
When DMPO was added to TCBQ/N-MeAHA system, a new compound with
retention time at 2.8 min and mass-to-charge ratio (m/z) of 185 was formed. Both the
MS and MS/MS results were in agreement with the formation of DMPO-N(CH3)Ac
adduct (Fig. 5a). In summary, both ESR and HPLC-MS analysis clearly showed that
Ac-(CH3)N• radical intermediate was produced from both TCBQ/N-MeAHA and
N-MeAHA/DCBQ systems.
(a) DMPO/N-MeAHA/DCBQ
(b) DMPO/N-MeAHA/TCBQ
(c-e) Simulations：
(c)=70%(d)+30%(e)
(d) DMPO/•N(CH3)-Ac
O
Ac
N
(e) Ac-(CH3)NO•
CH3
(G)
3480
3500
3520
3540
Fig. 4. Experimental and simulative ESR spectra of DMPO-nitrogen-centered radicals and
Ac-(CH3)NO• radicals produced by N-MeAHA/TCBQ (DCBQ) in the presence of DMPO.
Experimental ESR spectra observed in DMPO/N-MeAHA/DCBQ system (a) and
12

ACCEPTED MANUSCRIPT
DMPO/N-MeAHA/TCBQ system (b) and combined ESR simulation spectrum (c) of
DMPO-nitrogen-centered radicals (d) and Ac-(CH3)NO• (e). The hyperfine coupling constants
for DMPO-nitrogen-centered radicals: aH = 21.62 G, aN1 = 14.86 G, aN2 = 2.37 G; and for
Ac-(CH3)NO•: aN = 7.74 G, aH1 = aH2 = aH3 = 8.9 G. The spectra were obtained 2 min after
mixing the indicated compounds in Chelex-treated phosphate buffer (100 mM, pH 7.4) at
room temperature. The reaction mixtures contained DMPO (100 mM), TCBQ or DCBQ (1
mM) and N-MeAHA (2 mM in (a), 1 mM in (b)). The addition of chemicals was indicated as
in the chart. Each experiment was repeated at least twice.
80k
HPLC-MS m/z: 185
(d)
143.1198
600k
185.1310
60k
(a)
RT=2.8 min
400k
200k
0
40k
20k
DMPO/N-MeAHA/TCBQ
M+H=185
DMPO/Adduct A
DMPO/N-MeAHA/DCBQ
0
3k
0
2
4
6
8
10
120
160
200
240
HPLC-MS m/z: 74
(e)
300k
200k
100k
0
RT=3.1 min
143.1197
(b)
2k
1k
N-MeAHA/TCBQ
-MeAHA/DCBQ
N
MSMS: 143
N
-MeAA Standard
6
111.0939
120
185.1310
0
0
2
4
8
10
90
150
180
210
80k
HPLC-MS m/z: 74
(f)
RT=3.1 min
74.0608
(c)
60k
40k
20k
0
300k
200k
100k
0
N-MeAHA/TCBQ
N-MeAHA/DCBQ
M+H=74
DMPO/N-MeAHA/TCBQ
DMPO/N-MeAHA/DCBQ
96.0438
90 100
70 m/z80
0
2
4
6
8
10
50
60
110
Rentention Time (min)
Fig. 5. Detection and identification of Ac-(CH3)N•/DMPO adduct (a) and N-MeAA (b-c) by
HPLC-MS. (a, d, e) The ESI-Q-TOF-MS spectra of the nitrone form of the
Ac-(CH3)N•/DMPO adduct with a retention time of 2.8 min in HPLC and m/z of 185 in MS;
(b, c f) The ESI-Q-TOF-MS spectra of N-MeAA with a retention time of 3.1 min in HPLC
and m/z of 74 in MS. The spectra were obtained in Chelex-treated ammonium acetate buffer
(100 mM, pH 7.4) at 37◦C for 3 h after the indicated mixing of 1 mM TCBQ, 1 mM DCBQ, 1
mM N-MeAHA, 1 mM N-MeAA, 1 mM adduct and 10 mM DMPO. Each experiment was
repeated at least twice.
13

ACCEPTED MANUSCRIPT
Nitroxide radical Ac-(CH3)NO• was unexpectedly found to be produced mainly from
hydrogen abstraction of the N-centered Ac-(CH3)N• radical from N-MeAHA
From the above results, we know that Ac-(CH3)N• radical was produced during
reaction of N-MeAHA and TCBQ or DCBQ. In the following kinetic studies, we
found that when N-MeAHA was in excess (N-MeAHA/TCBQ = 4:1), as the reaction
prolonged, the ESR signal intensity of N-centered Ac-(CH3)N• radical gradually
decreased, with the concurrent gradual increase of the intensity of nitroxide
Ac-(CH3)NO• radical (Fig. 6a, b). These results indicated that the production of
Ac-(CH3)NO• radical might be related to the formation of Ac-(CH3)N• radical. We
further found that the DMPO/•N(CH3)-Ac signal produced by N-MeAHA/TCBQ
system can be reduced dramatically by the addition of excess N-MeAHA. As the ratio
of N-MeAHA/TCBQ increased gradually from 1:1 to 10:1 during the same reaction
time, the DMPO/•N(CH3)-Ac signal decreased and finally disappeared, with the
concurrent formation of Ac-(CH3)NO• signal which increased gradually as the
concentration of N-MeAHA increased (Fig. 6c, 6d). These results indicate that the
generation of Ac-(CH3)NO• radical is in good correlation with not only the amidyl
radical production, but also the concentration of N-MeAHA.
To further examine the potential role of O2 and other reactive oxygen species
(O2 and derived H2O2) on the formation of Ac-(CH3)N• and Ac-(CH3)NO• in
•-
N-MeAHA/TCBQ system, we carried out the experiments under N2 and O2
atmosphere, or with the addition of superoxide dismutase (SOD), catalase (CAT) and
bovine serum albumin (BSA) for comparison. We found that compared with the
reactions in air, N2 bubbling had no obvious effect on the production of both
Ac-(CH3)N• and Ac-(CH3)NO• in N-MeAHA/TCBQ system (Fig. S6a, b). However,
O2 bubbling could slightly depressed the ESR peak heights of both Ac-(CH3)N• and
Ac-(CH3)NO• (Fig. S6a, b) and broadened the ESR line-widths of Ac-(CH3)NO• (Fig.
S7a). Interestingly, similar effect of O2 on Ac-(CH3)NO• was observed in the known
one-electron oxidation system N-MeAHA/Cu(II)-BCS (Fig. S7b). It has been reported
that increase of O2 in free radical-producing systems can broaden the ESR line-widths
of radicals and thus decrease the ESR signal intensity because of Heisenberg spin
exchange of paramagnetic O2 and radicals, and based on this theory, many radical
probes such as nitroxide radicals, carbon-centered radicals and also N-centered
radicals have been synthesized and applied for EPR oximetry technique [48]. Besides,
14

ACCEPTED MANUSCRIPT
O2 can also react with free radicals including carbon-centered radicals and N-
centered radicals [49, 50]. Therefore, we assume that the ESR signal changes of
Ac-(CH3)NO• and Ac-(CH3)N• caused by O2 bubbling may be due to interactions of
O2 with Ac-(CH3)NO• and Ac-(CH3)N•. In addition, we found that addition of SOD,
CAT or BSA did not affect the production of Ac-(CH3)N• and Ac-(CH3)NO• in
N-MeAHA/TCBQ system (Fig. S6). In conclusion, these results suggested that O2 and
other derived reactive oxygen species (superoxide radicals and derived H2O2) did not
contribute to nitroxide formation in N-MeAHA/TCBQ reaction system and the
possibility that a source of the Ac-(CH3)NO• by O2 addition to Ac-(CH3)N• could be
eliminated.
DMPO:N-MeAHA:TCBQ=100:4:1
6k
(a)
N-MeAHA:TCBQ=4:1
Reaction Time (min)
(b)
Amidyl radical (Peak 1)
Nitroxide radical (Peak 2)
Peak 1
5k
4k
3k
2k
1k
0
1.5 min
6 min
Peak 2
20 min
0
10
20
30
40
50
60
3470
3480
3490
3500
3510
(G)
3520
3530
3540
Reaction Time (min)
Nitroxide radical (Peak 2)
Amidyl radical (Peak 1)
12k
10k
8k
6k
4k
2k
0
(c)
(d)
8k
6k
4k
2k
0
N-MeAHA:TCBQ
N-MeAHA:TCBQ
1:1
2:1
3:1
4:1
5:1
10:1
1:1
2:1
3:1
4:1
5:1
10:1
0
10
20
30
40
50
60
70
0
10
20
30
40
50
60
Reaction Time (min)
Reaction Time (min)
Fig. 6. The formation and decay of the Ac-(CH3)N•/DMPO and Ac-(CH3)NO• was
dependent on the concentration of N-MeAHA and the reaction time. The ESR signal
intensity of amidyl radical and nitroxide radical were quantified by the height of Peak 1 and
Peak 2, respectively, as shown in (a). The spectra were obtained 1.5 min after the mixing of
DMPO (100 mM), TCBQ (1 mM) and N-MeAHA as indicated in Chelex-treated phosphate
buffer (100 mM, pH 7.4) at room temperature. Each experiment was repeated at least twice.
The hydrogen abstraction reaction of amidyl radical has been widely reported and
used in the organic sythesis [51]. Based on the above results and the chemistry of
amidyl radicals, we proposed that Ac-(CH3)NO• should be produced by the hydrogen
15

ACCEPTED MANUSCRIPT
abstraction reaction between Ac-(CH3)N• radical and excess N-MeAHA (Scheme 1).
According to Scheme 1, N-MeAA was also expected to be produced concomitantly
with Ac-(CH3)NO• via H-abstraction reaction of Ac-(CH3)N•. To further confirm this
hypothesis, HPLC-MS analysis was used to identify N-MeAA formation with the
authentic standard. We found that this is indeed the case (Fig. 5b & 5f). Interestingly,
the yield of N-MeAA was found to increase markedly with the increasing
concentrations of N-MeAHA during its reaction with both TCBQ and DCBQ (Fig.
S8a, b).
We also found that the addition of DMPO can decrease the yield of N-MeAA
dramatically (Fig. 5c), possibly because excessive DMPO can compete with
N-MeAHA to trap Ac-(CH3)N•. These results further demonstrated that N-MeAA was
produced via hydrogen abstraction of Ac-(CH3)N•.
Scheme 1. Nitroxide radical (Ac-(CH3)NO•) formation via hydrogen abstraction by
Ac-(CH3)N• radical from N-MeAHA
According to our previous studies [29-31, 38], we speculated that an unstable
adduct, i.e., trichloro-1,4-benzoquinone-O-N(CH3)-Ac (TrCBQ-O-N(CH3)-Ac),
should be formed via nucleophilic substitution reaction between N-MeAHA and
TCBQ. TrCBQ-O-N(CH3)-Ac then can decompose homolytically to produce amidyl
radical and quinone enoxy radical intermediates. However, due to its extreme
unstability, we couldn’t detect TrCBQ-O-N(CH3)-Ac by HPLC or HPLC-MS.
Therefore, in order to verify this hypothesis, the relative stable cojugation adduct,
Ac-N(CH3)-O-CBQ (for simplicity, this cojugation adduct was referred as Adduct A)
was isolated and purified from reaction of N-MeAHA with DCBQ (a lower
chlorinated benzoquinone) (Scheme 2) by semi-preparative HPLC seperation. If the
above hypothesis were correct, then the N-centered amidyl radical Ac-(CH3)N• and
oxygen-centered quinone enoxy radical CBQ-O• should be produced by direct
homolysis of N-O bond in Adduct A. Indeed, the same N-centered radical can be
detected from Adduct A (Fig. 7), which was furhter confirmed by HPLC-MS analysis
16

ACCEPTED MANUSCRIPT
as Ac-(CH3)N• (Fig. 5a). Based on these results, we confirmed that Ac-(CH3)N•
radicals could be produced directly via N-O homolysis of the conjugation adducts
TrCBQ-O-N(CH3)-Ac (or CBQ-O-N(CH3)-Ac) during reaction of N-MeAHA and
TCBQ (or DCBQ). According to Scheme 1, Ac-(CH3)NO• should be produced when
excessive N-MeAHA was added into Adduct A system, since Adduct A should
decompose to produce Ac-(CH3)N•. This was confirmed by ESR studies (Fig. S9): in
the reaction systems containing only Adduct A (without N-MeAHA), the decay of
amidyl radicals was much more slower and no nitroxide radicals were formed;
However, when N-MeAHA was added into Adduct A solution, DMPO/Ac-(CH3)N•
signal produced via decomposition of Adduct A was found to decrease quickly, while
the ESR signal of Ac-(CH3)NO• appeared and increased gradually. In addition, the
formation of H-abstration product of Ac-(CH3)N•, i.e., N-MeAA, was found to
increase with the addition of increasing concentrations of N-MeAHA into Adduct A
decomposition system (Fig. S8c). These results suggest that both the significant
decrease of amidyl radical and production of nitroxide radical were due to addition of
N-MeAHA into this system.
In conclusion, the above results fully confirmed that Ac-(CH3)NO• can be indeed
produced from hydrogen abstraction of Ac-(CH3)N• from N-MeAHA, thus completely
excluding the possibility that Ac-(CH3)NO• generated from N-MeAHA/TCBQ
(TCHQ) was due to the one-electron oxidation of N-MeAHA by TCSQ• or TCBQ, i.e.,
due to direct scavenging of TCSQ• as previously proposed (36, 37).
Scheme 2. Adduct A formed via nucleophilic reaction between N-MeAHA and DCBQ
17

ACCEPTED MANUSCRIPT
(a) DMPO/N-MeAHA/DCBQ=100:2:1
(b) DMPO/Adduct A=100:1
?
?
?
?
(c) Combined Simulation for (b)
82%(d)+18%(e)
(d) Single Simulation
DMPO-N radical
?
?
(e) Single Simulation
?
?
DMPO-radical adduct
3480
3490
3500
3510
3520
3530
3540
(G)
Fig. 7. ESR spectra of the radical species produced from homolysis of pure conjugation
adduct (Adduct A) formed from N-MeAHA/DCBQ. (a-b) ESR spectra of the radical species
produced in N-MeAHA/DCBQ reaction system (a) or Adduct A decomposition system (b)
in the presence of DMPO. (c-e) The ESR simulation of the radical species produced in (b),
i.e., Adduct A decomposition system. The hyperfine coupling constants of the radicals were
as follows: (d) DMPO-N centered radical: aH = 22.26 G, aN1 = 15.00 G, aN2 = 2.22 G; (e)
Unidentified DMPO radical adduct: aH = 14.69 G, aN = 14.87 G. The spectra were obtained 8
min after mixing the indicated compounds in Chelex-treated phosphate buffer (100 mM, pH
7.4) at room temperature. Simples contained 100 mM DMPO with 1 mM DCBQ and 2 mM
N-MeAHA or 1 mM Adduct A. Each spectrum was repeated at least twice.
The major pathway for the production of nitroxide radicals Ac-(CH3)NO• is via
H-abstraction of N-MeAHA by amidyl radical Ac-(CH3)N•, while the minor
pathway is via H-abstraction of N-MeAHA by enoxy radicals
Beside the strong and predominant ESR signal for DMPO/Ac-(CH3)N•, a minor
DMPO/radical adduct with weak ESR signal was also observed in both
N-MeAHA/DCBQ reaction system and Adduct A decomposition system (Fig. 7).
However, because of the interference by the strong DMPO/Ac-(CH3)N• signal, only a
weak 4-line ESR signal could be observed, and we assumed it to be a quartet signal
with an intensity ratio of 1:2:2:1 (aH = 14.69 G, aN = 14.87 G) based on simulation
results. According to our previously proposed mechanism for halogenated
18

ACCEPTED MANUSCRIPT
quinone-enhanced decomposition of hydroperoxides [31, 34] and the simulation
parameters, we expect that this unidentified 4-line ESR radical adduct might be the
DMPO adduct with the oxygen-centered enoxy radical CBQ-O•, which should be
produced concurrently with Ac-(CH3)N• by homolysis of N-O bond from molecular
Adduct A. If so, similar oxygen-centered enoxy radical TrCBQ-O• should also be
formed from N-MeAHA/TCBQ. Indeed, the experimental and simulation results
indicated a DMPO adduct with an undentified radical may be formed in
N-MeAHA/TCBQ (Fig. S10), which might be the oxygen-centered enoxy TrCBQ-O•.
However, the ESR intensity of possible oxygen-centered enoxy radical TrCBQ-O•
produced from N-MeAHA/TCBQ was much lower than CBQ-O• produced from
N-MeAHA/DCBQ, possibly due to the lower stability of TrCBQ-O• than the lowly
chlorinated CBQ-O•. Further investigations and more definitive evidence are needed
to fully characterize these minor radical species in our future research.
Since the detected ESR signal of enoxy radical was much lower than that of
amidyl radicals, in order to evaluate the H-abstraction capabilities of amidyl radicals
and enoxy radicals, we conducted some related computational work: 1) we found that
the H-abstraction reaction of N-MeAHA with Ac-(CH3)N• is thermodynamically more
favorable than that of TrCBQ-O• or CBQ-O• (145.9 kJ/mol, 81.8 kJ/mol, 100.1 kJ/mol,
respectively) (Fig. S11); 2) we also compared the bond dissociation energies (BDEs)
of the N-H bond and O-H bond for the responding H-abstraction products. We found
that BDE298 (N-H) in N-MeAA (432.8 kJ/mol) was much higher than the BDE298 (O-H)
in both TrCBQ-OH (371.9 kJ/mol) and CBQ-OH (390.2 kJ/mol) (Fig. S12). Since the
enoxy radicals can be stabilized by conjugated benzoquinone ring system, we believe
that the more reactive amidyl radicals are more likely to react with N-MeAHA than
enoxy radicals.
As we have discussed above (Fig, S8), we found that the formation of N-MeAA
increased markedly when N-MeAHA was added into N-MeAHA/TCBQ (or DCBQ)
reaction system or Adduct A, respectively. Besides the amidyl radials, we found
interstingly that the decay of enoxy radical CBQ-O• can also be accelerated by
addition of N-MeAHA to Adduct A decomposition system (Fig. S9d), suggesting that
H-abstraction of N-MeAHA by enoxy radicals may be also partly responsible for
Ac-(CH3)NO• generation.
Therefore, taken both the experimental and the computational results together, we
think that the major pathway for the production of nitroxide radicals Ac-(CH3)NO• is
19

ACCEPTED MANUSCRIPT
via H-abstraction of N-MeAHA by Ac-(CH3)N• radical, while the minor pathway is
via H-abstraction of N-MeAHA by enoxy radicals.
Molecular mechanism for Ac-(CH3)NO• formation from N-MeAHA/TCHQ via
hydrogen abstraction from N-MeAHA by Ac-(CH3)N•
Based on the above results, a unique mechanism for Ac-(CH3)NO• production from
N-MeAHA and TCHQ was proposed (Scheme 3). According to the mechanism:
TCHQ first generates TCSQ• by auto-oxidation, and then further oxidizes to TCBQ
(See Scheme S1). Then a nucleophilic reaction takes place between N-MeAHA and
TCBQ, forming a unstable transient intermediate of TrCBQ-O-N(CH3)-Ac, which can
decompose homolytically to produce Ac-(CH3)N• and TrCBQ-O•. The major pathway
for the production of less-reactive and less-toxic nitroxide radical Ac-(CH3)NO• is via
H-abstraction of N-MeAHA by Ac-(CH3)N• radical, while the minor pathway is via
H-abstraction of N-MeAHA by enoxy radicals. Meanwhile, TrCBQ-O• may also
disproportionate to form one of the major reaction product TrCBQ-OH, which can
react with N-MeAHA through a similar pathway to form DDBQ.
20

ACCEPTED MANUSCRIPT
O
O
O
Nucleophilic
Substitution
O
O
Cl
Cl
Cl
Cl
Cl
Cl
O
H
CH3
CH3
N
H C
+
N
3
Cl
CH3
HCl
O
O
TCBQ
N-MeAHA
N-O
Homolysis
O
H
N
N-MeAA
H3C
H3C
Major
O
CH3
Ac-(CH3)N
H3C
N
CH3
H-Abstraction
O
O
H
O
O
N
H3C
N
Ac-(CH3)NO
TrCBQ-OH
+
O
CH3
CH3
O
Cl
Minor
TrCBQ-O
O
H-Abstraction Cl
OH
Cl
Cl
Cl
O
Cl
O
Dismutation
O
O
O
O
Cl
Cl
O
N-MeAHA
TrCBQ-O
DDBQ
Cl
Cl
Cl
O
O
Scheme 3. Proposed new pathway for nitroxide radical production from N-MeAHA/TCBQ:
The major pathway for the production of less-reactive and less-toxic nitroxide radical
Ac-(CH3)NO• is via H-abstraction of N-MeAHA by amidyl radicals Ac-(CH3)N•,
while the minor pathway is via H-abstraction of N-MeAHA by enoxy radicals.
DFO nitroxide radical (DFO•) generated from DFO/TCBQ system was also found
to be produced via hydrogen abstraction of DFO by N-centered DFO amidyl radical
When N-MeAHA was replaced by DFO, however, we failed to detect the
corresponding N-centered DFO amidyl radical in DFO/TCBQ with DMPO and
another spin-trapping agent 5-t-butoxycarbonyl 5-methyl-1-pyrroline N-oxide
(BMPO). To our delight, when another spin-trapping agent PBN
(N-t-butyl-α-phenylnitrone) [52] was used, a PBN/N-centered radical was detected
from DFO/TCBQ (Fig. S13): at the beginning of the reaction of DFO/TCBQ with
PBN, a combined ESR signal of a N-centered radical and a DFO nitroxide radical
(DFO•) were detected (Fig. S13a), which agreed well with the simulation results (Fig.
21

ACCEPTED MANUSCRIPT
S13b-c). We speculated that the N-centered radical was probably the corresponding
N-centered amidyl radical from DFO. As reaction time prolonged, DFO• decreased
gradually, and only the relative stable PBN/N-centered radical was left 1 hour later
(Fig. S13d). As we have discussed above, we expect that the nitroxide radical DFO•
may also be formed from hydrogen abstraction of DFO amidyl radical from excessive
DFO. Due to the stronger stablility of PBN/N-centered radical adduct and the relative
instablity of DFO•, we could not observe DFO• formation in a concentration- and
time-dependent manner. Even so, however, we have detected the corresponding
hydrogen abstraction product of amidyl radical of DFO (Fig. S14). Therefore, these
results suggeste that, in general, DFO nitroxide radical DFO• generated from
DFO/TCBQ system should also be produced via hydrogen abstraction of DFO by
N-centered DFO amidyl radical.
The mechanism for nitroxide formation via hydrogen abstraction by N-centered
amidyl radical is universal for other halogenated quinones and N-alkyl hydroxamic
acids
Interestingly, when N-MeAHA was substituted by other N-alkyl hydroxamic
acids
including
DFO,
N-methylbenzohydroxamic
acid
(N-MeBHA),
N-cyclohexylbenzohydroxamic acid (N-CyBHA) and N-benzylbenzohydroxamic acid
(N-BnBHA), the corresponding nitroxyl radicals were also observed during the
reaction with TCBQ (Fig. S15) and Adduct A (Fig. S16). All these nitroxyl radicals
observed in ESR experiment are identical with the signal produced by one-electron
oxidation of the corresponding hydroxamic acids by Cu(II)-BCS complex (Fig. S17).
More importantly, we found that N-centered amidyl radicals can also be produced
from N-MeBHA/TCBQ (Fig. S18) and N-BnBHA/TCBQ systems by ESR spin
trapping method using DMPO and PBN as spin-trapping agents (Fig. S19). It is worth
noting that the stability of the nitroxide radicals and amidyl radicals vary considerably
among different N-alkyl substitution groups. For example, since the nitroxide radicals
and amidyl radicals produced from N-CyBHA and N-BnBHA are less stable than that
of N-MeAHA and N-MeBHA, the corresponding nitroxide radicals can only be
observed in solution with higher concentration of acetonitrile (40% ACN), and amidyl
radicals can not be detected in N-CyBHA/TCBQ system even after we tried different
spin-trapping agents. Interestingly, amidyl radicals and nitroxide radicals could also
be detected when TCBQ was substituted by other halogenated quinones including
22

ACCEPTED MANUSCRIPT
2-chloro-1,4-benzoquinone (2-CBQ), 2,3-dichloro-1,4-benzoquinone (2,3-DCBQ),
2,6-dichloro-1,4-benzoquinone (2,6-DCBQ), trichloro-1,4-benzoquinone (TriCBQ),
tetrabromo-1,4-benzoquinone (TBrBQ) and tetrafluoro-1,4-benzoquinone (TFBQ)
(Fig. S20). These results indicated that the mechanism for nitroxide formation via
hydrogen abstraction is universal in the reactions between all halogenated quinones
and N-alkyl hydroxamic acids.
N-centered amidyl radical was found to be responsible for DNA damage induced by
N-MeAHA/TCBQ
As we mentioned at the beginning, we found that even a 100-fold higher
concentration of N-MeAHA could not completely inhibited the DNA strand breaks
induced by TCBQ/H2O2 (Fig. S1a), and more surprisingly, N-MeAHA/TCBQ can
induce DNA stand breaks even in the absence of H2O2 (Fig. S1a). Based on these
results, we speculated that the much reactive amidyl radical, but not more stable
nitroxide radical, produced from N-MeAHA/TCBQ, may be responsible for this
damage. We found that this is indeed the case: not only N-MeAHA/TCBQ, but also
Adduct A, can induce clear DNA stand breaks; and more importantly, addition of
N-MeAHA and spin trapping agents like DMPO and PBN to both systems provided
strong inhibition (Fig. S1). In contrast, the nitroxide radical (but not its corresponding
amidyl radical), which can be produced from Cu(II)-BCS/N-MeAHA (DFO) systems,
induced no DNA damage as compared with the respective controls. These results
suggest that amidyl radical is the main reactive radical species responsible for
N-MeAHA/TCBQ-induced DNA damage.
Since guanine is the main DNA target due to its lowest ionization potential and
being the most susceptible to oxidation [52], we speculated that the reactive amidyl
radical may react with dG, leading to dG adduct formation. We found that addition of
dG into N-MeAHA/TCBQ/DMPO system can remarkably inhibit the ESR signal of
N-centered amidyl radicals (Fig. S21), and a dG-amidyl adduct was indeed detected
by HPLC-MS (Fig. S22). These results clearly showed that amidyl radicals can react
23

ACCEPTED MANUSCRIPT
with dG and further induce oxidative (via H-abstraction from dG) and covalent DNA
damage (via forming dG adduct).
Possible biological and environmental relevance
Though PCP was banned or restricted in many countries, PCP can now be
detected in the air, water, and soil throughout the world as well as in the blood, urine,
seminal fluid, breast milk, and adipose tissue of general human populations [53]. It
has reported that even low level of PCP exposure could cause various health risks
including a decrease in thyroid hormone levels and adverse effects on cognitive and
behavioral outcomes in children [54]. As metabolites of PCP, though the levels of
TCHQ and TCBQ formed in vivo may be lower than PCP, the toxicity of XQs has
been proved to be much stronger than chlorophenols. Toxicological studies in vivo
and in vitro have shown that XQs can lead to production of reactive oxygen species
(ROS) and further induce oxidative damage to DNA including the formation of
8-hydroxydeoxyguanosine (8-oxodG), DNA strand breaks, and apurinic/apyrimidinic
(AP) sites and glutathione depletion. XQs were also found to form DNA adducts,
affect genome-wide DNA methylation, and inhibit DNA repair enzymes [55-58].
Furthermore, twelve XQs including TCBQ have been detected as new disinfection
byproducts (DBPs) in chlorinated drinking water (0.5 to 275 ng/L) and in swimming
pool water (3.9 to 299 ng/L) [58], confirming XQs as DBPs with high occurrence
frequency. Epidemiological studies and quantitative structure-toxicity relationship
(QSTR) analysis have indicated that XQs have contributed to an increased risk of
developing bladder cancer linked with the consumption of chlorinated water [58, 59].
These findings suggest that XQs are highly cytotoxic, potentially genotoxic and
carcinogenic and need further research.
DFO has been widely used clinically for the treatment of iron overload
syndromes, and the administration concentration of 50 mg/kg (nearly 0.1 mM) body
weight per day appears safe in thalassaemic patients [60]. However, it has been
reported that DFO may cause different levels of side effects including ocular, auditory,
renal, pulmonary and cerebral toxicity as well as mucormycosis, which may be related
24

ACCEPTED MANUSCRIPT
the formation of DFO nitroxide radicals in vivo [61-64]. DFO nitroxide radicals have
been reported to react rapidly with ascorbate, tocopherol, and sulfhydryl groups, cause
protein damage, and inactivation of alcohol dehydrogenase and Ca2+-ATPase [61,
65-67]. Interestingly, besides DFO, another N-alkyl hydroxamic acid, i.e., N-methyl
benzohydroxamic acid (N-MeBHA), has also been shown to initiate lipid peroxidation
via forming its corresponding nitroxide radical [68, 69]. Therefore, the relative
reactive nitroxide radicals generated from oxidation of DFO and other N-alkyl
hydroxamic acids have been suggested to account for some side-toxic effects of DFO
[21, 68, 70], although a direct link between the production of DFO nitroxide radical
and the toxicity of DFO has not yet been proven.
In our present study, we found that even a 100-fold higher concentration of
N-MeAHA could not be completely inhibit the DNA strand breaks induced by
TCBQ/H2O2, and N-MeAHA/TCBQ can induce DNA stand breaks even in the
absence of H2O2. Our further study identified that the amidyl radical can cause direct
DNA damage (both DNA strand breaks and dG adducts). We found that DNA strand
breakage could still be observed when pBR322 DNA was exposed to very low
concentrations (as low as 5 µM) of N-MeAHA/TCBQ or Adduct A and the DNA
breakage became more severe when the concentrations were up to 50 µM (Fig. S23).
In addition, DMPO-amidyl radical and DMPO-enoxy radical adducts can also be
detected by ESR method even when the concentration of Adduct A was as low as 5
µM (Fig. S24). As mention above, since DFO concentration can reach up to 100 µM
in thalassaemic patients [60], we believe that our study should be potentially
biologically relevant. Therefore, besides nitroxide radical, we speculate that amidyl
radicals produced from chemically-activated hydroxamic acids may also be partly
responsible for the potential side effects of DFO and other hydroxamic acid drugs.
Further investigations are needed to study this hypothesis in our future research.
Acknowledgments
This work was supported by Strategic Priority Research Program of CAS Grant
(XDB01020300); NSF China Grants (21836005, 21577149, 21477139, 21621064,
21407163, 21777180); and National Institutes of Health (ES11497, RR01008 and
ES00210).
25

ACCEPTED MANUSCRIPT
References
[1] W.-w. Zheng, H. Yu, X. Wang, W.-d. Qu, Systematic review of pentachlorophenol occurrence in the
environment and in humans in China: Not a negligible health risk due to the re-emergence of
schistosomiasis, Environ. Int. 42 (2012) 105-116.
[2] K.Z. Guyton, D. Loomis, Y. Grosse, F. El Ghissassi, V. Bouvard, L. Benbrahim-Tallaa, N. Guha, H.
Mattock, K. Straif, Carcinogenicity of pentachlorophenol and some related compounds, Lancet Oncol.
17(12) (2016) 1637-1638.
[3] Toxicological profile for Pentachlorophenol (Update), in: P.H.S. Department of Health and Human
Services (Ed.) Agency for Toxic Substances and Disease Registry (ATSDR), Atlanta, GA: U.S. , 2001.
[4] Risk profile on pentachlorophenol and its salts and esters, in: U.N.E.P. (UNEP). (Ed.) Geneva,
2013.
[5] B.-Z. Zhu, G.-Q. Shan, Potential Mechanism for Pentachlorophenol-Induced Carcinogenicity: A
Novel Mechanism for Metal-Independent Production of Hydroxyl Radicals, Chem. Res. Toxicol. 22(6)
(2009) 969-977.
[6] J.L. Bolton, M.A. Trush, T.M. Penning, G. Dryhurst, T.J. Monks, Role of quinones in toxicology,
Chem. Res. Toxicol. 13(3) (2000) 135-160.
[7] Y. Song, B.A. Wagner, J.R. Witmer, H.-J. Lehmler, G.R. Buettner, Nonenzymatic displacement of
chlorine and formation of free radicals upon the reaction of glutathione with PCB quinones, Proc. Natl.
Acad. Sci. U. S. A. 106(24) (2009) 9725-9730.
[8] D. Pal, S. Saha, Hydroxamic acid–A novel molecule for anticancer therapy, J. Adv. Pharm. Technol.
Res. 3(2) (2012) 92.
[9] S.A. Gupta S.P., Hydroxamic Acids: A unique family of chemicals with multiple biological
activities, Springer, Berlin, Heidelberg, 2013.
[10] N.M. Archin, A.L. Liberty, A.D. Kashuba, S.K. Choudhary, J.D. Kuruc, A.M. Crooks, D.C. Parker,
E.M. Anderson, M.F. Kearney, M.C. Strain, D.D. Richman, M.G. Hudgens, R.J. Bosch, J.M. Coffin, J.J.
Eron, D.J. Hazuda, D.M. Margolis, Administration of vorinostat disrupts HIV-1 latency in patients on
antiretroviral therapy, Nature 487 (2012) 482.
[11] B. Li, J. Wu, W. Zhang, Z. Li, G. Chen, Q. Zhou, S. Wu, Synthesis and biological activity of
salinomycin-hydroxamic acid conjugates, Bioorg. Med. Chem. Lett. 27(7) (2017) 1624-1626.
[12] E.M. Muri, M.J. Nieto, R.D. Sindelar, J.S. Williamson, Hydroxamic acids as pharmacological
agents, Curr. Med. Chem. 9(17) (2002) 1631-1653.
[13] G.D. McLaren, W.A. Muir, R.W. Kellermeyer, Iron overload disorders: natural history,
pathogenesis, diagnosis, and therapy, Crit. Rev. Clin. Lab. Sci. 19(3) (1983) 205-266.
[14] N. Mobarra, M. Shanaki, H. Ehteram, H. Nasiri, M. Sahmani, M. Saeidi, M. Goudarzi, H.
Pourkarim, M. Azad, A Review on Iron Chelators in Treatment of Iron Overload Syndromes, Int J
Hematol Oncol Stem Cell Res 10(4) (2016) 239-247.
[15] Y. Yu, J. Wong, D.B. Lovejoy, D.S. Kalinowski, D.R. Richardson, Chelators at the cancer coalface:
desferrioxamine to Triapine and beyond, Clin. Cancer. Res. 12(23) (2006) 6876-6883.
[16] E.S.R. Green, H. Evans, P. Rice-Evans, M.J. Davies, N. Salah, C. Rice-Evans, The efficacy of
monohydroxamates as free radical scavenging agents compared with di- and trihydroxamates, Biochem.
Pharmacol. 45(2) (1993) 357-366.
[17] I. Morel, J. Cillard, G. Lescoat, O. Sergent, N. Pasdeloup, A.Z. Ocaktan, M.A. Abdallah, P. Brissot,
P. Cillard, Antioxidant and free radical scavenging activities of the iron chelators pyoverdin and
hydroxypyrid-4-ones in iron-loaded hepatocyte cultures: comparison of their mechanism of protection
26

ACCEPTED MANUSCRIPT
with that of desferrioxamine, Free Radic. Biol. Med. 13(5) (1992) 499-508.
[18] D.M. Van Reyk, R.T. Dean, The iron-selective chelator desferal can reduce chelated copper, Free
Radic. Res. 24(1) (1996) 55-60.
[19] J. Kanner, S. Harel, Desferrioxamine as an Electron Donor. Inhibition of Membranal Lipid
Peroxidation Initiated by H202–Activated Metmyoglobin and Other Peroxidizing Systems, Free Radic.
Res. Commun. 3(1-5) (1987) 309-317.
[20] K.M. Morehouse, W.D. Flitter, R.P. Mason, The enzymatic oxidation of Desferal to a nitroxide
free radical, FEBS Lett. 222(2) (1987) 246-250.
[21] M. Soriani, S. Mazzuca, V. Quaresima, M. Minetti, Oxidation of deseferrioxamine to nitroxide
free radical by activated human neutrophils, Free Radic. Biol. Med. 14(6) (1993) 589-599.
[22] A. Denicola, J.M. Souza, R.M. Gatti, O. Augusto, R. Radi, Desferrioxamine inhibition of the
hydroxyl radical-like reactivity of peroxynitrite: role of the hydroxamic groups, Free Radic. Biol. Med.
19(1) (1995) 11-9.
[23] D.F. Minor, W.A. Waters, J.V. Ramsbottom, The electron spin resonance spectra of some
hydroxylamine free radicals. Part III. Radicals formed by oxidations in alkaline solution, Journal of the
Chemical Society B: Physical Organic (0) (1967) 180-184.
[24] A. Forrester, M. Ogilvy, R. Thomson, Mode of action of carcinogenic amines. Part I. Oxidation of
N-arylhydroxamic acids, J. Chem. Soc. C (8) (1970) 1081-1083.
[25] M. Perkins, 4. One-Electron Oxidation of N-Substituted Hydroxamic Acids: Structure and
Chemistry of Acyl Nitroxides, in: H. Kehl (Ed.), Chemistry and Biology of Hydroxamic Acids, Karger
Publishers1982, pp. 29-36.
[26] C.P. Carstens, J.K. Blum, I. Witte, The role of hydroxyl radicals in tetrachlorohydroquinone
induced DNA strand break formation in PM2 DNA and human fibroblasts, Chem. Biol. Interact. 74(3)
(1990) 305-314.
[27] P.J. O'Brien, Molecular mechanisms of quinone cytotoxicity, Chem.-Biol. Interact. 80(1) (1991)
1-41.
[28] G.J. Halliwell B, Free Radicals in Biology and Medicine, Oxford University Press, Oxford, 2007.
[29] B.-Z. Zhu, B. Kalyanaraman, G.-B. Jiang, Molecular mechanism for metal-independent
production of hydroxyl radicals by hydrogen peroxide and halogenated quinones, Proc. Natl. Acad. Sci.
U. S. A. 104(45) (2007) 17575-175758.
[30] B.-Z. Zhu, H.T. Zhao, B. Kalyanaraman, J. Liu, G.-Q. Shan, Y.-G. Du, B. Frei, Mechanism of
metal-independent decomposition of organic hydroperoxides and formation of alkoxyl radicals by
halogenated quinones, Proc. Natl. Acad. Sci. U. S. A. 104(10) (2007) 3698-3702.
[31] B.-Z. Zhu, G.-Q. Shan, C.-H. Huang, B. Kalyanaraman, L. Mao, Y.-G. Du, Metal-independent
decomposition of hydroperoxides by halogenated quinones: detection and identification of a quinone
ketoxy radical, Proc. Natl. Acad. Sci. U. S. A. 106(28) (2009) 11466-11471.
[32] H. Qin, C.-H. Huang, L. Mao, H.-Y. Xia, B. Kalyanaraman, J. Shao, G.-Q. Shan, B.-Z. Zhu,
Molecular mechanism of metal-independent decomposition of lipid hydroperoxide 13-HPODE by
halogenated quinoid carcinogens, Free Radic. Biol. Med. 63 (2013) 459-466.
[33] J. Shao, C.H. Huang, B. Kalyanaraman, B.Z. Zhu, Potent methyl oxidation of
5-methyl-2'-deoxycytidine by halogenated quinoid carcinogens and hydrogen peroxide via a
metal-independent mechanism, Free Radic. Biol. Med. 60 (2013) 177-182.
[34] C.-H. Huang, G.-Q. Shan, L. Mao, B. Kalyanaraman, H. Qin, F.-R. Ren, B.-Z. Zhu, The first
purification and unequivocal characterization of the radical form of the carbon-centered quinone
27

ACCEPTED MANUSCRIPT
ketoxy radical adduct, Chem. Commun. 49(57) (2013) 6436-6438.
[35] C.-H. Huang, F.R. Ren, G.-Q. Shan, H. Qin, L. Mao, B.-Z. Zhu, Molecular mechanism of
metal-independent decomposition of organic hydroperoxides by halogenated quinoid carcinogens and
the potential biological implications, Chem. Res. Toxicol. 28(5) (2015) 831-837.
[36] B.-Z. Zhu, R. Har-El, N. Kitrossky, M. Chevion, New modes of action of desferrioxamine:
scavenging of semiquinone radical and stimulation of hydrolysis of tetrachlorohydroquinone, Free
Radic. Biol. Med. 24(2) (1998) 360-369.
[37] I. Witte, B.-Z. Zhu, A. Lueken, D. Magnani, H. Stossberg, M. Chevion, Protection by
desferrioxamine and other hydroxamic acids against tetrachlorohydroquinone-induced cyto- and
genotoxicity in human fibroblasts, Free Radic. Biol. Med. 28(5) (2000) 693-700.
[38] B.Z. Zhu, J.G. Zhu, L. Mao, B. Kalyanaraman, G.Q. Shan, Detoxifying carcinogenic
polyhalogenated quinones by hydroxamic acids via an unusual double Lossen rearrangement
mechanism, Proc. Natl. Acad. Sci. U. S. A. 107(48) (2010) 20686-20690.
[39] H. Ulrich, A.A.R. Sayigh, 202. Hydroxyamino-derivatives from formaldehyde. Their reaction with
acyl halides, J Chem Soc (Resumed) (0) (1963) 1098-1101.
[40] E. Boyland, R. Nery, The oxidation of hydroxamic acids, J. Chem. Soc. C (0) (1966) 354-358.
[41] A. Samuni, S. Goldstein, One-Electron Oxidation of Acetohydroxamic Acid: The Intermediacy of
Nitroxyl and Peroxynitrite, J Phys Chem A 115(14) (2011) 3022-3028.
[42] C.L. Ritter, D. Malejka-Giganti, C.F. Polnaszek, Cytochrome c/H2O2-mediated one electron
oxidation of carcinogenic N-fluorenylacetohydroxamic acids to nitroxyl free radicals, Chem.-Biol.
Interact. 46(3) (1983) 317-334.
[43] B.Z. Zhu, L. Mao, C.H. Huang, H. Qin, R.M. Fan, B. Kalyanaraman, J.G. Zhu, Unprecedented
hydroxyl radical-dependent two-step chemiluminescence production by polyhalogenated quinoid
carcinogens and H2O2, Proc. Natl. Acad. Sci. U. S. A. 109(40) (2012) 16046-16051.
[44] K.P. Madden, H. Taniguchi, The role of the DMPO-hydrated electron spin adduct in DMPO-•OH
spin trapping, Free Radic. Biol. Med. 30(12) (2001) 1374-1380.
[45] N.A. Bauer, E. Hoque, M. Wolf, K. Kleigrewe, T. Hofmann, Detection of the formyl radical by
EPR spin-trapping and mass spectrometry, Free Radic. Biol. Med. 116 (2018) 129-133.
[46] C.L. Hawkins, M.J. Davies, The role of reactive N-bromo species and radical intermediates in
hypobromous acid-induced protein oxidation, Free Radic. Biol. Med. 39(7) (2005) 900-912.
[47] S. Bhattacharjee, L.J. Deterding, S. Chatterjee, J. Jiang, M. Ehrenshaft, O. Lardinois, D.C.
Ramirez, K.B. Tomer, R.P. Mason, Site-specific radical formation in DNA induced by Cu(II)-H2O2
oxidizing system, using ESR, immuno-spin trapping, LC-MS, and MS/MS, Free Radic. Biol. Med.
50(11) (2011) 1536-1545.
[48] R. Ahmad, P. Kuppusamy, Theory, instrumentation, and applications of electron paramagnetic
resonance oximetry, Chem. Rev. 110(5) (2010) 3212-3236.
[49] B. Maillard, K.U. Ingold, J.C. Scaiano, Rate constants for the reactions of free radicals with
oxygen in solution, J. Am. Chem. Soc. 105(15) (1983) 5095-5099.
[50] P.M. Gannett, X. Shi, T. Lawson, C. Kolar, B. Toth, Aryl radical formation during the metabolism
of arylhydrazines by microsomes, Chem. Res. Toxicol. 10(12) (1997) 1372-1377.
[51] A. Martín, I. Pérez-Martín, E. Suárez, Intramolecular hydrogen abstraction promoted by amidyl
radicals. Evidence for electronic factors in the nucleophilic cyclization of ambident amides to
oxocarbenium ions, Org. Lett. 7(10) (2005) 2027-2030.
[52] J. Cadet, T. Douki, J.-L. Ravanat, Oxidatively Generated Damage to the Guanine Moiety of DNA:
28

ACCEPTED MANUSCRIPT
Mechanistic Aspects and Formation in Cells, Acc. Chem. Res. 41(8) (2008) 1075-1083.
[53] H.J. Geyer, I. Scheunert, F. Korte, Distribution and bioconcentration potential of the
environmental chemical pentachlorophenol (PCP) in different tissues of humans, Chemosphere 16(4)
(1987) 887-899.
[54] W.W. Zheng, X. Wang, H. Yu, X.G. Tao, Y. Zhou, W.D. Qu, Global Trends and Diversity in
Pentachlorophenol Levels in the Environment and in Humans: A Meta-Analysis, Environ. Sci. Technol.
45(11) (2011) 4668-4675.
[55] R. Huang, W. Wang, Y. Qian, J.M. Boyd, Y. Zhao, X.F. Li, Ultra pressure liquid
chromatography-negative electrospray ionization mass spectrometry determination of twelve
halobenzoquinones at ng/L levels in drinking water, Anal. Chem. 85(9) (2013) 4520-4529.
[56] R. Yin, D. Zhang, Y. Song, B.-Z. Zhu, H. Wang, Potent DNA damage by polyhalogenated
quinones and H2O2 via a metal-independent and Intercalation-enhanced oxidation mechanism, Sci. Rep.
3 (2013) 1269.
[57] B. Zhao, Y. Yang, X. Wang, Z. Chong, R. Yin, S.-H. Song, C. Zhao, C. Li, H. Huang, B.-F. Sun, D.
Wu, K.-X. Jin, M. Song, B.-Z. Zhu, G. Jiang, J.M. Rendtlew Danielsen, G.-L. Xu, Y.-G. Yang, H. Wang,
Redox-active quinones induces genome-wide DNA methylation changes by an iron-mediated and
Tet-dependent mechanism, Nucleic Acids Res. 42(3) (2014) 1593-1605.
[58] J. Li, W. Wang, B. Moe, H. Wang, X.F. Li, Chemical and toxicological characterization of
halobenzoquinones, an emerging class of disinfection byproducts, Chem. Res. Toxicol. 28(3) (2015)
306-318.
[59] C.M. Villanueva, K.P. Cantor, J.O. Grimalt, N. Malats, D. Silverman, A. Tardon, R. Garcia-Closas,
C. Serra, A. Carrato, G. Castano-Vinyals, R. Marcos, N. Rothman, F.X. Real, M. Dosemeci, M.
Kogevinas, Bladder cancer and exposure to water disinfection by-products through ingestion, bathing,
showering, and swimming in pools, Am. J. Epidemiol. 165(2) (2007) 148-156.
[60] B. Halliwell, Protection against tissue damage in vivo by desferrioxamine: What is its mechanism
of action?, Free Radic. Biol. Med. 7(6) (1989) 645-651.
[61] M.J. Davies, R. Donkor, C.A. Dunster, C.A. Gee, S. Jonas, R.L. Willson, Desferrioxamine
(Desferal) and superoxide free radicals. Formation of an enzyme-damaging nitroxide, Biochem. J.
246(3) (1987) 725-729.
[62] G. Koren, Y. Bentur, D. Strong, et al., Acute changes in renal function associated with
deferoxamine therapy, Am. J. Dis. Child. 143(9) (1989) 1077-1080.
[63] M. Tenenbein, S. Kowalski, A. Sienko, D.H. Bowden, I.Y.R. Adamson, M. Tenenbein, Pulmonary
toxic effects of continuous desferrioxamine administration in acute iron poisoning, The Lancet
339(8795) (1992) 699-701.
[64] J.R. Boelaert, M. de Locht, Side-effects of desferrioxamine in dialysis patients, Nephrol Dial
Transplant 8(supp1) (1993) 43-46.
[65] M. Kiyose, C.I. Lee, E. Okabe, Inhibition of skeletal sarcoplasmic reticulum Ca2+-ATPase
activity by deferoxamine nitroxide free radical, Chem. Res. Toxicol. 12(2) (1999) 137-143.
[66] A. Mordente, E. Meucci, G.A. Miggiano, G.E. Martorana, Prooxidant action of desferrioxamine:
enhancement of alkaline phosphatase inactivation by interaction with ascorbate system, Arch. Biochem.
Biophys. 277(2) (1990) 234-240.
[67] M. Tada, Y. Niwano, M. Kohno, Generation Mechanism of Deferoxamine Radical by
Tyrosine-Tyrosinase Reaction, Anal. Sci. 31(9) (2015) 911-916.
[68] S. Bergamini, C. Rota, M. Staffieri, A. Tomasi, A. Iannone, Prooxidant activity of ferrioxamine in
29