HKOH-1: A Highly Sensitive and Selective Fluorescent Probe for Detecting Endogenous Hydroxyl Radicals in Living Cells

Article abstract of DOI:10.1002/anie.201705873

The hydroxyl radical (^.OH), one of the most reactive and deleterious reactive oxygen species (ROS), has been suggested to play an essential role in many physiological and pathological scenarios. However, a reliable and robust method to detect e

Full text of DOI:10.1002/anie.201705873

Angewandte
Communications
Chemie
International Edition: DOI: 10.1002/anie.201705873
German Edition: DOI: 10.1002/ange.201705873
Bioimaging
Hot Paper
HKOH-1: A Highly Sensitive and Selective Fluorescent Probe for
Detecting Endogenous Hydroxyl Radicals in Living Cells
Xiaoyu Bai, Yueyang Huang, Mingyang Lu, and Dan Yang*
Abstract: The hydroxyl radical (COH), one of the most reactive
and deleterious reactive oxygen species (ROS), has been
suggested to play an essential role in many physiological and
pathological scenarios. However, a reliable and robust method
to detect endogenous COH is currently lacking owing to its
extremely high reactivity and short lifetime. Herein we report
a fluorescent probe HKOH-1 with superior in vitro selectivity
and sensitivity towards COH. With this probe, we have
calibrated and quantified the scavenging capacities of a wide
range of reported COH scavengers. Furthermore, HKOH-1r,
which was designed for better cellular uptake and retention, has
performed robustly in detection of endogenous COH generation
by both confocal imaging and flow cytometry. Furthermore,
this probe has been applied to monitor COH generation in
HeLa cells in response to UV light irradiation. Therefore,
HKOH-1 could be used for elucidating COH related biological
functions.
distinguish it from other radicals. Furthermore, such methods
are not suitable for detecting COH in live cells.
Fluorescent probes are useful tools for detecting ROS in
live cells and tissues with extraordinary temporal and spatial
resolution. Indeed, there have been several reported fluores-
cent or luminescent probes for COH detection.[7] However, the
use of these probes has been doomed by their indirect
detection method, weak sensitivity or poor selectivity. Among
all of the reported COH probes, commercially available 2-[6-
(4’-hydroxy)phenoxy-3H-xanthen-3-on-9-yl]benzoic
acid
(HPF) is most widely used.[7a] It relies on an O-dearylation
of the masked fluorescein by an electron-rich phenol moiety.
However, HPF demonstrated poor selectivity over other
ROS, especially ONOOÀ (730-fold increase in fluorescence
intensity towards 10 equiv of Fentonꢀs reagent vs. a 120-fold
increase towards 0.3 equiv of ONOOÀ). Considering the low
concentration and high reactivity of COH in cells, it would be
difficult to confirm the real origin of fluorescence response in
live cell imaging. Some controversial results were obtained
when HPF was used as a COH probe and thiourea as a COH
scavenger.[8] Therefore the major challenges in developing
a desired COH fluorescence probe are to efficiently differ-
entiate COH from other ROS, especially strong oxidants like
HOCl and ONOOÀ, and competently capture intracellular
COH in the presence of various endogenous COH scavengers.
Herein we report the development of fluorescent probes
Compound A and HKOH-1 with excellent sensitivity, selec-
tivity, and extremely rapid turn-on response toward COH in
live cells in both confocal imaging and flow cytometry
experiments.
R
eactive oxygen species (ROS), including the hydroxyl
radical (COH), hypochlorous acid (HOCl), peroxynitrite
(ONOOÀ), superoxide (O2CÀ), nitric oxide (NO), singlet
oxygen (1O2), and hydrogen peroxide (H2O2), play essential
roles in various physiological and pathological processes.[1]
Among all ROS, COH is considered to be the most reactive
and harmful species. It has a short lifetime (approximately
10À9 s) and can react with numerous biomolecules such as
DNA bases, lipids, and proteins at a diffusion-controlled
rate.[2] Its excessive production leads to cell damage and has
been implicated in various diseases.[3] On the other hand,
more evidence suggests that generation of COH and other
ROS can be applied to cancer treatment.[4] Therefore,
monitoring intracellular COH is of paramount importance in
understanding its biological impact and further investigating
its therapeutic utilization. However, selective detection of
endogenous COH is highly challenging, given its short lifetime
and low concentration in cells. Over the past decades, several
methods to detect COH have been reported, for example,
those based on electron spin resonance (ESR) spectroscopy[5]
and aromatic hydroxylation reactions with salicylate,[6]
respectively. However, these methods are limited as they
are not sensitive enough towards COH and cannot selectively
Among the various highly reactive oxygen species
(hROS), the hydroxyl radical stands out from the rest,
namely HOCl and ONOOÀ, as a reactive radical species.
Based on previous reports, COH favors an electron transfer
reaction when reacts with a bulky phenol such as diiodophe-
nol to form a phenoxyl radical, which finally leads to
decomposition of aromatic compounds.[9] Thus, we designed
two novel probes, Compound A and HKOH-1 (Scheme 1), by
introducing two iodine atoms at the ortho position of the
phenolic hydroxyl group. While the diiodophenol would act as
a quencher of fluorescein or 2’,7’-dichlorofluorescein, the
[*] X. Bai, Y. Huang, M. Lu, Prof. D. Yang
Morningside Laboratory for Chemical Biology
Department of Chemistry, The University of Hong Kong
Pokfulam Road, Hong Kong (P. R. China)
E-mail: yangdan@hku.hk
Supporting information and the ORCID identification number(s) for
the author(s) of this article can be found under:
https://doi.org/10.1002/anie.201705873.
Scheme 1. Design of Compound A and HKOH-1.
Angew. Chem. Int. Ed. 2017, 56, 12873 –12877
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
12873

Angewandte
Communications
Chemie
steric hindrance of the two iodine atoms would block
oxidation by other strong oxidant like HOCl or ONOOÀ,
thus avoiding the selectivity problem of HPF. After oxidation
with COH, fluorescein or 2’,7’-dichlorofluorescein would be
released, leading to a “turn-on” of fluorescence.
Compound A and HKOH-1 were prepared in good yield
in three steps from commercially available starting materials
(Supporting Information, Schemes S1 and S2). The Fenton
reaction is the most common used method for COH generation
in vitro. Zhu and co-workers also reported that COH could
also be produced by reaction of tetrachloro-1,4-benzoquinone
(TCBQ) and H2O2.[10] For chemical characterization, we
applied both methods to investigate the reactivity of Com-
pound A and HKOH-1 towards COH. Compound A and
HKOH-1 showed absorption peaks at 490 nm and 500 nm,
respectively (Supporting Information, Figure S1). In absence
of COH, both probes were nonfluorescent; upon treatment
with 1 equiv of Fenton reagent ([H2O2]:[FeCl2] = 10:1) or
TCBQ/H2O2 ([H2O2]:[TCBQ] = 10:1), extremely rapid time-
dependent enhancement in fluorescence intensity was
observed (Figure 1a,b; Supporting Information, Fig-
ure S2a,b). HKOH-1 showed > 30-fold and > 180-fold
increases in fluorescence intensity upon treatment with
Fenton reagent and TCBQ/H2O2, respectively, while Com-
pound A showed > 40-fold and > 200-fold increases in
fluorescence intensity upon treatment with Fenton reagent
and TCBQ/H2O2, respectively. Both Compound A and
HKOH-1 completed the oxidation by Fenton reagent within
10 s, while the oxidation with TCBQ/H2O2 completed within
3 min yet with even stronger enhancement in fluorescence
intensity, illustrating the remarkable sensitivity and fast
response of these two probes.
Next, different equivalents of
Fenton reagent and TCBQ/H2O2
were added to the testing solution
of Compound A (Supporting Infor-
mation, Figure S2) and HKOH-
1 (Figure 1c–e). The fluorescence
intensity was found to increase
linearly with the concentration of
Fenton reagent (Supporting Infor-
mation, Figure S2; Figure 1d) and
the detection limit was as low as
490 nm and 390 nm (3s/k), respec-
tively, for Compound A and
HKOH-1. However, no such rela-
tionship was observed between
fluorescence intensity enhancement
and TCBQ/H2O2 concentration as
the generation mechanism of COH
from TCBQ/H2O2 is still not clear.
In spite of this, the results confirm
that our probes are highly sensitive
towards COH.
The selectivity of Compound A
and HKOH-1 was examined by
measuring the response upon treat-
ment against various analytes. As
shown in Figure 1 f, Compound A
displayed > 7-fold and > 40-fold
increases in fluorescence intensity
when treated with Fenton reagent
and TCBQ/H2O2, respectively, over
other ROS/RNS (HOCl, ONOO-,
O2CÀ, NO, 1O2, ROOC, TBHP, and
H2O2), FeCl2, FeCl3, and TCBQ
alone (Supporting Information, Fig-
ure S2). HKOH-1 exhibited > 20-
Figure 1. Characterization of HKOH-1 performance in chemical systems. Fluorescence spectra of
HKOH-1 (5 mm) in 0.1m potassium phosphate buffer (pH 7.4, 0.1% DMF) at 258C for 30 min.
a),b) Time course of the fluorescence response of HKOH-1 towards COH: a) Fenton reagent;
b) TCBQ/H2O2. c) Fluorescence emission spectra of HKOH-1 upon treatment with different amounts
of Fenton reagent. d) Fluorescence intensity of HKOH-1 as a function of concentration of Fenton
reagent (at 0.5–7 equiv). e) Fluorescence emission spectra of HKOH-1 upon treatment with different
amounts of TCBQ/H2O2. f) Fluorescence responses of the probe HKOH-1 toward various ROS/RNS.
The fluorescence spectra were measured at 520 nm with an excitation at 500 nm.
fold and > 100-fold increases in
fluorescence intensity when treated
with Fenton reagent and TCBQ/
H2O2, respectively, over other ROS/
RNS and analysts (Figure 1 f). Col-
lectively our probes, especially
HKOH-1, showed significantly
12874 www.angewandte.org
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2017, 56, 12873 –12877

Angewandte
Communications
Chemie
improved selectivity for COH over other ROS/RNS compared
with previously reported COH probes.
for endogenous COH generation, we detected the response of
HKOH-1 towards Fenton reagent in the presence of different
concentrations of scavengers. The results shown in Figure 2
implicated that the order of scavenging efficiency could be
described as follows: T4 > T3 > TEMPO > GSH ꢀ thiourea >
NAC @ DMSO. The high efficiency of T3 and T4 was
attributed to their direct reaction with COH as Compound A
and HKOH-1 do. Thus we have successfully confirmed and
compared the antioxidant activities of several representative
COH scavengers in vitro with our highly sensitive probe
HKOH-1.
The excellent performance of our probes, especially the
HKOH-1, in chemical system prompted us to explore its
biological applications. HKOH-1r was thus synthesized for
better cellular uptake and retention by introducing a dimethyl
diester group, which could be hydrolyzed in live cells by
various esterases. Cell viability study revealed that HKOH-1r
was not toxic even at 50 mm in both RAW264.7 mouse
macrophages and HeLa cells (Supporting Information, Fig-
ure S5).
For the reactions of Compound A and HKOH-1 with
COH, the fluorescent products were confirmed to be fluo-
rescein and dichlorofluorescein, respectively, by ESI-MS
analysis (Supporting Information, Figure S3). Furthermore,
the stability of HKOH-1 toward pH change was examined. As
shown in the Supporting Information, Figure S4, HKOH-
1 exhibited significant response across physiological pH range
(6.0–8.5) when treated with Fenton reagent. This result
implies the possibility of using HKOH-1 in detecting COH in
biological systems.
A wide range of chemical entities possess COH scavenging
abilities, for instance, DMSO,[11] thiourea,[12] glutathione
(GSH),[13] N-acetylcysteine (NAC),[14] 2,2,6,6-tetramethyl-1-
piperidinyloxy (TEMPO)[15] and thyroid hormones,[16] such as
triiodothyronine (T3) and its prohormone, thyroxine (T4),
among which a great proportion are biologically relevant.
However, a reliable and robust method to quantify their
reactivity is currently lacking, impeding the understanding of
various fundamental processes in chemical and biological
systems. With our probe HKOH-1 in hand, calibration and
quantification of the scavenging efficiency of reported
scavengers would therefore be possible, allowing further
investigation of the antioxidant capacity of cells. Considering
that the Fenton reaction is widely accepted to be responsible
To explore the potential of HKOH-1r for COH detection in
live cells, we first examined its performance in RAW264.7
mouse macrophages. PMA (phorbol myristate acetate),
a protein kinase C (PKC) activator, was applied to induce
endogenous COH. Cells were co-incubated with HKOH-1r
(5 mm) and PMA (500 ngmLÀ1) for 30 min. In contrast to the
control group, a robust fluorescence
signal enhancement was observed
upon treatment with PMA (Fig-
ure 3a; Supporting Information,
Figure S6). To confirm the fluores-
cence increase was a result of
endogenous COH generation, we
preloaded cells with 10 mm thio-
urea (a recognized hydroxyl radical
scavenger; see above) for 30 min
before treated with PMA. A re-
markable decrease in fluorescence
intensity was observed (Figure 3a;
Supporting
Information,
Fig-
ure S6), which indicates that the
alteration of fluorescence is a reflec-
tion of intracellular COH concentra-
tion change. Rapid time-lapse mon-
itoring of intracellular COH with
HKOH-1r confirmed that there
was no sign of photoactivating and
photobleaching (Supporting Infor-
mation, Figure S7).
To detect endogenous COH
quantitatively, we explored the ap-
plication of HKOH-1r in flow
cytometry. RAW264.7 cells were
co-incubated
with
HKOH-1r
(5 mm) with or without PMA
(500 ngmLÀ1) for 30 min, followed
by flow cytometry analysis (Fig-
ure 3b,c). PMA treatment induced
a significant fluorescence intensity
Figure 2. Investigating the COH scavenging ability of a) T4; b) T3; c) TEMPO; d) thiourea; e) GSH;
f) NAC and g) DMSO with HKOH-1 (5 mm) in 0.1m potassium phosphate buffer at pH 7.4. Hydroxyl
radical source: 5 mm Fenton reagent. The fluorescence spectra were recorded at 30 min with an
excitation at 500 nm and emission at 520 nm.
Angew. Chem. Int. Ed. 2017, 56, 12873 –12877
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 12875

Angewandte
Communications
Chemie
probe HKYellow-AM,[19] we successfully profiled
simultaneous COH and ONOOÀ generation in
HeLa cells. HKOH-1r and HKYellow-AM were
preloaded in HeLa cells for 30 min and then the
cells were exposed to UVA irradiation (365 nm)
for different periods. Compared to the untreated
control, robust fluorescence intensity enhance-
ments were observed in both channels of the UV
treatment group, and the fluorescence intensity
increased with UV irradiation time, which suggests
that both COH and ONOOÀ are generated in HeLa
cells upon UVA exposure (Figure 4; Supporting
Information, Figure S10). Meanwhile, COH scav-
enger thiourea and ONOOÀ scavenger urate
obviously inhibited the formation of COH and
ONOOÀ, respectively. Hence, with HKOH-1r,
endogenous COH generation in UV-irradiated
cancer cells can be readily detected, and together
with other probes like HKYellow-AM, we can
further investigate UVA-related biological pro-
cesses.
In summary, based on a selective O-dearyla-
tion reaction of 2,6-diiodophenol toward COH, we
have developed a highly sensitive and selective
probe HKOH-1 for monitoring COH formation in
live cells, with > 30-fold and > 75-fold increase in
fluorescence intensity upon treatment with Fenton
reagent or TCBQ/H2O2, respectively. We have also
successfully established the utility of HKOH-1r in
confocal imaging and flow cytometry by detecting
endogenous COH generated in PMA treated
RAW264.7 cells. Moreover, HKOH-1r could be
robustly applied in detecting COH generation in
UVA irradiated HeLa cells. Therefore, our probe
HKOH-1 is a powerful tool for the discovery of
COH related biological processes.
Figure 3. a) Detection of endogenous COH by HKOH-1r in RAW264.7 mouse macro-
phages in confocal imaging: a1) cells were incubated with HKOH-1r (5 mm) alone;
a2) cells were co-incubated with HKOH-1r (5 mm) and PMA (500 ngmLÀ1); a3) cells
were pre-treated with thiourea (10 mm) for 30 min before PMA (500 ngmLÀ1
)
stimulation. b),c) Detection of endogenous COH in flow cytometry. RAW264.7 cells
were co-incubated with HKOH-1r (5 mm) in the presence or absence of thiourea for
30 min. The unstained and untreated group were taken as control. b) The histogram
of fluorescence intensity changes towards PMA (500 ngmLÀ1) stimulation and
thiourea treatment. c) Relative fluorescence intensity increase calculated based on
the left histogram. Results are representative of at least three independent experi-
ments. Scale bar: 10 mm.
Acknowledgements
This work was supported by The University of
enhancement (> 20 fold) when compared with the untreated
group. Another group of cells were preloaded with thiourea
(10 mm) for 30 min, followed by PMA stimulation. The result
showed that thiourea efficiently blunted the fluorescence
increase. Overall, the FACS result showed that HKOH-1r
could be employed to detect endogenous COH in the flow
cytometry platform.
It is well-documented that UV irradiation from the
sunlight generates ROS, especially highly reactive COH and
ONOOÀ which induce skin damage and even skin cancers.
According to previous reports, UVA (320–400 nm) radiation
is more damaging than UVB (290–320 nm) to cells through
generation of ROS and RNS as delayed biological process-
es.[17,18] Furthermore, the COH generated contributes to DNA
degradation through both DNA base oxidation and DNA
single strand breaks. However, the mechanism of ROS
generation and downstream processes for each cell type is
different. Combined with the use of our established ONOOÀ
Hong Kong, the University Development Fund, Morningside
Foundation, and Hong Kong Research Grants Council under
General Research Fund Scheme (HKU 17305714) and Area
of Excellence Scheme (AoE/P-705/16). We thank Dr. Jing
Guo and Dr. Pei-Shin Emily Peng for technical assistance, the
HKU Li Ka Shing Faculty of Medicine Faculty Core Facility,
and the HKU School of Biological Sciences Central Facilities
Laboratory for support in confocal microscopy and flow
cytometry.
Conflict of interest
The authors declare no conflict of interest.
Keywords: bioimaging ·flow cytometry ·fluorescent probes ·
hydroxyl radicals ·live cells
12876 www.angewandte.org
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2017, 56, 12873 –12877

Angewandte
Communications
Chemie
[7] a) K.-i. Setsukinai, Y. Urano, K. Kakinuma, H. J. Majima, T.
Nagano, J. Biol. Chem. 2003, 278, 3170 –3175; b) N. Soh, K.
Makihara, E. Sakoda, T. Imato, Chem. Commun. 2004, 496 –497;
c) N. B. Yapici, S. Jockusch, A. Moscatelli, S. R. Mandalapu, Y.
Itagaki, D. K. Bates, S. Wiseman, K. M. Gibson, N. J. Turro, L.
Bi, Org. Lett. 2012, 14, 50 –53; d) M. Kim, S.-K. Ko, H. Kim, I.
Shin, J. Tae, Chem. Commun. 2013, 49, 7959 –7961; e) L. Meng,
Y. Wu, T. Yi, Chem. Commun. 2014, 50, 4843 –4845; f) Z. Li, T.
Liang, S. Lv, Q. Zhuang, Z. Liu, J. Am. Chem. Soc. 2015, 137,
11179 –11185; g) F. Liu, J. Du, D. Song, M. Xu, G. Sun, Chem.
Commun. 2016, 52, 4636 –4639; h) W. Zhou, Y. Cao, D. Sui, C.
Lu, Angew. Chem. Int. Ed. 2016, 55, 4236 –4241; Angew. Chem.
2016, 128, 4308 –4313.
[8] a) M. A. Kohanski, D. J. Dwyer, B. Hayete, C. A. Lawrence, J. J.
Collins, Cell 2007, 130, 797 –810; b) I. Keren, Y. Wu, J.
Inocencio, L. R. Mulcahy, K. Lewis, Science 2013, 339, 1213 –
1216; c) Y. Liu, J. A. Imlay, Science 2013, 339, 1210 –1213;
d) D. J. Dwyer, P. A. Belenky, J. H. Yang, I. C. MacDonald, J. D.
Martell, N. Takahashi, C. T. Y. Chan, M. A. Lobritz, D. Braff,
E. G. Schwarz, J. D. Ye, M. Pati, M. Vercruysse, P. S. Ralifo,
K. R. Allison, A. S. Khalil, A. Y. Ting, G. C. Walker, J. J. Collins,
Proc. Natl. Acad. Sci. USA 2014, 111, E2100 –E2109.
[9] X. Fang, H.-P. Schuchmann, C. von Sonntag, J. Chem. Soc.
Perkin Trans. 2 2000, 1391 –1398.
Figure 4. Detection of endogenous COH and ONOOÀ generation upon
UVA irradiation by HKOH-1r and HKYellow-AM in HeLa cells. a) Cells
were co-incubated with HKOH-1r (10 mm) and HKYellow-AM (10 mm)
without further treatment. b)–d) Cells were incubated with HKOH-1r
(10 mm) and HKYellow-AM (10 mm) for 30 min first and then washed
with HBSS buffer, followed by exposure to UVA (365 nm) for different
time periods: b) 5 min; c) 15 min; d) 30 min. e),f) Cells were pre-
treated with thiourea (10 mm) or urate (100 mm) for 30 min before
exposure to UVA. Scale bar: 10 mm. Results are representative of at
least three independent experiments.
[10] B.-Z. Zhu, B. Kalyanaraman, G.-B. Jiang, Proc. Natl. Acad. Sci.
USA 2007, 104, 17575 –17578.
[11] M. K. Eberhardt, R. Colina, J. Org. Chem. 1988, 53, 1071 –1074.
[12] M. Wasil, B. Halliwell, M. Grootveld, C. P. Moorhouse, D. C. S.
Hutchison, H. Baum, Biochem. J. 1987, 243, 867 –870.
[13] A. Yadav, P. C. Mishra, J. Mol. Model. 2013, 19, 767 –777.
[14] N. Agnihotri, P. C. Mishra, J. Phys. Chem. B 2009, 113, 12096 –
12104.
[15] H. Yu, L. Cao, F. Li, Q. Wu, Q. Li, S. Wang, Y. Guo, RSC Adv.
2015, 5, 63655 –63661.
[16] L. Oziol, P. Faure, C. Vergely, L. Rochette, Y. Artur, P. Chomard,
P. Chomard, J. Endocrinol. 2001, 170, 197 –206.
[17] J. Taira, K. Mimura, T. Yoneya, A. Hagi, A. Murakami, K.
Makino, J. Biochem. 1992, 111, 693 –695.
[18] J. Cadet, T. Douki, J.-L. Ravanat, Photochem. Photobiol. 2015,
91, 140 –155.
[19] T. Peng, X. Chen, L. Gao, T. Zhang, W. Wang, J. Shen, D. Yang,
Chem. Sci. 2016, 7, 5407 –5413.
How to cite: Angew. Chem. Int. Ed. 2017, 56, 12873–12877
Angew. Chem. 2017, 129, 13053–13057
[1] B. Halliwell, J. M. C. Gutteridge in Free Radicals in Biology and
Medicine, Oxford University Press, Oxford, 2015.
[2] a) H. Sies, Eur. J. Biochem. 1993, 215, 213 –219; b) H. Wiseman,
B. Halliwell, Biochem. J. 1996, 313, 17 –29; c) B. Balasubrama-
nian, W. K. Pogozelski, T. D. Tullius, Proc. Natl. Acad. Sci. USA
1998, 95, 9738 –9743; d) A. Ayala, M. F. MuÇoz, S. Argüelles,
Oxid. Med. Cell. Longev. 2014, 2014, 31.
[3] D. A. Allen, S. M. Harwood, M. Varagunam, M. J. Raftery,
M. M. Yaqoob, FASEB J. 2003, 17, 908 –910.
[4] J. L. Cleveland, M. B. Kastan, Nature 2000, 407, 309 –311.
[5] a) F. Scholz, G. López de Lara Gonzµlez, L. Machado de Car-
valho, M. Hilgemann, K. Z. Brainina, H. Kahlert, R. S. Jack,
D. T. Minh, Angew. Chem. Int. Ed. 2007, 46, 8079 –8081; Angew.
Chem. 2007, 119, 8225 –8227; b) B. E. Britigan, T. J. Coffman,
G. R. Buettner, J. Biol. Chem. 1990, 265, 2650 –2656.
Manuscript received: June 8, 2017
Accepted manuscript online: August 28, 2017
Version of record online: September 14, 2017
[6] M. Grootveld, B. Halliwell, Biochem. J. 1986, 237, 499 –504.
Angew. Chem. Int. Ed. 2017, 56, 12873 –12877
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 12877