Hydrogen peroxide is the major oxidant product of xanthine oxidase

Article abstract of DOI:10.1016/j.freeradbiomed.2009.11.012

Xanthine oxidase (XO) is a critical source of reactive oxygen species (ROS) in inflammatory disease. Focus, however, has centered almost exclusively on XO-derived superoxide (O₂^?-), whereas direct H₂O₂ production from XO has been less well investigated. Therefore, we examined the relative quantities of O₂^?- and H₂O₂ produced by XO under a range (1-21%) of O₂ tensions. At O₂ concentrations between 10 and 21%, H₂O₂ accounted for ～75% of ROS production. As O₂ concentrations were lowered, there was a concentration-dependent increase in H₂O₂ formation, accounting for 90% of ROS production at 1% O₂. Alterations in pH between 5.5 and 7.4 did not affect the relative proportions of H₂O₂ and O₂^?- formation. Immobilization of XO, by binding to heparin-Sepharose, further enhanced relative H₂O₂ production by ～30%, under both normoxic and hypoxic conditions. Furthermore, XO bound to glycosaminoglycans on the apical surface of bovine aortic endothelial cells demonstrated a similar ROS production profile. These data establish H₂O₂ as the dominant (70-95%) reactive product produced by XO under clinically relevant conditions and emphasize the importance of H₂O₂ as a critical factor when examining the contributory roles of XO-catalyzed ROS in inflammatory processes as well as cellular signaling.

Full text of DOI:10.1016/j.freeradbiomed.2009.11.012

Free Radical Biology & Medicine 48 (2010) 493–498
Contents lists available at ScienceDirect
Free Radical Biology & Medicine
journal homepage: www.elsevier.com/locate/freeradbiomed
Original Contribution
Hydrogen peroxide is the major oxidant product of xanthine oxidase
b
a
b
Eric E. Kelley a, , Nicholas K.H. Khoo , Nicholas J. Hundley , Umair Z. Malik ,
⁎
Bruce A. Freeman b, Margaret M. Tarpey a,c,
Department of Anesthesiology, University of Pittsburgh School of Medicine, Pittsburgh, PA 15213, USA
⁎
a
b
Department of Pharmacology and Chemical Biology, University of Pittsburgh School of Medicine, Pittsburgh, PA 15213, USA
c
Pittsburgh VA Medical Center, Pittsburgh, PA15213, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Xanthine oxidase (XO) is a critical source of reactive oxygen species (ROS) in inflammatory disease. Focus,
however, has centered almost exclusively on XO-derived superoxide (O2•−), whereas direct H2O2 production
Received 3 October 2009
Revised 6 November 2009
Accepted 15 November 2009
Available online 24 November 2009
•−
from XO has been less well investigated. Therefore, we examined the relative quantities of O2 and H2O2
produced by XO under a range (1–21%) of O2 tensions. At O2 concentrations between 10 and 21%, H2O2
accounted for ∼75% of ROS production. As O2 concentrations were lowered, there was a concentration-
dependent increase in H2O2 formation, accounting for 90% of ROS production at 1% O2. Alterations in pH
Keywords:
•−
between 5.5 and 7.4 did not affect the relative proportions of H2O2 and O2 formation. Immobilization of
Xanthine oxidase
Reactive oxygen species
Hydrogen peroxide
Oxygen
Superoxide
Hypoxia
XO, by binding to heparin–Sepharose, further enhanced relative H2O2 production by ∼30%, under both
normoxic and hypoxic conditions. Furthermore, XO bound to glycosaminoglycans on the apical surface of
bovine aortic endothelial cells demonstrated a similar ROS production profile. These data establish H2O2 as
the dominant (70–95%) reactive product produced by XO under clinically relevant conditions and emphasize
the importance of H2O2 as a critical factor when examining the contributory roles of XO-catalyzed ROS in
inflammatory processes as well as cellular signaling.
Free radicals
Published by Elsevier Inc.
The molybdoflavin enzyme xanthine oxidoreductase (XOR) cata-
lyzes the terminal two steps of purine degradation (hypoxanthine →
xanthine → uric acid) in humans. XOR is transcribed as a single gene
product, xanthine dehydrogenase (XDH). Substrate-derived electrons
at the Mo cofactor of XDH are transferred via Fe/S centers to a FAD
moiety where NAD+ is reduced to NADH. Under inflammatory
conditions, posttranslational modification by oxidation of critical
cysteine residues or limited proteolysis converts XDH to xanthine
oxidase (XO) [1,2]. The key difference between XDH and XO is the
structural conformation and electrostatic microenvironment sur-
rounding the FAD resulting in a decreased affinity for NAD+ and
enhancement of affinity for O2 [3]. Substrate-derived electrons at the
Mo cofactor of XO reduce O2 at the FAD cofactor both univalently,
generating superoxide (O2•−), and divalently, forming hydrogen
peroxide (H2O2). However, conversion to XO is not requisite for ROS
production, as XDH displays partial oxidase activity under conditions
in which NAD+ levels are diminished such as the ischemic/hypoxic
microenvironment encountered in vascular inflammation [4]. This
same inflammatory milieu leads to enhanced XO levels and thus
increased XO-derived ROS formation resulting in activation of redox-
dependent cell signaling reactions and alterations in vascular function.
Evidence of this role for XO is exemplified by numerous studies in
which XO inhibition attenuates symptoms of vascular disease
including congestive heart failure, sickle cell anemia, and diabetes
[5–8].
Reports of XO-derived ROS production frequently address XO as the
O2•−-producing form of XDH and H2O2 is produced as a secondary by-
product of spontaneous or enzymatic dismutation of O2•−. A crucial
concept is often overlooked, specifically, that under relatively physio-
logic conditions (21% O2 and pH 7.0) XO catalyzes the reduction of O2 to
•−
H2O2 and O2•− at a ratio of 4:1 (H2O2:O2•−) or ∼80% H2O2 and ∼20% O2
,
whereas production of 100% O2•− requires an environment of 100% O2 at
pH 10 [9]. Although some studies have acknowledged this characteristic
of XO [10–15], it is vastly underappreciated in the literature, in which
focus remains fixed on O2•− as the key reactive product derived from XO.
In addition, a limited number of biochemical studies addressing XO-
mediated H2O2 production have centered on hyperoxia and/or alkaline
conditions, which are less reflective of pathophysiologic conditions
under which XOR most likely exerts significant influences [9,16–18].
With renewed attention being focused on XO-derived ROS in numerous
inflammatory processes, the relationship between O2 concentration and
•−
XO-catalyzed H2O2/O2 formation is crucial for the evaluation of
Abbreviations: BAEC, bovine aortic endothelial cells; CAT, catalase; GAG, glycos-
aminoglycan; ROS, reactive oxygen species; SOD, superoxide dismutase; XDH, xanthine
dehydrogenase; XO, xanthine oxidase; XOR, xanthine oxidoreductase.
contributory roles of XO and subsequent design of pharmacological
approaches for treatment. In toto, these issues affirm the need for the
examination of XO-derived ROS under clinically relevant conditions as
performed herein.
⁎
Corresponding authors. Fax: 412 648 9587.
E-mail addresses: ekelley@pitt.edu (E.E. Kelley), mtarpey+@pitt.edu (M.M. Tarpey)
0891-5849/$ – see front matter. Published by Elsevier Inc.
doi:10.1016/j.freeradbiomed.2009.11.012

494
E.E. Kelley et al. / Free Radical Biology & Medicine 48 (2010) 493–498
Materials and methods
molecular O2 were determined polarographically using an Apollo
4000 free radical analyzer (World Precision Instruments, Sarasota, FL,
USA). Experiments were performed at standard temperature (25°C)
and pressure (1 atm).
Materials
Xanthine, allopurinol, diphenyleneiodonium chloride, Chelex
resin, catalase, and uric acid were from Sigma (St. Louis, MO, USA).
Medium 199 and fetal bovine serum (FBS) were from Invitrogen
(Carlsbad, CA, USA). Superoxide dismutase (CuZnSOD) was from Oxis
International (Portland, OR, USA).
Cellular studies
Bovine aortic endothelial cells (BAEC) were isolated as previously
described [22]. Primary cell culture, routine passage, and experimen-
tal manipulations were all conducted in the absence of proteases. Cells
were propagated by subculturing (1:4 ratios) in Medium 199
containing 10% FBS and thymidine (10 μM). Cells were utilized
between passages 4 and 8 and were monitored visually for typical
cobblestone morphology indicative of endothelial cells and by
staining for von Willebrand factor expression. For O2 consumption
studies, confluent BAEC were exposed to XO (5 mU/ml) for 20 min at
25°C, harvested by mechanical dissociation, washed thoroughly
(three times with ice-cold PBS, pH 7.4), resuspended as a single-cell
suspension (2×106 cells/ml), and placed on ice (for less than 1 h)
until warmed to 25°C immediately before evaluation. This method
minimizes cellular internalization of the enzyme as we have
previously demonstrated [23]. Cell viability was 93% after 1 h on ice
as determined by trypan blue dye exclusion. Oxygen consumption
studies were performed under various O2 tensions, as above, in the
presence and absence of 50 U/ml catalase (CAT) and/or SOD.
Buffer treatment
Buffers for all experiments were prepared from Milli-Q H2O and
treated with Chelex resin to remove adventitious metals and thus
minimize loss of ROS by metal-catalyzed reactions.
XO activity
Enzyme was purified from fresh bovine cream by the method of
Rajagopalan and stored in ammonium sulfate at 4°C until
immediately before use [19]. Enzymatic activity was determined
either spectrophotometrically by the rate of uric acid formation
monitored at 292 nm in 50 mM potassium phosphate (KPi), pH 7.4
(ɛ=11 mM−1 cm−1), or electrochemically via reverse-phase HPLC
analysis of uric acid production (ESA CoulArray System, Chelmsford,
MA, USA) (1 unit=1 μmol urate/min) as previously described [20].
XDH activity was distinguished from XO activity by incubation with
Statistics
NAD+ as previously described [21]. Formation of O2 was assessed
•−
by the SOD-inhibitable reduction of cytochrome c (550 nm) [9].
Data were analyzed using one-way analysis of variance followed
by Tukey's range test for multiple pair-wise comparisons. Significance
was determined as pb0.05.
Univalent/divalent flux
Univalent flux was determined as previously reported [9].
Briefly, under saturating xanthine concentrations, the total electron
flux through the enzyme to O2 was calculated as the rate of uric acid
formation. Under saturating cytochrome c concentrations, the rate
of SOD-inhibitable cytochrome c reduction represents a measure of
electron flux via the univalent reduction of O2 to O2•−. Dividing the
cytochrome c reduction rate (1 e−) by the uric acid formation rate
(2 e−) gives the percentage of univalent flux (cytochrome c/ (2 x
uric acid)×100). As O2 is the sole oxidizing substrate in the system,
the divalent flux is derived by subtracting the percentage of
univalent flux from 100.
Results
Univalent/divalent flux
At 21% O2, oxidation of xanthine (100 μM) to uric acid by XO
(5 mU/ml) resulted in 28.1 1.4% univalent flux (O•2− formation) in
accordance with our previous reports (Table 1) [20,24]. Between 21
and 5% O2, univalent flux remained relatively constant at ∼70%. As
O2 concentration was lowered below 5%, there was an O2-dependent
•−
decrease in univalent flux so that O2 formation accounted for only
10% of electron flow through the enzyme at 1% O2. Likewise, XO-
dependent divalent reduction of O2 to H2O2 increased from 72%
(21% O2) to 90% (1% O2). As O2 tensions dropped below the Km for
O2 (46 μM) an O2-dependent decrease in the total electron flux (as
determined by the formation of uric acid) through the enzyme was
observed (1%, 13 μmol/min; 2.5%, 18 μmol/min; and 5%, 2.3 μmol/
min) (1, 2.5, and 5% O2 =∼13, 29, and 59 μM O2, respectively) [25].
This was reflected by a decrease in the overall rates of both uric acid
XO binding to GAGs
Xanthine oxidase was bound to heparin–Sepharose 6B (HS6B) as
previously described [20]. Briefly, XO (2 mg/ml) was added to a fixed
amount of gel (0.05 g dry wt) and the mixture gently stirred in 5 mM
KPi, pH 7.4 (2 ml final volume), at 25°C for 30 min. The suspension was
centrifuged at 10,000 g for 5 min, washed and the pellet was
resuspended in 5 mM KPi, pH 7.4. A quantity of HS6B–XO, equaling
5 mU/ml of XO activity, was added to PBS, pH 7.4, in a 3-ml cuvette
containing a small stir bar. Continuous gentle stirring was maintained
with a Helma electronic stirrer placed inside the spectrophotometer
cavity.
formation and cytochrome
c reduction. Control experiments
Table 1
Oxygen dependence of univalent vs divalent flux for XO
−
[O2] (%)
% univalent flux (O2
)
% divalent flux (H2O2)
1
2.5
5
10
15
21
10.4 0.6
14.3 0.9
24.6 2.3
28.5 0.7
27.8 0.5
28.1 1.4
90
86
76
71
72
72
Oxygen tension experiments
Experiments at specific oxygen tensions were performed in a
table-top glove box (Coy Instruments, Grass Lake, MI, USA) purged
with N2. All buffers were equilibrated N18 h before use. Glove box
atmospheric O2 conditions were followed with an O2 monitor
(Maxtec, Salt Lake City, UT, USA) and O2 concentrations verified
with a clinical blood gas analyzer. Spectrophotometric determinations
were carried out in gas-tight cuvettes. Real-time concentrations of
Purified XO (5 mU/ml) was exposed to xanthine (100 μM). Values for univalent flux
were calculated by monitoring the SOD-inhibitable reduction of cytochrome
(λ=550 nm) as described under Materials and methods. Divalent flux was derived
by subtracting the percentage univalent flux values from 100. Experiments were
performed at standard temperature (25 °C) and pressure (1 atm).
c

E.E. Kelley et al. / Free Radical Biology & Medicine 48 (2010) 493–498
495
when SOD or CAT was added to the sample before xanthine, CAT
significantly decreased the rate of O2 consumption, whereas SOD did
not (Fig. 2B). However, the addition of both SOD and CAT resulted in
an additional decrease in the O2 consumption rate compared to that
produced by CAT alone, as diagrammed in Fig. 2C. Plotting the O2
consumption rates in the presence of SOD + CAT as a percentage of
the CAT-only rates for each O2 tension revealed an O2-dependent
decrease in the SOD + CAT rates from 1 to 5% O2 (Fig. 2D). For all O2
consumption studies, addition of boiled CAT or SOD did not affect O2
evolution or rates of O2 consumption (not shown).
Effects of pH
The effects of pH (5.5–7.4) on the relative proportions of O2•− and
H2O2 formation by XO were determined under various O2 concentra-
tions (1–21%) at saturating xanthine concentrations (100 μM). There
was no effect of pH on univalent/divalent flux at any of the O2
concentrations examined (data not shown).
XO immobilization
At 21% O2, immobilization of XO by binding to heparin–Sepharose
6B reduced the rates of urate formation and cytochrome c reduction
as we reported previously [20] (Table 2). In addition to an
immobilization-induced reduction of total electron flow through the
enzyme, the relative proportions of O2•− and H2O2 were also altered,
reducing univalent electron transfer to O2 by 30% compared to XO in
solution. This immobilization-induced reduction in univalent flux was
consistent for all O2 tensions examined, such that at 1% O2, univalent
flux was only 6.9% for bound XO compared to 10.4% univalent flux for
free XO.
Fig. 1. H2O2 is the major ROS produced by XO at low O2 tensions. (A) XO (20 mU/ml,
PBS, pH 7.4) was exposed to xanthine (100 μM) in a 1% O2 atmosphere and O2
concentration was monitored polarographically over time. (B) During the linear phase
of O2 consumption 50 U/ml catalase (CAT) (2H2O2 → 2H2O + O2) was added to the
reaction. (C) Scheme showing total O2 evolved (2.1 μM) by the addition of 50 U/ml CAT
at 1% O2 and the effect of increasing O2 tension (1–5%) on the quantity of CAT-
dependent O2 evolution. (D) Using the quantity of O2 evolved at 1% O2 as maximum
(100%), CAT-dependent O2 evolution values are plotted as % of maximum for O2
concentrations (1–21%). For each O2 tension, CAT was added to the sample immediately
after the same XO-dependent depletion of O2 (4.7 μM). Values represent the means of at
least four independent determinations.
conducted in the presence of CAT demonstrated no H2O2-mediated
reoxidation of cytochrome c during the initial 60 s of the reaction
(the time frame for all univalent flux studies herein) that could
•−
account for the observed reduction in O2 formation. Additional
control experiments demonstrated that XO preparations were not
contaminated with SOD as determined by both Western blot and
activity assays. When O2 tension was elevated to 95%, the
percentage univalent flux increased to 47.2 2.3 (data not shown).
Oxygen consumption
Initiation of XO turnover by the addition of xanthine (100 μM), at 1%
O2, resulted in a rapid rate of O2 consumption (Fig. 1A). Increasing the O2
concentration from 1 to 5% resulted in an O2-dependent increase in the
rate of O2 consumption similar to the effects observed for univalent flux
studies. At 1% O2, the addition of CAT (50 U/ml) after 4.7 μM O2 was
consumed produced an immediate and pronounced evolution of O2
(2.1 μM) from Reaction (1) (Fig. 1B). Similar experiments carried out at
different O2 concentrations revealed an O2-dependent diminution of
CAT-induced O2 evolution between1 and 5% O2 withnofurther decrease
in O2 evolved above 5% O2 (Figs. 1C and 1D). It is important to note that
the quantity of H2O2 consumed by CAT is represented by 2×O2 evolved
from Reaction (1). For example, at 1% O2, CAT addition resulted in 2.1 μM
O2 evolution, representing the enzymatic catalysis of 4.2 μM H2O2, and
thus 89% of the O2 consumed was due to H2O2 formation. Similar
calculations for percentage H2O2 formation from O2 consumption at the
various O2 tensions yielded 89 (1% O2), 77 (2.5% O2), 72 (5% O2), 64 (10%
O2), 68 (15% O2), and 65% (21% O2).
Fig. 2. Oxygen dependence of O2•− formation from XO. (A) XO (20 mU/ml, PBS, pH 7.4)
was exposed to xanthine (100 μM) at 1% O2 as for Fig. 1. During the linear phase of O2
consumption 50 U/ml SOD (O2•− + O2•− → H2O2 + O2) was added to the reaction. (B)
Either SOD or CAT was added to the sample before the initiation of enzyme turnover
with xanthine (100 μM) and compared to control in the absence of antioxidant. (C)
Shown is a representative diagram of experiments conducted similar to those in (B),
but with SOD and CAT added together before xanthine. (D) The SOD + CAT-dependent
decreases in the rate of O2 consumption are plotted as percentage of the CAT-only
values (CAT only=100%) at each O2 concentration (1–21%). These SOD + CAT-
dependent decreases in the rate of O2 depletion result from evolution of O2 from the
reactions of SOD (O2•− + O2•− → H2O2 + O2) and CAT (2H2O2 → 2H2O + O2). Shown
are representative experiments of at least four independent determinations.
2H2O2Y2H2O + O2
ð1Þ
ð2Þ
•−
O2 + O2•−YH2O2 + O2
Addition of SOD (50 mU/ml) during enzyme turnover did not alter
the rate of O2 consumption from Reaction (2) (Fig. 2A). At 1% O2,

496
E.E. Kelley et al. / Free Radical Biology & Medicine 48 (2010) 493–498
Table 2
Discussion
Oxygen dependence of univalent vs divalent flux for GAG-immobilized XO
•−
Studies on reactive species derived from XO have confirmed the
[O2] %
% O2
Free
% H2O2
Free
•−
biological production of O2
[26]. However, under normoxic
Bound
Bound
conditions at neutral pH, XO produces significantly greater quantities
of H2O2 [9]. Perhaps of more significance is that under the lower O2
tensions and pH encountered during inflammation or ischemia, H2O2
accounts for 90–95% of XO-derived ROS formation (Tables 1 and 2).
These data confirm an O2 dependence of ROS formation in that lower
O2 concentrations lead to even greater H2O2 formation by XO.
Conceptually, this can be illustrated by consideration of the overall
oxidation/reduction state of the enzyme (Fig. 4). This scheme, albeit a
very simplified representation of a complex series of electronic
interactions between cofactors, summarizes the concept that a more
reduced XO favors H2O2 production and a more oxidized XO favors
1
5
10
21
10.4 0.6
24.6 2.3
28.5 0.7
28.1 1.4
6.9 1.2
90
76
71
72
93
83
81
80
17.4 4.2
18.9 2.1
19.8 1.1
XO was immobilized on heparin–Sepharose 6B and exposed to xanthine (100 μM) and
univalent flux (O2 formation) was determined by the SOD-inhibitable reduction of
cytochrome c (λ=550 nm) under the indicated O2 tensions as described under
Materials and methods. Divalent flux (H2O2 formation) was derived by subtracting the
percentage univalent flux values from 100. Experiments were performed at standard
temperature (25°C) and pressure (1 atm).
•−
•−
•−
Endothelial cell-bound XO
O2 formation. This effect of the redox state of XO on H2O2 vs O2
production has been previously reported with purified enzyme at pH
8.5 and confirms that high concentrations of xanthine and low O2
tensions result in greater H2O2 production, whereas low concentra-
tions of xanthine and high O2 tensions result in greater O2•− formation
[18,27–29]. This is critical to note as hypoxia/inflammation leads to
lower O2 tensions, increased XO expression, and increased hypoxan-
thine/xanthine levels from ATP catabolism and thus sets the stage for
enhanced vascular H2O2 production. Therefore, recognizing that both
reducing substrate concentration and O2 tension are key factors in
determining the identity of XO-mediated ROS formation, we
examined the effects of O2 tension in the presence of a saturating
xanthine concentration (100 μM). We chose this xanthine level as it
reflects the concentrations encountered in vivo under ischemic/
hypoxic conditions in which ATP breakdown results in elevation of
hypoxanthine/xanthine levels from ∼2 to 50–100 μM, well above its
Km (6.5 μM) [20]. In aggregate, our studies were designed to confirm
and expand the observations of previous studies under clinically
relevant conditions and to refocus attention on XO-derived H2O2.
Electron flux studies utilize uric acid formation and cytochrome c
reduction to determine univalent flux; however, they do not directly
measure divalent flux and thus it is a derived value. In a simple system
in which O2 is the sole electron acceptor, this is a valid calculation;
Xanthine oxidase was bound to extracellular GAGs on the apical
surface of BAEC as described under Materials and methods. Cellular
O2 consumption studies were performed at various O2 tensions (1–
21%) in which CAT (200 U/ml), SOD (200 U/ml), or both were
present before the addition of xanthine. The effect of the presence
of the antioxidants on the rate of XO-dependent O2 consumption
was determined and plotted as percentage decrease compared to
control rates (no SOD or CAT) at each O2 tension (Fig. 3). The
presence of CAT significantly decreased the rate of O2 consumption
at all O2 tensions, resulting in a 27% decrease at 1% O2, whereas
SOD produced small, but significant, decreases (N5%) at O2 levels
above 2.5%. The presence of both SOD and CAT did not significantly
alter the rates of O2 consumption compared to samples containing
only CAT. Rates of O2 consumption in the absence of xanthine
were determined and subtracted from all values reported. When
XO-treated cells were washed and then exposed to trypsin (0.25%)
for 3 min at 37 °C to remove/inactivate GAG-associated enzyme,
no O2 consumption was observed upon addition of xanthine,
demonstrating that ROS formation was from cell-associated XO.
Furthermore, treatment with the XO-specific inhibitor Febuxostat
(25 μM) before addition of xanthine abolished XO-dependent O2
consumption (not shown).
Fig. 4. Effects of O2 on XO-derived ROS formation. Purine-derived electrons enter XO via
the Mo cofactor and egress through the FAD cofactor by reaction with molecular O2.
Therefore, at physiological pH, the oxidation/reduction status of XO is dependent upon
the concentrations of both purine (xanthine/hypoxanthine) and O2. Under saturating
purine concentrations, O2 tension is the prime determinant of this redox state of XO.
Shown are two cartoons depicting the extreme redox states (either fully oxidized or
reduced) of the various cofactors of XO in the presence of saturating xanthine
concentrations. On the left, under anoxia, the enzyme is fully reduced. Upon
oxygenation, the reduced FAD cofactor reacts with O2 divalently to produce H2O2. As
O2 levels increase so does O2-mediated electron egress via the FAD cofactor until it
begins to out-compete electron entry at the Mo cofactor, resulting in an increase in the
overall oxidation of the enzyme. This increase in the oxidation of the enzyme shifts FAD
Fig. 3. H2O2 is the major ROS produced by endothelial cell-associated XO. BAEC were
exposed to XO (5 mU/ml) as described under Materials and methods. Cell
suspensions (2×106 cells/ml) were monitored for O2 consumption upon the
addition of xanthine. Either SOD (200 U/ml) or CAT (200 U/ml) or both were
added to the samples before xanthine as indicated. The effects of the presence of
CAT, SOD, or CAT + SOD are represented as % decrease in the rate of O2 consumption
compared to control samples without antioxidant and are plotted for the indicated
O2 tensions. Values represent the means of three independent determinations.
⁎pb0.05 compared to CAT only and CAT + SOD.
•−
cofactor–O2 reactions from divalent to univalent, producing more O2 and less H2O2.

E.E. Kelley et al. / Free Radical Biology & Medicine 48 (2010) 493–498
497
however, an alternative approach that measures XO-derived H2O2
formation would confirm conclusions based on this derivation.
Oxygen consumption was chosen to examine XO-dependent H2O2
formation in the presence and absence of CAT and SOD, the enzymes
produced a similar ∼30% reduction in univalent flux, so that at 1%
O2, 93% of the ROS formed was H2O2. These data suggest that XO-
derived ROS production is almost exclusively H2O2 when XO is bound
to endothelial GAGs in a hypoxic vascular milieu. To test this
hypothesis in a cellular model, BAEC were exposed to purified XO,
and XO-dependent O2 consumption was monitored over time in the
presence of SOD, CAT, or CAT + SOD (Fig. 3). The presence of SOD
slightly altered O2 consumption rates only when O2 tensions
exceeded 5%. In contrast, CAT significantly decreased the rates of O2
consumption as XO-produced H2O2 was converted to H2O and O2.
These CAT-dependent decrements in O2 consumption rates were seen
at all O2 tensions, with the greatest effects observed below 5% at
which H2O2 production would be expected to account for a
preponderance of total ROS formation. The presence of both CAT
and SOD produced O2 consumption rates that were not statistically
different from rates produced by samples containing only CAT.
Although differences between the CAT + SOD and the CAT-only
values did not demonstrate statistical significance, the individual
mean values for CAT + SOD from 5 to 21% O2 were all greater than the
corresponding CAT-only values, possibly suggesting contributions
from SOD near the limit of detection. Combined, these data
demonstrate that quantification and characterization of ROS from
these cellular studies is problematic owing to the loss of both H2O2
and O2•− before reaction with CAT and SOD from many mechanisms,
including reactions with cellular constituents such as GSH. However,
they do reveal trends similar to those observed in the biochemical
studies and qualitatively demonstrate that H2O2 is the major ROS
product of XO, especially at O2 tensions relevant to vascular
pathology.
•−
most often used to substantiate contributions of H2O2 and O2 in
biological systems (Fig. 1). This avoids possible direct interactions
with XOR, ROS specificity issues, and the potential of redox cycling,
which limit the specificity and sensitivity of several dye-based H2O2
assays [30]. At 1% O2, CAT addition resulted in significant evolution of
O2, indicating that 89% of the O2 consumed by XO was reduced to
H2O2. Values obtained at other O2 tensions were also similar to, yet
slightly lower than, the divalent fluxes listed in Table 1, supporting the
validity of this approach. Lower yields for the O2 evolution experi-
ments may be due to the extended time (∼2 min) of enzyme turnover
and thus potential loss of H2O2 before the addition of CAT and/or
subtle variations in temperature and barometric pressure.
Because CAT addition experiments confirmed the proportion of
H2O2 formed during XO-catalyzed O2 consumption, the remainder of
O2 consumption should be attributable to O2•− formation. However, at
1% O2, the presence of SOD before or after (Fig. 2) the addition of
xanthine resulted in neither O2 evolution nor alteration of the O2
•−
consumption rate. It is hypothesized that loss of O2
from
spontaneous dismutation (∼2×105 M−1 s−1), reaction with protein
in the sample, and degree of assay sensitivity all contributed to the
absence of observable effects. To further address this issue, experi-
ments were performed in which both SOD and CAT were present
before initiation of enzyme turnover. At all O2 tensions, addition of
SOD + CAT before xanthine decreased the rate of O2 consumption
compared to samples containing only CAT (Figs. 2C and 2D). It is
assumed that both O2 evolution and the formation of additional CAT
substrate (H2O2) by SOD (Reaction (2)) served to augment total O2
evolution. The CAT + SOD-induced decreases in O2 consumption rates
were not O2-dependent above 5% O2 (Fig. 2D), which is consistent
with data from univalent flux studies showing no significant increase
in O2•− formation from 5 to 21% O2. This result is also consistent with
the argument that as O2 levels exceed the Km of O2 with FAD (46 μM),
a constant rate of O2•− formation would ensue when the presence of
SOD offers no additional contribution to O2 evolution and subsequent
reduction in the rate of O2 consumption in this O2 concentration range
(46–235 μM or 5–21%).
Over the past several years, there has been renewed interest in
the role of XO in vascular inflammation as the number of studies
demonstrating salutary effects of XO inhibition accumulate. To
successfully evaluate the contribution of XO in these and future
studies, it is crucial to understand the identity and relative
proportions of the ROS produced by this complex enzyme under
conditions reflective of the microenvironment in which they are
formed. Herein, we revisit the underappreciated detail that H2O2 is
the major (∼75%) ROS produced by XO under normal aerobic
conditions and expand this observation to demonstrate that under
pathophysiologic conditions XO-derived H2O2 formation
approaches 95%. These results affirm the need to more critically
evaluate the role of H2O2 in studies in which XO-derived ROS have
been proposed to play a contributory role. For example, many of
Inflammation is often characterized by both hypoxia and lowered
pH; thus, it is important to consider the effects of H+ ion
concentration on ROS production by XO. The oxidation of xanthine
at the Mo cofactor is base-catalyzed and thus sensitive to changes in
pH in the physiological range as well as substrate concentration [27].
However, under saturating xanthine concentrations, the relative
•−
the vessel studies in which XO-derived O2 reduces NO-mediated
vasodilatation were conducted using 95% O2. This O2 tension is
significantly higher than O2 levels experienced in vivo and also
•−
•−
proportion of H2O2 to O2 at each O2 tension examined was not
serves to greatly enhance XO-derived O2 formation and possibly
altered by pH (5.5–7.4).
lead to misconceptions regarding the role of XO in this model.
Furthermore, assigning a causative role to XO-derived O2 based
•−
During inflammatory conditions XDH is released into the circula-
tion, rapidly converted to XO, and avidly binds (Kd =6 nM) to
negatively charged GAGs on the surface of vascular endothelial cells
[24,31–33]. This sequestration of XO by GAGs substantially amplifies
local enzyme concentration and diminishes its rotational and
translational mobility, thus altering kinetic properties and conferring
resistance to both product-induced and pharmacological inhibition
[20,24]. GAG association of XO decreases substrate binding affinity
and thus increases the Km for xanthine (6.5 μM free vs 21.2 μM bound)
and Ki for oxypurinol (85 nM free vs 451 nM bound) compared to XO
in solution [24,34]. These characteristics indicate that XO bound to
endothelial cell GAGs could serve as a long-lived source of ROS in this
microenvironment. It is in this setting that XO-generated ROS can
critically impact vessel function, emphasizing the need to more clearly
define the effects of O2 tension on ROS production by GAG-bound XO.
Our previous studies demonstrate that immobilization of XO on
heparin–Sepharose reduces univalent flux by 30% in room air [20]. At
every O2 concentration examined, GAG immobilization of XO
upon the beneficial actions of allopurinol administration, in the
absence of controls for XO-derived H2O2, may result in neglecting
pivotal signaling events mediated by H2O2, a molecule with an
expanding number of targets in various cell signaling pathways
[35–37].
Combined, the results from this report refocus attention on
previous and commonly overlooked studies reporting H2O2 as the
major ROS product of XO under normoxia. Furthermore, these data
demonstrate that under pathophysiologic conditions XO-derived
H2O2 formation approaches 95% and thus emphasize the danger of
failing to appreciate this attribute when investigating contributory
roles for XO in pathology.
Acknowledgments
This study was supported by a University of Pittsburgh, Depart-
ment of Anesthesiology, Seed Grant (E.E.K.) and National Institutes of

498
E.E. Kelley et al. / Free Radical Biology & Medicine 48 (2010) 493–498
[17] Kellogg III, E. W.; Fridovich, I. Superoxide, hydrogen peroxide, and singlet oxygen
in lipid peroxidation by a xanthine oxidase system. J. Biol. Chem. 250:8812–8817;
1975.
[18] Porras, A. G.; Olson, J. S.; Palmer, G. The reaction of reduced xanthine oxidase with
oxygen: kinetics of peroxide and superoxide formation. J. Biol. Chem. 256:
9006–9103; 1981.
Health Grants HL8115 and HL64937 (BAF). The authors thank Joe S.
Moritz, Ph.D., and Keith Inskeep, Ph.D., Division of Animal and
Veterinary Science, West Virginia University (Morgantown, WV, USA)
for both fresh cream and chicken livers used in enzyme purification.
[19] Waud, W. R.; Brady, F. O.; Wiley, R. D.; Rajagopalan, K. V. A new purification
procedure for bovine milk xanthine oxidase: effect of proteolysis on the subunit
structure. Arch. Biochem. Biophys. 169:695–701; 1975.
References
[1] Amaya, Y.; Yamazaki, K.; Sato, M.; Noda, K.; Nishino, T. Proteolytic conversion of
xanthine dehydrogenase from the NAD-dependent type to the O2-dependent
type: amino acid sequence of rat liver xanthine dehydrogenase and identification
of the cleavage sites of the enzyme protein during irreversible conversion by
trypsin. J. Biol. Chem. 265:14170–14175; 1990.
[2] Waud, W. R.; Rajagopalan, K. V. The mechanism of conversion of rat liver xanthine
dehydrogenase from an NAD+-dependent form (type D) to an O2-dependent
form (type O). Arch. Biochem. Biophys. 172:365–379; 1976.
[3] Enroth, C.; Eger, B. T.; Okamoto, K.; Nishino, T.; Nishino, T.; Pai, E. F. Crystal
structures of bovine milk xanthine dehydrogenase and xanthine oxidase:
structure-based mechanism of conversion. Proc. Natl. Acad. Sci. USA 97:
10723–10728; 2000.
[4] Harris, C. M.; Massey, V. The oxidative half-reaction of xanthine dehydrogenase
with NAD: reaction kinetics and steady-state mechanism. J. Biol. Chem. 272:
28335–28341; 1997.
[20] Kelley, E. E.; Trostchansky, A.; Rubbo, H.; Freeman, B. A.; Radi, R.; Tarpey, M. M.
Binding of xanthine oxidase to glycosaminoglycans limits inhibition by oxypur-
inol. J. Biol. Chem. 279:37231–37234; 2004.
[21] Suzuki, H.; DeLano, F. A.; Parks, D. A.; Jamshidi, N.; Granger, D. N.; Ishii, H.;
Suematsu, M.; Zweifach, B. W.; Schmid-Schonbein, G. W. Xanthine oxidase
activity associated with arterial blood pressure in spontaneously hypertensive
rats. Proc. Natl. Acad. Sci. USA 95:4754–4759; 1998.
[22] Kelley, E. E.; Hock, T.; Khoo, N. K. H.; Richardson, G. R.; Johnson, K. K.; Powell, P. C.;
Giles, G. I.; Agarwal, A.; Lancaster Jr., J. R.; Tarpey, M. M. Moderate hypoxia induces
xanthine oxidoreductase activity in arterial endothelial cells. Free Radic. Biol. Med.
40:952–959; 2006.
[23] Kelley, E. E.; Batthyany, C. I.; Hundley, N. J.; Woodcock, S. R.; Bonacci, G.; Del
Rio, J. M.; Schopfer, F. J.; Lancaster Jr., J. R.; Freeman, B. A.; Tarpey, M. M. Nitro-
oleic acid, a novel and irreversible inhibitor of xanthine oxidoreductase. J. Biol.
Chem. 283:36176–36184; 2008.
[5] Butler, R.; Morris, A. D.; Belch, J. J.; Hill, A.; Struthers, A. D. Allopurinol normalizes
endothelial dysfunction in type 2 diabetics with mild hypertension. Hypertension
35:746–751; 2000.
[6] Desco, M. C.; Asensi, M.; Marquez, R.; Martinez-Valls, J.; Vento, M.; Pallardo, F. V.;
Sastre, J.; Vina, J. Xanthine oxidase is involved in free radical production in type 1
diabetes: protection by allopurinol. Diabetes 51:1118–1124; 2002.
[7] Farquharson, C. A.; Butler, R.; Hill, A.; Belch, J. J.; Struthers, A. D. Allopurinol
improves endothelial dysfunction in chronic heart failure. Circulation 106:
221–226; 2002.
[24] Radi, R.; Rubbo, H.; Bush, K.; Freeman, B. A. Xanthine oxidase binding to
glycosaminoglycans: kinetics and superoxide dismutase interactions of immobi-
lized xanthine oxidase–heparin complexes. Arch. Biochem. Biophys. 339:125–135;
1997.
[25] Saito, T.; Nishino, T. Differences in redox and kinetic properties between NAD-
dependent and O2-dependent types of rat liver xanthine dehydrogenase. J. Biol.
Chem. 264:10015–10022; 1989.
[26] McCord, J. M.; Fridovich, I. The reduction of cytochrome c by milk xanthine
oxidase. J. Biol. Chem. 243:5753–5760; 1968.
[8] Aslan, M.; Ryan, T. M.; Adler, B.; Townes, T. M.; Parks, D. A.; Thompson, J. A.;
Tousson, A.; Gladwin, M. T.; Patel, R. P.; Tarpey, M. M.; Batinic-Haberle, I.; White,
C. R.; Freeman, B. A. Oxygen radical inhibition of nitric oxide-dependent vascular
function in sickle cell disease. Proc. Natl. Acad. Sci. USA 98:15215–15220; 2001.
[9] Fridovich, I. Quantitative aspects of the production of superoxide anion radical by
milk xanthine oxidase. J. Biol. Chem. 245:4053–4057; 1970.
[27] Hille, R.; Massey, V. Studies on the oxidative half-reaction of xanthine oxidase.
J. Biol. Chem. 256:9090–9095; 1981.
[28] Olson, J. S.; Ballow, D. P.; Palmer, G.; Massey, V. The reaction of xanthine oxidase
with molecular oxygen. J. Biol. Chem. 249:4350–4362; 1974.
[29] Olson, J. S.; Ballou, D. P.; Palmer, G.; Massey, V. The mechanism of action of
xanthine oxidase. J. Biol. Chem. 249:4363–4382; 1974.
[10] McCutchan, H. J.; Schwappach, J. R.; Enquist, E. G.; Walden, D. L.; Terada, L. S.;
Reiss, O. K.; Leff, J. A.; Repine, J. E. Xanthine oxidase-derived H2O2 contributes to
reperfusion injury of ischemic skeletal muscle. Am. J. Physiol. 258:H1415–H1419;
1990.
[11] Brown, J. M.; Terada, L. S.; Grosso, M. A.; Whitmann, G. J.; Velasco, S. E.; Patt, A.;
Harken, A. H.; Repine, J. E. Xanthine oxidase produces hydrogen peroxide which
contributes to reperfusion injury of ischemic, isolated, perfused rat hearts. J. Clin.
Invest. 81:1297–1301; 1988.
[30] Tarpey, M. M.; Fridovich, I. Methods of detection of vascular reactive species:
nitric oxide, superoxide, hydrogen peroxide, and peroxynitrite. Circ. Res. 89:
224–236; 2001.
[31] Adachi, T.; Fukushima, T.; Usami, Y.; Hirano, K. Binding of human xanthine oxidase
to sulphated glycosaminoglycans on the endothelial-cell surface. Biochem. J. 289:
523–527; 1993.
[32] Fukushima, T.; Adachi, T.; Hirano, K. The heparin-binding site of human xanthine
oxidase. Biol. Pharm. Bull. 18:156–158; 1995.
[12] Siems, W.; Schmidt, H.; Muller, M.; Henke, W. Gerber, G. H2O2 formation during
nucleotide degradation in the hypoxic rat liver: a quantitative approach. Free
Radic. Res. Commun. 1:289–295; 1986.
[13] Patt, A.; Harken, A. H.; Burton, L. K.; Rodell, T. C.; Piermattei, D.; Schorr, W. J.;
Parker, N. B.; Berger, E. M.; Horesh, I. R.; Terada, L. S. Xanthine oxidase-derived
hydrogen peroxide contributes to ischemia reperfusion-induced edema in gerbil
brains. J. Clin. Invest. 81:1556–1562; 1988.
[33] Houston, M.; Estevez, A.; Chumley, P.; Aslan, M.; Marklund, S.; Parks, D. A.;
Freeman, B. A. Binding of xanthine oxidase to vascular endothelium: kinetic
characterization and oxidative impairment of nitric oxide-dependent signaling.
J. Biol. Chem. 274:4985–4994; 1999.
[34] Camejo, G.; Olsson, U.; Hurt-Camejo, E.; Baharamian, N.; Bondjers, G. The
extracellular matrix on atherogenesis and diabetes-associated vascular disease.
Atherosclerosis (Suppl. 3):3–9; 2002.
[14] Fatokun, A. A.; Stone, T. W.; Smith, R. A. Hydrogen peroxide mediates damage by
xanthine and xanthine oxidase in cerebellar granule neuronal cultures. Neurosci.
Lett. 416:34–38; 2007.
[35] Rhee, S. G.; Kang, S. W.; Jeong, W.; Chang, T. S.; Yang, K. S.; Woo, H. A. Intracellular
messenger function of hydrogen peroxide and its regulation by peroxiredoxins.
Curr. Opin. Cell Biol. 17:183–189; 2005.
[15] Lee, H.; Carlson, J. D.; McMahon, K. K.; Moyer, T. P.; Fischer, A. G. Xanthine oxidase:
a source of hydrogen peroxide in bovine thyroid glands. Life Sci. 20:453–458; 1977.
[16] Lacy, F.; Gough, D. A.; Schmid-Schonbein, G. W. Role of xanthine oxidase in
hydrogen peroxide production. Free Radic. Biol. Med. 25:720–727; 1998.
[36] Forman, H. J.; Torres, M. Reactive oxygen species and cell signaling: respiratory
burst in macrophage signaling. Am. J. Respir. Crit. Care Med. 166:4S–8S; 2002.
[37] Cai, H. Hydrogen peroxide regulation of endothelial function: origins, mechan-
isms, and consequences. Cardiovasc. Res. 68:26–36; 2005.