In vitro oxidative metabolism of 6-mercaptopurine in human liver: Insights into the role of the molybdoflavoenzymes aldehyde oxidase, xanthine oxidase, and xanthine dehydrogenase

Article abstract of DOI:10.1124/dmd.114.058107

Anticancer agent 6-mercaptopurine (6MP) has been in use since 1953 for the treatment of childhood acute lymphoblastic leukemia (ALL) and inflammatory bowel disease. Despite being available for 60 years, several aspects of 6MP drug metabolism and pharmacok

Full text of DOI:10.1124/dmd.114.058107

Supplemental material to this article can be found at:
http://dmd.aspetjournals.org/content/suppl/2014/05/13/dmd.114.058107.DC1.html
1
D
521-009X/42/8/1334–1340$25.00
RUG METABOLISM AND DISPOSITION
http://dx.doi.org/10.1124/dmd.114.058107
Drug Metab Dispos 42:1334–1340, August 2014
Copyright ª 2014 by The American Society for Pharmacology and Experimental Therapeutics
In Vitro Oxidative Metabolism of 6-Mercaptopurine in Human Liver:
Insights into the Role of the Molybdoflavoenzymes Aldehyde
s
Oxidase, Xanthine Oxidase, and Xanthine Dehydrogenase
Kanika V. Choughule, Carlo Barnaba, Carolyn A. Joswig-Jones, and Jeffrey P. Jones
Department of Chemistry, Washington State University, Pullman, Washington
Received March 13, 2014; accepted May 13, 2014
ABSTRACT
Anticancer agent 6-mercaptopurine (6MP) has been in use since 1953 Both AO and XO were involved in the metabolism of the 6TX
for the treatment of childhood acute lymphoblastic leukemia (ALL) and intermediate, whereas only XO was responsible for the conversion of
inflammatory bowel disease. Despite being available for 60 years, 6TX to 6TUA. These findings were further confirmed using purified
several aspects of 6MP drug metabolism and pharmacokinetics in
humans are unknown. Molybdoflavoenzymes such as aldehyde
human AO and Escherichia coli lysate containing expressed recombi-
nant human XO. Xanthine dehydrogenase (XDH), which belongs to the
oxidase (AO) and xanthine oxidase (XO) have previously been family of xanthine oxidoreductases and preferentially reduces nicotin-
+
implicated in the metabolism of this drug. In this study, we investigated amide adenine dinucleotide (NAD ), was shown to contribute to the
the in vitro metabolism of 6MP to 6-thiouric acid (6TUA) in pooled
human liver cytosol. We discovered that 6MP is metabolized to 6TUA 6TUA in the presence of NAD in human liver cytosol. In conclusion, we
through sequential metabolism via the 6-thioxanthine (6TX) inter- present evidence that three enzymes, AO, XO, and XDH, contribute to
mediate. The role of human AO and XO in the metabolism of 6MP was the production of 6TX intermediate, whereas only XO and XDH are
overall production of the 6TX intermediate as well as the final product
+
established using the specific inhibitors raloxifene and febuxostat.
involved in the conversion of 6TX to 6TUA in pooled HLC.
Introduction
in vivo (Keuzenkamp-Jansen et al., 1996; Rowland et al., 1999). There
is contradictory evidence on whether 6MP is converted to 6TUA via
6
-Mercaptopurine (6MP) is a thiopurine drug with antitumor activity
6TX or 8-oxo-6MP in vivo. Early pharmacokinetic studies revealed that
that has been in use as a remission-inducing agent for the treatment of
childhood acute lymphoblastic leukemia (Burchenal et al., 1953). It has
also been used as an immunosuppressive agent in combination with its
prodrug, azathioprine, for the treatment of inflammatory bowel disease
such as ulcerative colitis and Crohn’s disease (Nielsen et al., 2001).
this drug was initially oxidized to 8-oxo-6-mercaptopurine before being
converted to 6-thiouric acid (Bergmann and Ungar, 1960; Elion, 1967;
Van Scoik et al., 1985). However, Zimm et al. (1984) identified
6-thioxanthine in urine of patients dosed with 6MP and proposed that this
metabolite could also be an intermediate in the formation of 6-thiouric
acid. Human xanthine oxidase (XO) and aldehyde oxidase (AO) are
very closely related molybdoflavoenzymes that have a high degree of
amino acid sequence identity, require the same cofactors (Garattini
6
MP is structurally related to endogenous purine bases such as adenine,
guanine, and hypoxanthine, and hence is metabolized by enzyme
systems and pathways that metabolize endogenous purines (Aarbakke
et al., 1997). Phosphoribosylation, oxidation, and methylation are the
major metabolic pathways of 6MP metabolism (Fig. 1). Phosphor- et al., 2003), and share a similar mechanism of action (Alfaro and Jones,
ibosylation is an anabolic pathway that results in the production of 2008). However, they still differ remarkably in their substrate
active metabolites that exert the antitumor effect of 6MP by interfering specificities (Garattini and Terao, 2012). Conversion of 6MP to
with purine ribonucleotide synthesis. As opposed to phosphoribosyla- 6TUA has been attributed to the activity of these molybdoflavoen-
tion, oxidation and methylation are catabolic pathways that produce zymes. 6MP has a low oral bioavailability because of extensive first
inactive metabolites. It has been known that 6MP is converted to pass metabolism by hepatic and intestinal enzymes. It is believed that
methylmercaptopurine (MeMP) by the action of thiopurine methyl- the drug is rapidly oxidized to its major in vivo metabolite, 6-thiouric
transferase by a pathway that is almost exclusive for thiopurines acid, by the action of XO in the liver and intestine. Administration of
(Giverhaug et al., 1999). Oxidative metabolism of 6MP results in 6- 6MP along with XO inhibitors have resulted in an increase in the
thiouric acid (6TUA), 6-thioxanthine (6TX), 8-oxo-6-mercaptopurine bioavailability of this drug (Balis et al., 1987; Giverhaug et al., 1999).
(8-oxo-6MP), and 6-methylmercapto-8-hydroxypurine (6Me-8OH-MP) Apart from this, 6MP is also converted to 6TUA by calf liver XO,
bovine milk XO (Krenitsky et al., 1972), and rabbit liver AO (Hall
and Krenitsky, 1986). However, the contribution of AO/XO in the
This work was supported by a grant from the National Institutes of Health
conversion of 6MP to its intermediate and subsequently to 6TUA in
humans is largely unknown. Evidence by Rashidi et al. (2007) suggests
that 6MP is sequentially metabolized to produce 6TUA through the
[
GM100874 (to J.P.J.)].
dx.doi.org/10.1124/dmd.114.058107.
s This article has supplemental material available at dmd.aspetjournals.org.
ABBREVIATIONS: 6MP, 6-mercaptopurine; 6TUA, 6-thiouric acid; 6TX, 6-thioxanthine; AO, aldehyde oxidase; HLC, human liver cytosol; HPLC,
high-performance liquid chromatography; LB, lysogeny broth; LC, liquid chromatography; MS/MS, tandem mass spectrometry; NAD, nicotinamide
adenine dinucleotide; XDH, xanthine dehydrogenase; XO, xanthine oxidase; XOR, xanthine oxidoreductase.
1334

Oxidative Metabolism of 6-Mercaptopurine
1335
Fig. 1. Reaction scheme showing metabolites of 6-Mercaptopurine formed by phosphoribosylation (anabolic), methylation and oxidation (catabolic) pathways.
intermediate metabolite 6TX in partially purified guinea pig liver.
Rashidi et al. also demonstrated that 6MP is metabolized to 6TX
Materials and Methods
6
-Mercaptopurine was purchased from Sigma-Aldrich (St. Louis, MO),
exclusively by XO and subsequently converted to 6TUA by a concerted whereas 6-thioxanthine and 6-thiouric acid were purchased from Toronto
action of XO and AO. One caveat to bear in mind is that all of these in Research Chemicals (Toronto, ON, Canada). Febuxostat was kindly donated by
vitro experiments were performed using AO/XO from nonhuman Dr. Eric Kelley from University of Pittsburgh (Pittsburgh, PA). Raloxifene was
purchased from Enzo Life Sciences (Farmingdale, NY). Sequence grade trypsin
sources. Major species differences have already been reported for
was acquired from Promega (Madison, WI). Pooled mixed-gender human liver
several compounds metabolized by AO and hence extrapolation of
samples (n = 8) devoid of allopurinol were obtained from St Jude’s Children’s
pharmacokinetic data from other mammalian species to humans is not
advisable (Choughule et al., 2013; Dalvie et al., 2013). Moreover,
mammalian xanthine oxidoreductase (XOR) exists in two interconvert-
ible forms, xanthine oxidase (XO) and xanthine dehydrogenase (XDH).
Hospital human liver bank (Barr et al., 2014). The samples were prepared and
stored as described in (Barr et al., 2014). Table 1 contains information on the
demographics of the liver samples used to prepare cytosol. Plasmid pTrc99A
containing gene encoding for human XO was kindly provided by Tomohiro
+
XDH preferentially reduces nicotinamide adenine dinucleotide (NAD )
Matsumura from Nippon Medical School (Tokyo, Japan).
Expression of Recombinant Human XOR in TP1000 Cells. Human XOR
+
and predominates in vivo, while XO cannot reduce NAD and prefers
molecular oxygen as an electron acceptor (Harrison, 2002). It has containing plasmid DNA (10 ng) was transformed into TP1000 cells (a gift
already been established that XO plays a role in the metabolism of from John Enemark’s laboratory, University of Arizona). Transformants were
6
MP. However, the contribution of XDH to the metabolism of this thio- plated on lysogeny broth (LB) agar plates containing 100 mg/ml ampicillin and
grown overnight at 37°C. A single colony was used to inoculate 100 ml of LB
purine drug still remains largely unknown.
To understand the in vitro oxidative metabolism of this well
established anticancer drug, we investigated the in vitro metabolism of
containing 100 mg/ml ampicillin. Cell stocks were prepared in 15% glycerol
and stored at –80°C until further use.
For expression of human XOR, glycerol stocks of TP1000 cells containing
plasmid pRTC99A encoding human XOR were plated onto LB agarose plates
containing 100 mg/ml ampicillin. A single colony was picked and grown overnight
at 30°C in 100 ml LB containing 100 mg/ml ampicillin and 50 mM sodium
molybdate. A 10-ml aliquot of the culture was used to inoculate 500 ml fresh LB
broth with 100 mg/ml ampicillin and 50 mM sodium molybdate. Cultures were
6
MP to 6TUA via the 6TX intermediate in pooled human liver
cytosol. The roles of XO and AO were explored using specific
inhibitors raloxifene and febuxostat. We confirmed the roles by using
purified human AO and Escherichia coli lysate containing expressed
recombinant human XO. Finally, we assessed the metabolism in the
+
presence and absence of NAD to determine the role of XDH to the grown at 37°C and 250 rpm for 2 hours until an absorbance of 1 at 600 nm was
overall metabolism of 6MP and 6TX.
reached. Isopropyl b-D-thiogalactoside (IPTG) at a final concentration of 1 mM was

1
336
Choughule et al.
1 mM 3,5 dibromo-4-hydroxybenzoic acid as internal standard. The samples
TABLE 1
were then centrifuged at 5000 rpm for 10 minutes in an Eppendorf centrifuge
5415D and the supernatant was collected for analysis.
Demographics of individual liver samples pooled to generate human liver cytosol
used in this study
HPLC-MS/MS Assays. The samples were analyzed using an 1100 series
HPLC system (Agilent Technologies, Santa Clara, CA) coupled to a API4000
tandem mass spectrometry system (Applied Biosystems/MDS Sciex, Foster City,
Liver
Sex
Age
Disease History
Hepatoma
Colorectal cancer, metastasis
Colorectal cancer, metastasis
Cholangio carcinoma
Colorectal cancer, metastasis
Colorectal cancer, metastasis
No information given
1
2
3
4
5
6
7
8
Female
Male
69
62
68
63
60
70
24
57
CA) on
Chromatography was performed on a Zorbax Eclipse XDB-C8 column (4.6 ꢀ
50 mm; 5 mm; Agilent Technologies, Santa Clara, CA). Mobile phase A
a turbospray interface (ESI) operating in negative ion mode.
Female
Female
Female
Female
Female
Male
1
consisted of 0.05% formic acid and 0.2% acetic acid in water and mobile phase B
contained 90% acetonitrile, 9.9% water, and 0.1% formic acid. The column was
first equilibrated with 99% mobile phase A for 3 minutes. Chromatographic
separation was achieved using a linear gradient to 25% mobile phase A over the
next 1.5 minutes. Mobile phase A was then held constant for the next 1.5 minutes
before a linear gradient back to 99% mobile phase A was achieved over a period
of 1 minute. Finally, the column was re-equilibrated to starting conditions over
the next 1 minute. The entire chromatographic assay was performed over 8
minutes per sample with a constant flow rate of 800 ml/min The retention times
for 6-thioxanthine, 6-thiouric acid, and the internal standard were 6.2, 4.5, and
Colorectal cancer, metastasis
used to induce cells, which were then allowed to continue growing at room
temperature for 24 hours at 150 rpm. Cells were harvested by centrifugation at
500g at 4°C for 30 minutes in a swinging bucket rotor (Allegra 6R centrifuge;
Beckman Coulter, Pasadena, CA). Cell paste was collected and resuspended in
equal volumes of 1 g of paste to 1 ml of buffer (100 mM potassium phosphate
3
6.8 minutes, respectively (see Supplemental Fig. 1). The optimized mass
spectrometer tune parameters for 6TX and 6TUA were as follows: collision
gas,10; curtain gas, 30; ion source gas 1, 50; ion source gas 2, 30; ion spray
voltage, 4500; desolvation temperature, 450; declustering potential, 70; entrance
potential, 15; collision energy, 30; collision cell exit potential, 2. The analytes,
buffer, pH 7.4). Cell suspension was lysed by adding lysozyme (5 mg/ml), MgCl
28.6 mg/ml), RNase, and DNase (10 mg/ml of each) followed by incubation at 4°C
for 60 minutes. The cell suspension was then disrupted by sonication and
subsequently subjected to centrifugation at 100,000g for 40 minutes at 4°C.
Supernatant containing recombinant human XO was collected and stored at –80°C
until further use.
Expression of Human AOX1 and Subsequent Protein Purification. The
method of human AOX1 gene expression and purification previously
developed in our laboratory was used in this study for the preparation of
purified human AO (Alfaro et al., 2009). Human AOX1 was overexpressed as
a fusion protein with an N-terminal hexa-His tag in TP1000 cells. Partial
purification was achieved by lysing the cells and passing the lysate through
a 1-ml HiTrap Chelating HP column (GE Healthcare, Little Chalfont, Bucking-
hamshire, UK). The purified protein was subsequently dialyzed into 100 mM
potassium phosphate buffer, pH 7.4, and snap frozen in liquid nitrogen and
stored at –80°C.
2
(
6
TX, 6TUA, and the internal standard, were detected using multiple reaction
monitoring mode by monitoring the m/z transition from 167.1 to 133.4, 183.0 to
40.2, and 294.8 to 251, respectively. Quantitation of the product was achieved
by comparison with a standard curve ranging from 1 nM to 10 mM for 6TX and
1
0.1 to 10 mM for 6TUA.
Data Analysis. Experimental data were fitted to appropriate nonlinear
regression models using GraphPad Prism (version 4.03; GraphPad Software
Inc., San Diego, CA).
Michaelis-Menten model:
Vmax½Sꢁ
V ¼
ð1Þ
Km þ ½Sꢁ
Quantitation of Purified AO by High-Performance Liquid Chromatography–
Tandem Mass Spectrometry with Electrospray Interface. The amount of AO in
purified samples was quantified using a technique previously developed in our
laboratory (Barr et al., 2013). A deuterium-labeled synthetic peptide standard was
used to determine the amount of AO in purified samples. A sample consisting of
Substrate inhibition model:
V ¼
Vmax½Sꢁ
Km þ ½Sꢁ 1þ
ꢀ
Sꢁꢁ
ð2Þ
½
Ki
2
5 ml of purified AO (1.5 mM) was mixed with an equal volume of denaturing
solution containing 8 M urea and 2 mM dithiothreitol and incubated at 60°C for
0 minutes. Sodium bicarbonate buffer (25 mM, pH 8.4) containing 100 nM
where V is the reaction velocity, [S] is the substrate concentration, Vmax is the
maximum reaction velocity, K is the Michaelis-Menten constant, and K is the
inhibition constant for the substrate.
6
m
i
internal standard peptide was added to the mixture so that the final reaction
volume was 250 ml. Ten milliliters of 20-mg trypsin was added to the reaction
mixture and incubated for 15 hours at 37°C. The reaction was quenched by
adding 50 ml of 50% v/v trifluoroacetic acid (TFA) in water to 200 ml of reaction
mixture. The samples were vortexed and centrifuged at 1460g for 10 minutes and
analyzed using high-performance liquid chromatography–tandem mass spec-
trometry (HPLC-MS/MS). The native peptide and hence the amount of purified
AO in a sample was quantified by comparing its MS peak area to the peak area of
the internal standard peptide of a known concentration.
Visual inspection of the velocity curve inflection (Figs. 3A and 5A), as well
as the biphasic nature of the kinetic displayed by Eadie-Hofstee plot (Fig. 3Ai),
led us to fit the data to the equation for two or more enzymes catalyzing the
same reaction. This kinetic approach considers the total velocity as the sum of
the contribution by each enzyme (Segel, 1993):
ꢂ
ꢃ
ꢂ
ꢃ
Vmax ½Sꢁ
Vmax ½Sꢁ
1
2
V ¼
þ
ð3Þ
Km þ½Sꢁ
Km þ½Sꢁ
1
2
Incubation Conditions. The incubation mixture consisted of substrates
Here, Vmax1 is the lower Vmax, Vmax2 is the higher Vmax, Km1 is the lower K
m2 is the higher K
Lineweaver-Burk plots were used to estimate the kinetic constants Vmax and K
m
and
6
MP or 6TX at a final concentration in the range of 10–1000 mM and 0.1–200
mM, respectively, in 25 mM potassium phosphate buffer (pH 7.4) containing
.1 mM EDTA. The final concentration of dimethylsulfoxide was 0.5% (v/v).
K
m
. Since nonlinear regression methods failed to converge,
m
0
for biphasic reactions. Reciprocal plots generated using data points from the
For the inhibition experiments, a substrate concentration of 100 mM was used.
Inhibitors raloxifene and febuxostat were used at a final concentration of 1 mM.
The substrate and inhibitor stocks were made up so that the total concentration
of dimethylsulfoxide in the reaction mixture was 1.5% (v/v). All incubations
for 6TX and 6TUA formation were allowed to proceed at 37°C for 15 and 30
minutes, respectively. The formation of 6TX was observed to be linear with
respect to time for 15 minutes and 6TUA for more than 30 minutes. The
reaction was initiated by addition of pooled human liver cytosol (HLC) at
a final concentration of 0.5 or 1 mg/ml and a known concentration of purified
region below the first inflection of the biphasic curve were used to obtain Km1 and
Vmax1, and points above the first inflection were used to predict Km2 and Vmax2.
Results and Discussion
Metabolism of 6MP to 6TUA by the molybdenum hydroxylases
requires two sequential oxidations. Evidence in the literature suggests
that 6MP can form 6TUA either via 6TX or 8-OH-6MP as an
human AO and human XO lysate. The final reaction volume was 80 ml. The intermediate in vivo in humans, but the enzymes responsible for the
reaction was quenched by the addition of 20 ml of 1 M formic acid containing reactions are not known (Bergmann and Ungar, 1960; Zimm et al.,

Oxidative Metabolism of 6-Mercaptopurine
1337
metabolism with human XO and AO (Sahi et al., 2008; Choughule
et al., 2013; Dalvie et al., 2013).
We initially explored the time course of 6MP metabolism by
monitoring the production of the intermediate 6TX and the final product
6TUA in pooled human liver cytosol. The results are shown in Fig. 2.
The intermediate 6TX is formed at a relatively linear rate for almost 25
minutes. Between 25 and 40 minutes the intermediate reaches steady-state
concentrations. The final product is formed after an initial lag of about
25–40 minutes and produced in a linear fashion after the intermediate
6TX reached steady-state. These results are consistent with sequential
oxidation with release of the intermediate prior to the second step. Thus,
the intermediate 6TX must rebind to the enzyme and does not simply
reorient in the active site. Due to the unavailability of the 8-OH-6MP
metabolite standard, it was not possible to investigate the contribution of
this intermediate. However, a recent crystallography study has shown that
both 6MP and hypoxanthine bind in two orientations that could lead to
either oxidation on the five-membered or six-membered ring of either
substrate (Cao et al., 2010). While 6MP metabolism was not explored,
this study concluded that xanthine (oxidation on the six-membered ring)
Fig. 2. Dynamics of the formation of the intermediate, 6TX and the final product,
6
TUA in human liver cytosol over a period of 150 minutes.
1
984; Van Scoik et al., 1985) (see Fig. 1). Studies using cow XO and was the major intermediate of hypoxanthine oxidation and that oxidation
guinea pig AO concluded that oxidation to the 6TX intermediate was on the five-membered ring was not observed. Thus, the formation of 6TX
mediated solely by XO. The second step was found to be mediated by (oxidation on the six-membered ring) is consistent with the results for
both guinea pig AO and cow XO (Rashidi et al., 2007). Given that hypoxanthine oxidation. Furthermore, the correlation between steady-state
many studies have shown that animal models are poor predictor of 6TX levels and 6TUA linear rates support 6TX as the major intermediate,
molybdenum oxidase activity we decided to explore the in vitro or that 6TX and 8-OH-6MP are formed with nearly identical kinetics.
+
Fig. 3. (A) Biphasic curve showing the formation of the intermediate 6TX from 6MP in human liver cytosol in the absence of NAD . Points represent mean and error bars
show standard error for triplicate experiments. Shown as an inset are Eadie Hofstee plot (A-I) and Lineweaver-Burke plot (A-II) representation of the same data. Points
represent an average of triplicate experiments. (B) Inhibition of 6TX formation from 100 mM 6MP by AO specific inhibitor raloxifene (1 mM) and XO specific inhibitor
febuxostat (1 mM) in HLC. Each reaction was performed in duplicate and values are expressed as percentage activity relative to the activity observed in the control (No
21
inhibitor added). Error bars represent 6 SEM. (C) Saturation plot for 6MP oxidation to 6TX in purified human AO showing Kcat (min ) versus substrate concentration. Data
are fit to a substrate inhibition model and are an average of duplicate experiments.

1
338
Choughule et al.
TABLE 2
Enzyme kinetic parameters for the conversion of 6-mercaptopurine to 6-thioxanthine and 6-thioxanthine to 6-thiouric
+
acid in pooled human liver cytosol in the presence and absence of NAD
Data are representative of at least duplicate measurements 6 S.E.M.
V
max
K
m
Reaction
Source
Enzyme Kinetics
V
max1
V
max2
K
m1
K
m2
nmol/min per milligram
mM
6
6
MP → 6TX
HLC (–NAD)
HLC (+NAD)
HLC (–NAD)
HLC (+NAD)
Biphasic
Biphasic
Michaelis-Menten
Michaelis-Menten
0.08 6 0.00
0.2 6 0.01
7.2 6 0.2
0.2 6 0.1
0.3 6 0.1
NA
88.2 6 4.4
83.4 6 2.9
3.0 6 0.4
8.0 6 1.6
500 6 188
339 6 169
NA
TX → 6TUA
15.5 6 0.7
NA
NA
With 6TX established as an intermediate in the formation of 6TUA, XO also converted 6MP to 6TX, but since the enzyme levels are very
the next step was to determine the kinetics of the 6TX formation. Fig. low in the unpurified lysate, it was not possible to perform saturation
A shows that biphasic kinetics was observed for 6TX formation. The kinetics and subsequently obtain a Km for this reaction for comparison
3
Eadie-Hofstee plot shown in Fig. 3Ai, also displays a biphasic profile. with purified human AO. Thus, it appears that unlike guinea pig AO,
Biphasic profiles are generally a result of a single enzyme in possession human AO mediates the formation of 6TX, while both human and cow
of a low- and a high-affinity catalytic binding site or two distinct XO can also mediate this reaction.
enzymes involved in the biotransformation of a single substrate, one
having low and the other having high affinity toward the substrate Michaelis-Menten kinetics were observed for the conversion of the
Hutzler and Tracy, 2002). After using Lineweaver-Burke plots to fit the intermediate 6TX to final product 6TUA in HLC. This indicates that
data (Fig. 3Aii), Vmax1/Vmax2 and Km1/Km2 were estimated to be 0.08/ a single enzyme, or more than one enzyme with similar Km values,
.2 nmol/min per milligram and 88/500 mM, respectively (Table 2). were involved in the catalysis of this step (Fig. 4A). Inhibition
Unlike 6TX formation, which followed biphasic kinetics,
(
0
Thus, one high-affinity enzyme/active site and one high-capacity experiments with raloxifene indicated that AO was not involved in the
enzyme/active site is involved in metabolism to the 6TX intermediate, conversion of this intermediate to the final product. Raloxifene
but the identity of the specific enzymes are unknown.
inhibited only 9% of 6TUA formation, whereas the XO inhibitor
Specific inhibitors were used to address whether both AO and XO febuxostat almost completely inhibited 6TUA formation from 6TX
are involved in 6TX production. Raloxifene and febuxostat specifi- (Fig. 4B). In addition to this, purified human AO did not turn over
cally inhibit drug metabolism by AO and XO, respectively, at 6TX to produce 6TUA but E. coli lysate containing expressed human
concentrations up to 1 mM. (Obach, 2004; Barr and Jones, 2013; XO converted 6TX to 6TUA following Michaelis-Menten kinetics
Weidert et al., 2014). Raloxifene inhibited the formation of 6TX by (Fig. 4C). These results underpin the importance of XO in the
approximately 39%, whereas febuxostat demonstrated 67% inhibition metabolism of the intermediate to the final product and the absence of
(
Fig. 3B). When both inhibitors were used in the same incubation they AO contribution to this step. This is again in contrast with the 6TX
inhibited 97% of the formation of 6TX (Fig. 3B). Neither raloxifene metabolism by guinea pig AO and cow XO (Rashidi et al., 2007), both
nor febuxostat independently resulted in complete inhibition of 6TX of which catalyze the reaction. Interestingly, human liver cytosol has
production, implying that both AO and XO are involved in the a higher Vmax and a lower K
m
value of 3 mM than the E. coli-expressed
generation of this intermediate metabolite. (Table 2). enzyme (K , 30 mM) (Table 3). Given the uncertainty in the protein
m
To identify the high- and low-Km enzyme, 6MP was incubated with concentration of XO, the rate differences between the cytosolic XO
either purified recombinant human aldehyde oxidase or E. coli cell and the expressed XO cannot be explained. However, it is surprising
lysate containing recombinant human xanthine oxidase. Purified that the K
m
values are different for the two enzyme systems. One
human AO metabolized 6MP to 6TX and showed substrate inhibition possibility is that the human liver cytosol may have a significant
2
1
kinetics with a Kcat of 0.2 minute and a Km of 572 mM (Fig. 3C). amount of xanthine dehydrogenase (XDH) as well as XO.
The K value of 572 mM obtained here (Table 3) is not significantly To assess the contribution, if any, of XDH to the formation of 6TUA
m
different from the Km2 value of 500 mM reported for 6TX formation in from 6TX, we performed kinetic experiments on human liver cytosol in
+
human liver cytosol above and hence it is probable that AO is the the presence of NAD . The kinetics of 6TUA formation from 6TX in the
+
high-K
m
enzyme associated with 6TX formation in human liver presence of NAD showed Michaelis-Menten kinetics with a doubling of
+
cytosol. This also means that XO is the likely low-K
m
enzyme
Vmax from 7 to 15 nmol/min per milligram relative to no NAD (Fig. 5A).
associated with this reaction. Escherichia coli lysate containing human An increase in the K value from 3 to 8 mM was observed in the presence
m
TABLE 3
Enzyme kinetic parameters for the conversion of 6-mercaptopurine to 6-thioxanthine by purified recombinant human aldehyde oxidase and 6-thioxanthine to 6-thiouric acid
by E. coli lysate containing recombinant human xanthine oxidase
Data are representative of at least duplicate measurements 6 S.E.M.
V
max
K
m
Reaction
Source
Enzyme Kinetics
K
cat
K
i
V
max1
V
max2
K
m1
K
m2
21
nmol/min per milligram
min
mM
mM
6
6
MP → 6TX
TX → 6TUA
Purified human AO
Human XO containing
E. coli lysate
Substrate Inhibition
Michaelis-Menten
NA
NA
NA
NA
0.2 6 0.0
NA
572 6 172
30 6 2
NA
NA
244 6 77
NA

Oxidative Metabolism of 6-Mercaptopurine
1339
Fig. 4. (A) Michaelis-Menten kinetics for the formation of the final product 6TUA from the intermediate, 6TX in HLC and (C) lysate containing expressed human xanthine
oxidase. Each experiment is an average of duplicate experiments (6 SEM). (B) Inhibition of 6TUA formation from 100 mM 6TX in HLC by AO specific inhibitor raloxifene
(1 mM) and XO specific inhibitor febuxostat (1 mM). Values are expressed as a percentage of activity relative to the activity observed in of the control (No inhibitor added).
Data are an average of duplicate experiments 6 SEM.
+
+
of NAD (Table 2). In short, adding NAD to human liver cytosol did between the cytosolic and E. coli-expressed metabolism. This brings into
increase the rate of the reaction for the formation of the final product from question whether XDH may also play a role in the first rate-limiting step,
the intermediate 6TX and can explain, at least in part, the difference the production of 6TX.
Fig. 5. (A) Metabolism of 6MP to form 6TUA
from 6TX and (B) 6TX from 6MP in the
+
presence (red) and absence (blue) of NAD .
HLC were used as a source of enzyme. Data
points are an average of duplicate experiments
6
SEM. Shown as (B-I) is a Lineweaver-Burke
plot of the same data used for estimation of
kinetic parameters such as Vmax and K
m
.

1
340
Choughule et al.
+
Bergmann F and Ungar H (1960) The enzymatic oxidation of 6-mercaptopurine to 6-thiouric acid.
J Am Chem Soc 82:3957–3960.
In the presence of NAD , kinetics of 6TX was still biphasic
(
Fig. 5B) but the kinetic parameters indicate a 2.5-fold increase in Vmax1
Burchenal JH, Murphy ML, Ellison RR, Sykes MP, Tan TC, Leone LA, Karnofsky DA, Craver
LF, Dargeon HW, and Rhoads CP (1953) Clinical evaluation of a new antimetabolite, 6-
mercaptopurine, in the treatment of leukemia and allied diseases. Blood 8:965–999.
Cao H, Pauff JM, and Hille R (2010) Substrate orientation and catalytic specificity in the action of
xanthine oxidase: the sequential hydroxylation of hypoxanthine to uric acid. J Biol Chem 285:
value, with no significant difference in the Vmax2 value (Table 2). (This
would be expected since the second Vmax value has been assigned to
AO.) Thus, it appears that XDH and XO contribute to the high-affinity
reaction, while AO is responsible for the high-capacity reaction.
In conclusion, 6MP is converted to 6TUA via the intermediate 6TX.
Three enzymes, AO, XO, and XDH, contribute to the production of
the 6TX intermediate, whereas only XO and XDH are involved in the
conversion of the intermediate to the final product. This is in direct
contrast with studies using nonhuman tissue in which XO was found
to be responsible for the first step, and both AO and XO where
responsible for the second step. This study establishes the enzymes
responsible for oxidation of 6MP, and underlines the potential
problems associated with using other species to predict the role of
human molybdenum oxidases.
2
8044–28053.
Choughule KV, Barr JT, and Jones JP (2013) Evaluation of rhesus monkey and guinea pig
hepatic cytosol fractions as models for human aldehyde oxidase. Drug Metab Dispos 41:
1
852–1858.
Dalvie D, Xiang C, Kang P, and Zhou S (2013) Interspecies variation in the metabolism of
zoniporide by aldehyde oxidase. Xenobiotica 43:399–408.
Elion GB (1967) Symposium on immunosuppressive drugs. Biochemistry and pharmacology of
purine analogues. Fed Proc 26:898–904.
Garattini E and Terao M (2012) The role of aldehyde oxidase in drug metabolism. Expert Opin
Drug Metab Toxicol 8:487–503.
Garattini E, Mendel R, Romão MJ, Wright R, and Terao M (2003) Mammalian molybdo-
flavoenzymes, an expanding family of proteins: structure, genetics, regulation, function and
pathophysiology. Biochem J 372:15–32.
Giverhaug T, Loennechen T, and Aarbakke J (1999) The interaction of 6-mercaptopurine (6-MP)
and methotrexate (MTX). Gen Pharmacol 33:341–346.
Hall WW and Krenitsky TA (1986) Aldehyde oxidase from rabbit liver: specificity toward
purines and their analogs. Arch Biochem Biophys 251:36–46.
Harrison R (2002) Structure and function of xanthine oxidoreductase: where are we now? Free
Acknowledgments
Radic Biol Med 33:774–797.
Hutzler JM and Tracy TS (2002) Atypical kinetic profiles in drug metabolism reactions. Drug
Metab Dispos 30:355–362.
Keuzenkamp-Jansen CW, van Baal JM, De Abreu RA, de Jong JG, Zuiderent R, and Trijbels JM
The authors thank Dr. Tomohiro Matsumura at Nippon Medical School
Tokyo, Japan) for providing us with the plasmid pTrc99A containing the gene
encoding human XO.
(
(
1996) Detection and identification of 6-methylmercapto-8-hydoxypurine, a major metabolite
of 6-mercaptopurine, in plasma during intravenous administration. Clin Chem 42:380–386.
Krenitsky TA, Neil SM, Elion GB, and Hitchings GH (1972) A comparison of the specificities of
xanthine oxidase and aldehyde oxidase. Arch Biochem Biophys 150:585–599.
Nielsen OH, Vainer B, and Rask-Madsen J (2001) Review article: the treatment of inflammatory
bowel disease with 6-mercaptopurine or azathioprine. Aliment Pharmacol Ther 15:1699–1708.
Obach RS (2004) Potent inhibition of human liver aldehyde oxidase by raloxifene. Drug Metab
Dispos 32:89–97.
Rashidi MR, Beedham C, Smith JS, and Davaran S (2007) In vitro study of 6-mercaptopurine
oxidation catalysed by aldehyde oxidase and xanthine oxidase. Drug Metab Pharmacokinet 22:
299–306.
Authorship Contributions
Participated in research design: Choughule, Joswig-Jones, Jones.
Conducted experiments: Choughule, Joswig-Jones.
Performed data analysis: Choughule, Jones, Barnaba.
Wrote or contributed to the writing of the manuscript: Choughule, Barnaba,
Jones.
Rowland K, Lennard L, and Lilleyman JS (1999) In vitro metabolism of 6-mercaptopurine by
human liver cytosol. Xenobiotica 29:615–628.
References
Sahi J, Khan KK, and Black CB (2008) Aldehyde oxidase activity and inhibition in hepatocytes
and cytosolic fractions from mouse, rat, monkey and human. Drug Metab Lett 2(3):176–183.
Segel I (1993) Kinetics of unireactant enzymes, in Enzyme Kinetics (Segal I ed.) pp 18–98, John
Wiley & Sons, New York.
Van Scoik KG, Johnson CA, and Porter WR (1985) The pharmacology and metabolism of the
thiopurine drugs 6-mercaptopurine and azathioprine. Drug Metab Rev 16:157–174.
Weidert ER, Schoenborn SO, Cantu-Medellin N, Choughule KV, Jones JP, and Kelley EE (2014)
Inhibition of xanthine oxidase by the aldehyde oxidase inhibitor raloxifene: implications for
identifying molybdopterin nitrite reductases. Nitric Oxide 37:41–45.
Aarbakke J, Janka-Schaub G, and Elion GB (1997) Thiopurine biology and pharmacology.
Trends Pharmacol Sci 18:3–7.
Alfaro JF and Jones JP (2008) Studies on the mechanism of aldehyde oxidase and xanthine
oxidase. J Org Chem 73:9469–9472.
Alfaro JF, Joswig-Jones CA, Ouyang W, Nichols J, Crouch GJ, and Jones JP (2009) Purification
and mechanism of human aldehyde oxidase expressed in Escherichia coli. Drug Metab Dispos
3
7:2393–2398.
Balis FM, Holcenberg JS, Zimm S, Tubergen D, Collins JM, Murphy RF, Gilchrist GS, Ham-
mond D, and Poplack DG (1987) The effect of methotrexate on the bioavailability of oral 6-
mercaptopurine. Clin Pharmacol Ther 41:384–387.
Zimm S, Grygiel JJ, Strong JM, Monks TJ, and Poplack DG (1984) Identification of 6-mer-
captopurine riboside in patients receiving 6-mercaptopurine as a prolonged intravenous in-
fusion. Biochem Pharmacol 33:4089–4092.
Barr JT and Jones JP (2013) Evidence for substrate-dependent inhibition profiles for human liver
aldehyde oxidase. Drug Metab Dispos 41:24–29.
Barr JT, Jones JP, Joswig-Jones CA, and Rock DA (2013) Absolute quantification of aldehyde
oxidase protein in human liver using liquid chromatography-tandem mass spectrometry. Mol
Pharm 10:3842–3849.
Barr JT, Choughule KV, Nepal S, Wong T, Chaudhry AS, Joswig-Jones CA, Zientek M, Strom
SC, Schuetz EG, and Thummel KE, et al. (2014) Why do most human liver cytosol prepa-
rations lack xanthine oxidase activity? Drug Metab Dispos 42:695–699.
Address correspondence to: Dr. Jeffrey P. Jones, Department of Chemistry,
Washington State University, P.O. Box 644630, Pullman, WA 99164-4630. E-mail:
jpj@wsu.edu