Spirodihydantoin is a minor product of 5-hydroxyisourate in urate oxidation

Article abstract of DOI:10.1021/ol048547w

(Chemical Equation Presented) Spirodihydantoin is a minor product from oxidation of uric acid (～0.15% yield), while spiroiminodihydantoin is a major product from oxidation of 8-oxo-7,8-dihydroguanine (37% yield, pH 10.2). High pH and temperature favor the formation of both spiro compounds. ¹⁸O labeling experiments and in situ generation and decomposition of 5-hydroxy-N7-methylisouric acid indicate that spirodihydantoin and allantoin and spiroiminodihydantoin and guanidinohydantoin are products of 5-hydroxyisourate and 5-hydroxy-8-oxo-7,8-dihydroguanine intermediates, respectively.

Full text of DOI:10.1021/ol048547w

ORGANIC
LETTERS
2004
Vol. 6, No. 19
3417-3420
Spirodihydantoin Is a Minor Product of
5-Hydroxyisourate in Urate Oxidation
Hongbin Yu,† Jacquin C. Niles,† John S. Wishnok,† and
,†,‡
Steven R. Tannenbaum*
Biological Engineering DiVision and Department of Chemistry,
Massachusetts Institute of Technology, 77 Massachusetts AVenue, Room 56-731A,
Cambridge, Massachusetts 02139
srt@mit.edu
Received July 26, 2004
ABSTRACT
Spirodihydantoin is a minor product from oxidation of uric acid (∼0.15% yield), while spiroiminodihydantoin is a major product from oxidation
of 8-oxo-7,8-dihydroguanine (37% yield, pH 10.2). High pH and temperature favor the formation of both spiro compounds. 18O labeling experiments
and in situ generation and decomposition of 5-hydroxy-N7-methylisouric acid indicate that spirodihydantoin and allantoin and spiroiminodi-
hydantoin and guanidinohydantoin are products of 5-hydroxyisourate and 5-hydroxy-8-oxo-7,8-dihydroguanine intermediates, respectively.
8-OxodG,1 an important biomarker of DNA oxidative
damage, is prone to further oxidation. Nucleosides dGh and
dSp are two major products2,3 from decomposition of 5-OH-
8-oxodG, by analogy to 5-OH-UA from oxidation of UA.4,5
Recently the long postulated 5-OH-8-oxoguanosine was
detected by NMR at low temperature6 and rearranged to Sp
nucleoside at room temperature. Hydration/decarboxylation
of C6-carbonyl of 5-OH-8-oxodG leads to dGh, similar to
the formation of Alla from 5-OH-UA, while the rearrange-
ment of 5-OH-8-oxodG leads to dSp. Interestingly, a similar
rearrangement of 5-OH-UA to Spd has not been reported.
In fact, during extensive study of oxidation of UA in the
past few decades, only one communication has provided
indirect evidence for the formation of Spd from oxidation
of UA.7 We report here detection and quantitation of this
compound from oxidation of UA by isotope dilution mass
spectrometry. Mechanistic studies to confirm Spd as a
product of 5-OH-UA and the comparison between the
product distribution of oxidation of 8-oxoG and UA are also
reported.
UA was oxidized by Na2IrCl6,3 peroxynitrite,8 NO2,9 and
uricase/O2,5 and 8-oxoG was oxidized by Na2IrCl6. The first
three oxidants are known to oxidize 8-oxodG to dSp via
5-OH-8-oxodG, and uricase/O2 generates 5-OH-UA inter-
mediate from oxidation of UA.5 Our preliminary investigation
indicated that the yield of Spd was very low. To unambigu-
ously detect and quantitate this compound, authentic Spd
† Biological Engineering Division.
‡ Department of Chemistry.
(1) Abbreviations used: 8-oxo-7,8-dihydro-2′-deoxyguanosine, 8-oxodG;
8-oxo-7,8-dihydroguanine, 8-oxoG; uric acid, UA; spirodihydantoin, Spd;
spiroiminodihydantoin, Sp; guanidinohydantoin, Gh; 5-hydroxyisourate,
5-OH-UA; 5-hydroxy-N7-methylisouric acid, 5-OH-N7-methylUA; 5-hy-
droxy-8-oxoguanine, 5-OH-8-oxoG; 5-hydroxy-8-oxodG, 5-OH-8-oxodG;
allantoin, Alla; horseradish peroxidase, HRP; myeloperoxidase, MPO.
(2) Luo, W.; Muller, J. G.; Rachlin, E. M.; Burrows, C. J. Chem. Res.
Toxicol. 2001, 14, 927-938.
(3) Luo, W.; Muller, J. G.; Rachlin, E. M.; Burrows, C. J. Org. Lett.
2000, 2, 613-616.
(4) Modric, N.; Derome, A. E.; Ashcroft, S. J. H.; Poje, M. Tetrahedron
Lett. 1992, 33, 6691-6694.
(7) Modric, N.; Poje, M.; Gojmerac-Ivsic, A. Bioorg. Med. Chem. Lett.
1994, 4, 1685-1686.
(5) Kahn, K.; Serfozo, P.; Tipton, P. A. J. Am. Chem. Soc. 1997, 119,
5435-5442.
(8) Niles, J. C.; Wishnok, J. S.; Tannenbaum, S. R. Org. Lett. 2001, 3,
963-966.
(6) McCallum, J.; Foote, C. S. Abstracts of Papers, 227th National
Meeting of the American Chemical Society, Anaheim, CA, March 28-
April 1, 2004; American Chemical Society: Washington, DC, 2004.
(9) Shafirovich, V.; Cadet, J.; Gasparutto, D.; Dourandin, A.; Geacintov,
N. E. Chem. Res. Toxicol. 2001, 14, 233-241.
10.1021/ol048547w CCC: $27.50 © 2004 American Chemical Society
Published on Web 08/25/2004

and 2-13C-1,3-15N2-Spd were synthesized on the basis of a
modification of Poje’s procedure.10 Addition of acetonitrile
as a cosolvent shortened the reaction time and increased the
yield of Spd. The identities of both Spd compounds were
confirmed by high-resolution MS (theoretical mass of [M
- H]- of Spd 183.0149, actual mass 183.0152; theoretical
mass of [M - H]- of labeled Spd 186.0149, actual mass
186.0123). In negative ion mode of ESI-TOF MS, unlabeled
and labeled Spd also gave fragment ions of 140.0 (from
Spd), 141.0, 142.0, and 143.0 (from labeled Spd) (Figure
1), corresponding to the loss of neutral fragments NHCO,
mediates. Alla and Spd and Gh and Sp bases were isolated
from oxidation of UA and 8-oxoG, respectively. The
identities of these compounds were confirmed by comparing
their HPLC retention times, UV spectra, and mass spectro-
metric properties with those of the authentic standards or
those reported in the literature.2,3 Alla, Gh, and Sp bases
were quantitated with the external standards by HPLC. Spd
was quantitated by isotope dilution mass spectrometry with
labeled Spd as the internal standard. The dose-response,
temperature, and pH dependence of the formation of Spd
and Alla and the pH dependence of the formation of Sp and
Gh bases are summarized in Table 1. Alla represented the
bulk of the products from oxidation of UA, and its yield
increased in a dose-dependent fashion up to 2 equiv of Na2-
IrCl6, which completely oxidized UA. Interestingly, Alla was
not detectable when >5 equiv of Na2IrCl6 was used due to
its further oxidation (products not determined). Similar
oxidation of dGh was observed by Burrows and co-workers.2
In the temperature and pH dependence studies, 2 equiv of
Na2IrCl6 was used to maximize the yield of Spd without
further oxidation of Alla. The yield of Alla was in the range
of 67-94% except at high pH and temperature, where it
dropped significantly, partly due to its decomposition under
these conditions.13 The yield of Spd was in the range of
0.08-0.92% and demonstrated a modest dose response.
However, the temperature and pH dependence were more
significant, with high pH and temperature favoring its
formation (no decomposition of Spd observed at pH 12.2 in
4 h). Control experiments with no addition of Na2IrCl6 were
conducted to accompany the above studies. No oxidation of
UA was observed, and Spd was not detectable. Gh and Sp
bases represented the major products from oxidation of
8-oxoG. Their formation showed significant pH dependence,
with high pH favoring the formation of Sp base and
disfavoring that of Gh base. No Sp base was detectable at
pH < 8, and its yield increased to 58% at pH 11.2. It is
important to note that similar pH and temperature dependence
were observed for the formation of dSp from oxidation of
8-oxodG.3,14
Figure 1. Mass spectra of Spd and 2-13C-1,3-15N2-Spd
15NHCO, or 15NH13CO from [M - H]-, which are typical
for compounds containing a hydantoin structure.3,11 The
isotopic purity of labeled Spd was 99.4% based on ESI-MS
measurements. Additionally, 13C NMR of Spd indicated a
resonance at 76.2 ppm, assigned as the spiro carbon because
of its similar position to quaternary carbon resonances in
other spiro compounds.3,7,12
Results of oxidation of UA by peroxynitrite (ONOO-) and
enzyme-mediated oxidants are summarized in Table 2 (HRP/
- 15
- 15
H2O2, HRP/H2O2/NO2 , MPO/H2O2, MPO/H2O2/NO2 ,
and uricase/O2 designated as 1-5, respectively). The dose
response of the formation of Spd and Alla from ONOO-
Na2IrCl6 was used extensively in oxidation of UA and
8-oxoG because of its ability to generate the 5-OH inter-
a
Table 1. Summary of Oxidation of UA and 8-oxoG by Na2IrCl6
dose response (pH 8, UA)
temp dependence (pH 8, UA)
pH dependence (UA and 8-oxoG)
Ir (equiv)
Sp d (%)
Alla (%)
T (°C)
Sp d (%)
Alla (%)
pH
Sp d (%)
Alla (%)
Sp (%)
Gh (%)
0.2
0.5
1
2
5
0.15
0.08
0.10
0.10
0.11
0.12
0.16
47
55
76
75
nd
nd
nd
0
24
50
62
84
0.15
0.11
0.34
0.55
0.92
67
78
80
82
40
6.7
7.4
8.0
0.09
0.09
0.10
0.15
0.17
0.19
0.16
94
81
74
84
84
28
15
nd
nd
minor
12
37
major
major
major
79
57
33
9.5
10.2
11.2
12.2
10
20
58
a Nd ) not detectable; all yields based on consumed UA and 8-oxoG.
3418
Org. Lett., Vol. 6, No. 19, 2004

in situ from hydration of 5-chloro-N7-methylUA in potas-
sium phosphate buffer.17 2-Methyl-Spd (minor product) and
3-methyl-Alla (major product) were isolated from the
decomposition of 5-chloro-N7-methylUA as well as from
the oxidation of N7-methylUA by Na2IrCl6. This unambigu-
ously demonstrates that these two compounds are products
of 5-OH-N7-methylUA in the oxidation of N7-methylUA
and strongly suggests that Spd and Alla are products of
5-OH-UA in the oxidation of UA.
Table 2. Summary of Oxidation of UA by ONOO- and
Enzyme-mediated Oxidantsa
ONOO-
(equiv)
Sp d (%) Alla (%) enzyme Sp d (%) Alla (%)
0.2
0.5
1.0
2.0
10
nd
nd
0.09
0.14
0.06
0.14
3
8
12
16
nd
nd
1
2
3
4
5
0.04
0.04
0.06
0.07
0.006
8
10
21
19
56
Formation of analogous intermediates and similar pH
dependence of formation of the spiro compounds suggest
that the oxidation of UA and 8-oxoG share similar pathways
(Scheme 1). The significant difference between the yields
20
a Nd ) not detectable.
oxidation was carried out in pH 7.2 potassium phosphate
and sodium bicarbonate buffer via bolus addition of ONOO-.
The yield of Alla was <16% when < 2 equiv of ONOO-
were used and not detectable at >10 equiv of ONOO-
treatment. The formation of Spd (∼0.1% yield) showed no
significant dose response. In the uricase/O2 oxidation, the
yield of Alla was 56%, while the yield of Spd (0.006%)
was significantly lower than that in oxidation of UA by Na2-
IrCl6. In the HRP- or MPO-mediated oxidation, the yields
of Alla and Spd were in the range of 8-21% and 0.04-
Scheme 1. Decomposition of 5-OH-UA and 5-OH-8-oxoG
-
0.07%, respectively, and did not differ significantly if NO2
was present, suggesting no significant oxidation by NO2.
Control experiments with no addition of ONOO- or enzymes
were carried out to ensure no false positive formation of Spd.
To confirm that Spd and Alla and Sp and Gh bases are
products of 5-OH-UA and 5-OH-8-oxoG intermediates,
respectively, we carried out oxidation of UA and 8-oxoG
by Na2IrCl6 both in H216O and H218O (95% enriched) in pH
9.5 phosphate buffer. The isolated Spd, Alla, Sp, and Gh
bases were analyzed by ESI-TOF MS in negative ion mode,
and those reactions done in H216O gave [M - H]- at 183.0,
157.0, 182.0, and 156.1, respectively. From oxidation in
H218O, [M - H]- ions at 185.0, 159.0, 184.0, and 158.1 were
observed (see Supporting Information), suggesting that the
OH group in 5-OH-UA and 5-OH-8-oxoG comes from water
and that the oxygen atom is retained in the products. This
supports the intermediacy of 5-OH-UA and 5-OH-8-oxoG.
The definitive proof that Spd is a product of 5-OH-UA
would require independent synthesis of 5-OH-UA and
isolation of Spd from its decomposition. Because 5-OH-
UA is very unstable, we used 5-OH-N7-methylUA as its
surrogate because oxidations of UA and N7-methylUA
proceed via similar intermediates, namely, 5-OH-UA and
5-OH-N7-methylUA.16 5-OH-N7-MethylUA was generated
of Spd and Sp base is very intriguing. The absence of
significant formation of 1-methyl-Spd (<9% yield) from
oxidation of N9-methylUA by Na2IrCl6 excluded the role
of N9-substituents in causing the low yield of Sp. The
transient existence of 5-OH-UA and 5-OH-8-oxoG has
limited efforts to probe their decomposition mechanism.
However, the physicochemical properties of stable analogues
suggest that deprotonation of 5-OH facilitates the formation
of spiro compounds via pathway II,18,19 and enolization of
the C6 carbonyl slows the hydrolysis of the N1-C6 amide
bond and pathway I.20
On the basis of observations from Burrows21,22 and our
group,23 we propose pathways I and II (Scheme 1) to be
competitive. The predominant form of 5-OH-UA in pH range
7-11 is the N9-deprotonated species (1)5,24 (Scheme 2),
while the predominant form of 5-OH-8-oxoG is the neutral
species (4)25 at pH 7. At pH ∼ 10, N9 of 5-OH-8-oxoG is
deprotonated (6) (N9 pKa ∼ 9 using structural similarity
(17) Subramanian, P.; Nguyen, N. T.; Dryhurst, G.; Hutzenlaub, W.;
Pfleiderer, W. J. Electroanal. Chem. Interfacial Electrochem. 1989, 263,
353-371.
(18) Iwata, M.; Bruice, T. C.; Carrell, H. L.; Glusker, J. P. J. Am. Chem.
Soc. 1980, 102, 5036-5044.
(10) Poje, M.; Paulus, E. F.; Rocic, B. J. Org. Chem. 1980, 45, 65-68.
(11) Volk, K. J.; Yost, R. A.; Brajter-Toth, A. Anal. Chem. 1989, 61,
1709-1717.
(19) Bruice, T. C.; Miller, A. J. Chem. Soc., Chem. Commun. 1980, 693-
694.
(12) Johnson, F.; Huang, C.-Y.; Yu, P.-L. EnViron. Health Perspect.
(Suppl.) 1994, 6, 143-149.
(20) Smith, S. B.; Bruice, T. C. J. Am. Chem. Soc. 1975, 97, 2875-
2881.
(13) Vogels, G. D.; De Windt, F. E.; Bassie, W. Recl. TraV. Chim. Pays-
Bas 1969, 88, 940-950.
(21) Luo, W.; Muller, J. G.; Burrows, C. J. Org. Lett. 2001, 2801-2804.
(22) Ye, Y.; Muller, J. G.; Luo, W.; Mayne, C. L.; Shallop, A. J.; Jones,
R. A.; Burrows, C. J. J. Am. Chem. Soc. 2003, 13926-13927.
(23) Niles, J. C.; Wishnok, J. S.; Tannenbaum, S. R. Chem. Res. Toxicol.
In press.
(24) Telo, J. P. Org. Biomol. Chem. 2003, 1, 588-592.
(25) Lazarus, R. A.; DeBrosse, C. W.; Benkovic, S. J. J. Am. Chem.
Soc. 1982, 104, 6871-6872.
(14) Leipold, M. D.; Muller, J. G.; Burrows, C. J.; David, S. S.
Biochemistry 2000, 39, 14984-14992.
(15) Van der Vliet, A.; Eiserich, J. P.; Halliwell, B.; Cross, C. E. J. Biol.
Chem. 1997, 272, 7617-7625.
(16) Wrona, M. Z.; Owens, J. L.; Dryhurst, G. J. Electroanal. Chem.
1979, 570, 371-387.
Org. Lett., Vol. 6, No. 19, 2004
3419

pH range 5-11. High pH favored the formation of Spd and
Sp base, suggesting involvement of deprotonation of 5-OH.
As the pH rises to >10, a significant portion of 5-OH-8-
oxoG is the anionic species (7). Enolization of the C6
carbonyl may generate intramolecular hydrogen bonding
between C6-O- and 5-OH, which may facilitate deproto-
nation of 5-OH since C6-O- can act as a proton acceptor
in proximity (Scheme 2). The pH dependence of formation
of the Sp and Gh bases supports this hypothesis. As the pH
rises, the anionic species of 5-OH-8-oxoG (7) becomes more
and more dominant along with a gradual increase in the yield
of Sp base and a decrease in the yield of Gh base. At pH >
11, N1 of 5-OH-UA may also be deprotonated to enolize
the C6 or C2 carbonyl (species 2 and 3 in Scheme 2). As in
the case of 5-OH-8-oxoG, enolization of C6 carbonyl would
slow pathway I and facilitate pathway II. Our data showed
that the yields of Alla at pH > 11 were significantly lower
than those at pH < 10.2 due to its decomposition. However,
a significant increase in the yield of Spd was not observed.
One possibility is that reaction pathways besides I and II
are in effect at such high pH and II is always a minor
pathway. Another possibility is that the enolization takes
place at the C2 carbonyl, thus facilitated deprotonation of
5-OH via intramolecular hydrogen bonding is not likely as
in the case of 5-OH-8-oxoG. It is important to note that
although our data are consistent with the above hypotheses,
more experimental or theoretical data are needed to fully
understand this intriguing chemistry.
Scheme 2. Ionization States of 5-OH-UA and 5-OH-8-oxoG
between 4a-OH-tetrahydropterins and 5-OH-8-oxoG26). Since
hydration of the C6 carbonyl is merely a hydrolysis of the
N1-C6 amide bond,5 the rate-limiting step is most likely
the collapse of the tetrahedral intermediates. For 5-OH-UA
at pH < 11, the collapse is facilitated by delocalization of
the negative charge onto the C2 oxygen (see Scheme 2 in
Supporting Information). For 5-OH-8-oxoG, at pH 7-9,
such facilitation is not available, although hydration of C6
carbonyl still takes place. At pH > 9, the anionic species of
5-OH-8-oxoG (6 and 7) predominates and enolization of C6
carbonyl slows the hydrolysis of N1-C6 amide bond due
to electrostatic repulsion between C6-O- and -OH.20
However, the N1-protonated species of 5-OH-8-oxoG (5)
predominates at pH < 7 and is susceptible to hydration since
the positively charged guanidine moiety is a good leaving
group. Thus, pathway I is favored at pH < 11 for 5-OH-
UA, while for 5-OH-8-oxoG it is favored at pH < 7 and
disfavored at pH > 9. This may partly explain why the yield
of Alla was above 70% at pH below 10.2 from the oxidation
of UA and the pH dependence of formation of Gh base in
Acknowledgment. This work was supported by NIH
Grant No. CA26731 and by the MIT Center for Environ-
mental Health Sciences with a grant from the NIEHS
(ES002109-26). We thank Agilent Technologies for access
to the 1100 MSD and ESI-TOF MS.
Supporting Information Available: Experimental pro-
cedures, mass spectra of products, LC-MS and HPLC
chromatograms, 13C NMR of 1-methyl-Spd and Spd, and
relevant schemes. This material is available free of charge
via the Internet at http://pubs.acs.org.
(26) Bailey, S. W.; Rebrin, I.; Boerth, S. R.; Ayling, J. E. J. Am. Chem.
Soc. 1995, 117, 10203-10211.
OL048547W
3420
Org. Lett., Vol. 6, No. 19, 2004