Synthesis of highly pure 14C-labelled DL-allantoin and 13C NMR analysis of labelling integrity

Article abstract of DOI:10.1002/jlcr.1614

A number of synthetic approaches are assessed to prepare allantoin labelled with ¹⁴C given certain requirements and technical limitations. A method that fulfils these criteria is described to achieve the synthesis of highly pure ¹⁴C-labelled allantoin with the label introduced to the ureido carbonyl group in the final step by reaction of 5-chlorohydantoin with [¹⁴C]urea. The chosen method favours high purity at the expense of radiochemical yield, which is achieved at a level of 8%. The integrity of the label is then investigated by performing an NMR analysis of ¹³C- labelled allantoin synthesized by the same method. The ¹³C NMR spectrum confirms partial scrambling of the label to the C-2 position by equilibration of the product via a putative bicyclic intermediate, which had been suggested by other workers. The ¹⁴C-labelled allantoin synthesized by this method is therefore assigned as DL-[H₂N ¹⁴CO/¹⁴C-2]allantoin. This study also includes the first full characterization of a side product, 5-hydroxy-5-methoxyhydantoin, obtained by the reaction of a 5-hydroxyhydantoin intermediate with the methanol solvent. Copyright

Full text of DOI:10.1002/jlcr.1614

Note
Received 5 December 2008,
Revised 15 April 2009,
Accepted 1 May 2009
Published online 26 June 2009 in Wiley Interscience
(www.interscience.wiley.com) DOI: 10.1002/jlcr.1614
14
Synthesis of highly pure C-labelled
13
DL-allantoin and C NMR analysis of
labelling integrity
Ã
Simon G. Patching
A number of synthetic approaches are assessed to prepare allantoin labelled with 14C given certain requirements and
technical limitations. A method that fulfils these criteria is described to achieve the synthesis of highly pure 14C-labelled
allantoin with the label introduced to the ureido carbonyl group in the final step by reaction of 5-chlorohydantoin with
[
14C]urea. The chosen method favours high purity at the expense of radiochemical yield, which is achieved at a level of 8%.
The integrity of the label is then investigated by performing an NMR analysis of 13C-labelled allantoin synthesized by the
same method. The 13C NMR spectrum confirms partial scrambling of the label to the C-2 position by equilibration of the
product via a putative bicyclic intermediate, which had been suggested by other workers. The 14C-labelled allantoin
synthesized by this method is therefore assigned as DL-[H2N14CO/14C-2]allantoin. This study also includes the first full
characterization of a side product, 5-hydroxy-5-methoxyhydantoin, obtained by the reaction of a 5-hydroxyhydantoin
intermediate with the methanol solvent.
Keywords: 5-substituted hydantoins; 14C-labelling; allantoin; 13C NMR analysis
(pre-treated with 10% H2SO4)7 on a 100 mg scale. In both cases
allantoin could not be obtained as a crystalline solid, without
Introduction
DL-5-monosubstituted hydantoins are important precursors to
optically pure D- or L-amino acids, which are intermediates in the
production of a number of drugs and pharmaceuticals via their
enzyme-catalysed hydrolyses.1,2 As part of our studies on the
structure–function relationships of hydantoin transporters in
pathogenic bacteria,3–6 we required DL-allantoin (DL-5-ureido-
hydantoin) labelled with 14C, which is not commercially
available. Here we describe our approach to the synthesis of
14C-labelled DL-allantoin 5, where the target site for the 14C label
was the ureido carbonyl group. A number of important
considerations in our choice of synthetic route were the costs
of starting compounds and 14C-labelled reagent, avoidance of
overcomplicated or technically demanding reactions (due to
limited facilities for radiolabelled work), introduction of the 14C
label at the final step (to minimize handling and to optimize
radiolabelled yield), production of the final compound as a solid
with 499% purity from a relatively small-scale (ꢀ100 mg)
reaction. Avoiding chromatography was also desirable to
minimize further handling. We investigated a number of routes
for the synthesis of 5 under these limitations.
purification by chromatography, and the product contained
some unreacted urea (1H NMR 5.41 ppm, 13C NMR 160.0 ppm);
therefore, this route was considered not suitable for the
synthesis of 5.
We also investigated the reaction of DL-5-aminohydantoin
with potassium cyanate9 for the synthesis of 5, which would use
[14C]KCNO as the source of radiolabel. For this, DL-5-aminohy-
dantoin was prepared as described by Sarges et al.10 with the
final removal of the benzyloxycarbonyl protecting group using
HBr-AcOH cleavage. However, on a small scale, the reaction with
potassium cyanate did not produce allantoin as a pure solid.
We then investigated a route to 5, which was devised
independently in this study, but has been described earlier by
Abblard and Meynaud11 and is shown in Scheme 1, where the
radiolabel is introduced in the last step by reaction of DL-5-
chlorohydantoin 4 with [14C]urea.
To produce 4, we began with the commercially available
parabanic acid 1, which was reduced with KBH4 in a similar
manner to that described by Abblard and Meynaud11 to give DL-
5-hydroxyhydantoin 2 in 77% yield. In this reaction, some of the
starting parabanic acid (molar equivalent to KBH4) was lost as
the potassium salt, which was removed as a solid. Some of the
Results and discussion
An attractive one-step synthesis of DL-allantoin is the reaction of
glyoxylic acid (OHC-COOH) with urea in the presence of an acid,
for which the use of heterogeneous catalysts has been studied,7
especially as the relatively cheap and available [14C]urea could
be used as the source of the radiolabel. We investigated this
reaction catalysed by both HCl8 and by Dowex 50W Â 8 resin
Astbury Centre for Structural Molecular Biology and Institute of Membrane and
Systems Biology, University of Leeds, Leeds LS2 9JT, UK
*Correspondence to: Simon G. Patching, Astbury Centre for Structural Molecular
Biology and Institute of Membrane and Systems Biology, University of Leeds,
Leeds LS2 9JT, UK.
E-mail: s.g.patching@leeds.ac.uk
J. Label Compd. Radiopharm 2009, 52 401–404
Copyright r 2009 John Wiley & Sons, Ltd.

S. G. Patching
O
OH
H
N
H
N
H
N
O
O
OH
O
5
1
H
N
H
NH2
NH2
OMe
H
N
H
N
KBH4 / MeOH
*
NH
OH
NH
HN
HN
N
O
O
C
C
4
O
+
2
*
O
O
N
H
N
H 3
N
H
O
O
O
2 (77%)
3
(8%)
1
N
H
N
H
O
O
*
O
SOCl2
H
N
NH2
H
N
H
N
Figure 1. Putative mechanism for the scrambling of a label at the ureido carbonyl
group of allantoin to the C-2 position. This type of mechanism had been
suggested earlier by Abblard and Meynaud.11
Cl
O
*
C
*
CO(NH2)2 / MeNO2
O
O
O
N
H
N
H
O
5
(13%)
4
(69%)
equivalents of urea, which will affect the theoretical maximum
yield possible. The relatively low yield was tolerated in order to
obtain highly pure DL-allantoin, important when performing
biological assays with radiolabelled compounds, and to fulfil all
of the limitations described above. The reaction was then
performed in the presence of [14C]urea (0.5% at 56 mCi/mmol) to
produce 34.4 mg (0.22 mmol) of 14C-labelled DL-allantoin 5, with
=
14C
*
H
N
NH2
C
H
N
*
O
O
N
H
O
a
14C specific activity of 166 mCi/mmol. This corresponds to a
Scheme 1. Synthesis of 14C-labelled DL-allantoin.
chemical yield of 13% from urea and a radiochemical yield of 8%.
The work performed by Abblard and Meynaud11 suggested
that a 14C label incorporated at the ureido carbonyl group of
allantoin undergoes scrambling to C-2, possibly by the attack of
the ureido primary amino group on the C-4 carbonyl function
required product was lost by reaction of 2 with the methanol
solvent to give 5-hydroxy-5-methoxyhydantoin 3, which was
also removed as a solid, in 8% yield. The loss of material to these
side products had already been suggested by Abblard and
Meynaud,11 but isolation of 3 was surprising; therefore, a full
NMR and physical analysis of this compound was performed,
which had not been reported previously. Compound 3 was
isolated as highly pure colourless crystals with melting point
222–2251C and with elemental analysis and high-resolution
mass spectrometry results entirely consistent with the
5-methoxy structure shown in Scheme 1. The NMR spectra for
this compound obtained in DMSO-d6 were also consistent with
and therefore via
a bicyclic symmetrical hydroxyglycouril
intermediate (Figure 1). To investigate potential scrambling
through this mechanism the synthesis was performed on the
same scale as for the radiolabelled synthesis, except using 100%
[13C]urea to give 13C-labelled DL-allantoin 5 in 17% yield. The 13
C
NMR spectrum of 13C-labelled 5 (Figure 2(A)) confirmed that a
label at the ureido carbonyl group does undergo partial
scrambling to C-2. Both the 13C and 1H NMR spectra of the
13C-labelled allantoin (Figure 2(A) and (B)) also confirm the high
purity of DL-allantoin produced by this method, especially with
regard to any residual urea, which would have been easily
detected (at 160.0 ppm) in the 13C spectrum, with it being 13C-
the given structure, except that C-5 was not observed in the 13
C
spectrum; this may be due to the absence of directly attached
protons at C-5. Observation of NMR spectra also demonstrated
that compound 3 was not stable for long in the DMSO solution.
Compound 3 was not sufficiently soluble to allow 13C NMR
analysis in alternative solvents, including methanol-d4; this was
not surprising as methanol is the solvent from which compound
3 cleanly crystallizes from the reaction mixture.
Substitution of 2 to produce DL-5-chlorohydantoin 4 was
achieved by reflux in thionyl chloride. As described by Sarges
et al. in their preparation of DL-5-[(benzyloxy-carbonyl)amino]-
hydantoin,10 4 was not isolated as it is unstable; instead the
reaction mixture was evaporated and the residue dissolved in
nitromethane for direct use in the next reaction. Importantly,
any unreacted 2 is insoluble in nitromethane and thus can be
removed by filtration, and the recovered 2 provided an
approximate and indirect measure for the yield of 4. This
approach avoids isolating the unstable compound 4 as a solid,
which had to be kept covered with hexane and used
contaminated with ꢀ10% 5-hydroxyhydantoin 2 in the method
described by Abblard and Meynaud.11
DL-allantoin 5 was produced by reflux of 4 with urea in
nitromethane in a similar manner to that described by Abblard
and Meynaud,11 except that a solution of 4 was used directly
from the previous reaction. The product was obtained as a pure
solid, with no contaminating urea, by its precipitation from the
reaction mixture and washing with water. The yield of DL-
allantoin 5 using this method was 15% from urea. The yield of 4
is only approximated; therefore, the quantity of 4 in the reaction
mixture is likely to be significantly less than the molar
1
labelled, and at 5.41 ppm in the H spectrum.
The 14C-labelled allantoin 5 produced using the method
chosen here therefore contains molecules with 14C either at the
ureido carbonyl group or at C-2 and so can be assigned as
DL-[H2N14CO/14C-2]allantoin. Having the label at a mixture of the
two positions does not cause any problems with its use in our
intended biological assays. This study has highlighted a simple
and reliable method to synthesize 14C-labelled DL-allantoin
under certain experimental limitations where high purity and
consistency was favoured over yield.
Experimental
General
Except where stated otherwise, chemicals, ion-exchange resins and
NMR solvents were from Sigma-Aldrich and reaction solvents used
were of analytical grade, and all used without further purification.
Column chromatography was performed on silica gel 60 using
general-purpose solvents. Melting points were measured using a
STUART Melting Point Apparatus. NMR spectra were recorded on a
Bruker Avance 300 spectrometer and chemical shifts (d) are given in
ppm relative to the internal standard TMS. High-resolution mass
spectra were obtained using a Waters GCT Premier instrument.
Liquid scintillation analysis was performed on a Packard Tri-Carb
2100TR instrument and using an emulsifier-safe scintillation cocktail
from Perkin-Elmer.
www.jlcr.org
Copyright r 2009 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2009, 52 401–404

S. G. Patching
156.9 (C-4), 77.2 (C-5). HRMS (EI1/TOF): m/z calcd. for
C3H4N2O3 = 116.0222; found: 116.0217.
160.0 ppm
(A)
H2NCO
C-2
157.7 ppm
157.1 ppm
DL-5-chlorohydantoin (4)10
5-Hydroxyhydantoin 2 (500 mg, 4.31 mmol) was refluxed with
vigorous stirring in thionyl chloride (8.7 ml) for 2 h. After cooling,
the mixture was evaporated and the residue dissolved in
nitromethane (10 ml) (5-chlorohydantoin dissolves to give a
yellow solution and unreacted 5-hydroxyhydantoin remains
insoluble). The solution was filtered to remove the contaminat-
ing 5-hydroxyhydantoin and 5 ml of the filtrate was used in the
next reaction. The yield of 4 was ꢀ401 mg, 2.98 mmol, 69%
based on the amount of recovered 5-hydroxyhydantoin.
Urea
C-4
173.9 ppm
DMSO-d6
C-5
62.8 ppm
Expansion
13C chemical shift (ppm)
DL-allantoin (5)7,11
A solution of 4 in nitromethane (5 ml containing ꢀ200 mg,
1.49 mmol, prepared as above) was added in portions over
ꢀ30 min to a solution of urea (100 mg, 1.66 mmol) in boiling
nitromethane (5 ml) and the mixture was refluxed for 1 h with
vigorous stirring. After allowing to cool, water (3 ml) was added and
the mixture was stood at 41C for ꢀ5 days. A pale precipitate
formed at the interface with the aqueous layer, which was collected
under vacuum, washed thoroughly with water and dried under
vacuum over phosphorus pentoxide for ꢀ2 days to give 5
(39.3 mg, 0.25 mmol, 15% from urea) as a pale buff powder. 1H NMR
(300 MHz, DMSO-d6): d= 10.53 (s, 1 H, N3-H), 8.05 (s, 1-H, N1-H), 6.88
(d, 8.4 Hz, 1 H, HNCO), 5.78 (s, 2 H, NH2), 5.24 (dd, 1 H, J=1.1 and
8.4 Hz, C5-H). 13C NMR (75 MHz, DMSO-d6): d= 173.2 (C-4), 156.9
(H2NCO), 156.3 (C-2), 61.99 (C-5). HRMS (EI1/TOF): m/z calcd. for
C4H6N4O3 = 158.0440; found: 158.0444. Anal. calcd. for C4H6N4O3: C,
30.39; H, 3.82; N, 35.43. Found: C, 30.55; H, 3.7; N, 35.15.
(B)
DMSO-d6
NH2
5.77 ppm
HNCO
6.88 ppm
N3-H
10.52 ppm
N1-H
8.04 ppm
C5-H
5.24 ppm
1H chemical shift (ppm)
DL-[H2N14CO/14C-2]allantoin (5)
Figure 2. NMR analysis of 13C-labelled DL-allantoin. 13C (A) and 1H (B) NMR spectra
of the synthesized 13C-labelled DL-allantoin 5 obtained in DMSO-d6 and using a
300 MHz magnet. The 13C NMR signal for urea, obtained for a separate sample of
unlabelled urea in DMSO-d6 and using the same magnet, is also shown in (A).
The procedure was as above for the unlabelled 5, except that
0.5 mg of the unlabelled urea was replaced by 0.5 mg [14C]urea
(American Radiolabelled Chemicals) with 14C specific activity
56 mCi/mmol or 0.93 mCi/mg. The chemical yield of the reaction
was 34.4 mg (0.22 mmol, 13% from urea). A total of 1.0 mg of the
solid 5 gave 14C counts (Packard Tri-Carb 2100TR) with an
average of 2 360 945 dpm = 39.35 kBq = 1.0635 mCi; therefore,
34.4 mg of the solid has 36.58 mCi and a 14C specific activity of
166 mCi/mmol or 0.166 mCi/mmol. The [14C]urea (0.5 mg) used in
the reaction had 465 mCi; the radiochemical yield of the reaction
was therefore 36.58/465 mCi  100 = 8%.
DL-5-hydroxyhydantoin (2)11
Potassium borohydride (273 mg, 5.1 mmol) was added in
portions over ꢀ30 min to a stirred solution of parabanic acid
1 (2 g, 17.5 mmol) in anhydrous methanol (44 ml) and the
mixture was stirred for a further 1 h at room temperature.
During this time, the potassium salt of 1 formed as a colourless
solid, which was removed by filtration. The filtrate was stood at
room temperature for ꢀ24 h, during which 5-hydroxy-5-
methoxyhydantoin 3 formed as colourless crystals (208 mg,
DL-[H2N13CO/13C-2]allantoin (5)
1
1.42 mmol, 8%). [M.p. 222–2251C. H NMR (300 MHz, DMSO-d6):
The procedure was as above for the unlabelled 5, except that
101.3 mg (1.66 mmol) of [13C]urea (Cambridge Isotope Laboratories)
was used in the reaction. The yield of 5 was 44.9 mg (0.28 mmol,
17% from urea). 1H NMR (300 MHz, DMSO-d6): d= 10.52 (s, 1 H, N3-H),
8.04 (s, 1-H, N1-H), 6.88 (d, 8.4Hz, 1 H, HNCO), 5.77 (s, 2 H, NH2), 5.24
(dd, 1 H, J= 1.1 and 8.4 Hz, C5-H). 13C NMR (75 MHz, DMSO-d6):
d= 173.9 (C-4), 157.7 (H2NCO), 157.1 (C-2), 62.8 (C-5). HRMS (ESI1):
m/z calcd. for C133CH6N4O31H = 160.0546; found: 160.0541.
d = 10.46 (s, 1 H, N3-H), 7.35 (s, 1 H, N1-H), 7.10 (s, 1 H, C5-OH),
3.77 (s, 3 H, C5-OCH3). 13C NMR (75 MHz, DMSO-d6): d = 160.2
(C-2), 152.7 (C-4), no C-5 detected, 53.3 OCH3. HRMS (EI1/TOF):
m/z calcd. for C4H6N2O4 = 146.0328; found: 146.0323. Anal. calcd.
for C4H6N2O4: C, 32.88; H, 4.14; N, 19.17. Found: C, 32.95; H, 4.1;
N, 19.4. Unstable to storage in DMSO]; these were removed by
filtration. The filtrate was evaporated to leave a colourless solid,
which was washed with methanol and dried under vacuum to
give 2 (1.1 g, 9.48 mmol, 77%) as a colourless solid. The yield is
corrected for the loss of 1 to the potassium salt (5.1 mmol). M.p.
Acknowledgement
1
1411C (lit. 140–1421C11). H NMR (300 MHz, DMSO-d6): d = 10.58
This work was funded by grants from the BBSRC [BB/C51725],
EPSRC [EP/C00664X/1] and the EV European Membrane Protein
(s, 1 H, N3-H), 8.31 (s, 1 H, N1-H), 6.67 (s, 1 H, C5-OH), 5.08 (d, 1 H,
J = 3.0 Hz, C5-H). 13C NMR (75 MHz, DMSO-d6): d = 174.6 (C-2),
J. Label Compd. Radiopharm 2009
Copyright r 2009 John Wiley & Sons, Ltd.
www.jlcr.org

S. G. Patching
consortium (E-MeP, contract LSHG-CT-2004-504601). S. G. P.
thanks Peter Henderson and Steve Homans (University of Leeds),
Malcolm Levitt (University of Southampton) and Shun’ichi
Suzuki (Ajinomoto Co., Inc.) for their support, and Pikyee Ma
for the unpublished information. Mass spectrometry and
elemental analyses were performed by the School of Chemistry,
University of Leeds.
J. K. Griffith, M. J. Stark, A. Ward, J. O’Reilly, N. G. Rutherford, M. K.
Phillips-Jones, P. J. F. Henderson, FEBS Lett. 2003, 555, 170–175.
[4] M. Saidijam, K. E. Bettaney, G. Szakonyi, G. Psakis, K. Shibayama,
S. Suzuki, J. Clough, V. Blessie, A. Abu-Bakr, S. Baumberg, J. Mueller,
C. K. Hoyle, S. L. Palmer, P. Butaye, K. Walravens, S. G. Patching,
J. O’Reilly, N. G. Rutherford, R. M. Bill, D. I. Roper, M. K. Phillips-
Jones, P. J. F. Henderson, Biochem. Soc. Trans. 2005, 33, 867–872.
[5] S. Suzuki, P. J. F. Henderson, J. Bacteriol. 2006, 188, 3329–3336.
[6] S. Weyand, T. Shimamura, S. Suzuki, O. Mirza, K. Krusong,
E. P. Carpenter, N. G. Rutherford, J. M. Hadden, J. O’Reilly,
P. Ma, M. Saidijam, S. G. Patching, R. J. Hope, H. T. Nobertczak,
P. C. J. Roach, S. Iwata, P. J. F. Henderson, A. D. Cameron, Science.
2008, 322, 709–713.
References
[7] C. Cativiela, J. M. Fraile, J. I. Garcia, B. Lazaro, J. A. Mayoral,
A. Pallares, Green Chem. 2003, 5, 275–277.
[8] Z. Cai, J. Wang, L. Qiu, M. Wang, Huaxue Shijie. 2004, 45, 42.
[9] K. M. Youssef, E. Al-Abdullah, H. El-Khamees, Med. Chem. Res. 2002,
11, 481–503.
[1] J. Altenbuchner, M. Siemann-Herzberg, C. Syldatk, Curr. Opin.
Biotechnol. 2001, 12, 559–563 (and references therein).
[2] A. S. Bommarius, M. Schwarm, K. Drauz, J. Mol. Catal. B Enzym.
1998, 5, 1–11.
[3] M. Saidijam, G. Psakis, J. L. Clough, J. Mueller, S. Suzuki, C. J. Hoyle, [10] R. Sarges, R. C. Schnur, J. L. Belletire, M. J. Peterson, J. Med. Chem.
S. L. Palmer, S. M. Morrison, M. K. Pos, R. C. Essenberg, 1988, 31, 230–243.
M. C. J. Maiden, A. Abu-Bakr, S. G. Baumberg, A. A. Neyfakh, [11] J. Abblard, A. Meynaud, Bull. Soc. Chim. Fr. 1971, 942–946.
www.jlcr.org
Copyright r 2009 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2009, 52 401–404