6-Deoxy-6-fluoro-l-ascorbic acid: Crystal structure and oxidative degradation

Article abstract of DOI:10.1016/S0008-6215(00)00183-X

Ascorbic acid and its oxidation products have been implicated in non-enzymatic modification of proteins in aging and diseases of oxidative stress. We have studied the feasibility of using 6-deoxy-6-fluoroascorbic acid (6) for identification of ascorbic acid degradation products by ¹⁹F NMR spectroscopy. Crystals of compound 6 from nitromethane belonged to the space group P2₁ with a = 5.547(2), b = 6.769(3), c = 9.302(2) A, β = 91.80(3)°and Z = 2. Atomic coordinates, bond lengths and angles, hydrogen coordinates, anisotropic and isotropic displacement parameters were similar if not identical with those of native ascorbic acid. Similarly, UV properties and oxidation kinetics by CuCl₂ at different pH values were essentially identical with ascorbic acid. Using 750 MHz ¹⁹F NMR spectroscopy, five to six new fluorinated products were detected after overnight oxidation of 6 with Cu²⁺, suggesting that 6 may be a powerful and sensitive tool for assessment of its catabolism in vivo. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0008-6215(00)00183-X

www.elsevier.nl/locate/carres
Carbohydrate Research 329 (2000) 477–485
Note
6-Deoxy-6-fluoro- -ascorbic acid: crystal structure and
L
oxidative degradation
Janusz Madaj a,1, Yoko Nishikawa b, V. Prakash Reddy c, Peter Rinaldi d,
Tadao Kurata b, Vincent M. Monnier a,*
a Institute of Pathology, Case Western Reser6e Uni6ersity, Cle6eland, OH, USA
b Institute of En6ironmental Science for Human Life, Ochanomizu Uni6ersity, Tokyo, Japan
c Department of Chemistry, Case Western Reser6e Uni6ersity, Cle6eland, OH, USA
d Department of Chemistry, Uni6ersity of Akron, Ohio 44325, Cle6eland, OH 44106, USA
Received 10 February 2000; received in revised form 25 May 2000; accepted 28 May 2000
Abstract
Ascorbic acid and its oxidation products have been implicated in non-enzymatic modification of proteins in aging
and diseases of oxidative stress. We have studied the feasibility of using 6-deoxy-6-fluoroascorbic acid (6) for
identification of ascorbic acid degradation products by 19F NMR spectroscopy. Crystals of compound 6 from
,
nitromethane belonged to the space group P21 with a=5.547(2), b=6.769(3), c=9.302(2) A, i=91.80(3)° and
Z=2. Atomic coordinates, bond lengths and angles, hydrogen coordinates, anisotropic and isotropic displacement
parameters were similar if not identical with those of native ascorbic acid. Similarly, UV properties and oxidation
kinetics by CuCl2 at different pH values were essentially identical with ascorbic acid. Using 750 MHz 19F NMR
spectroscopy, five to six new fluorinated products were detected after overnight oxidation of 6 with Cu2+, suggesting
that 6 may be a powerful and sensitive tool for assessment of its catabolism in vivo. © 2000 Elsevier Science Ltd.
All rights reserved.
Keywords: Ascorbic acid; Glycation; Maillard reaction; Aging; Diabetes
threo-hex-2,3-diulosonic acid (2,3-diketo-
1. Introduction
L
-gulonic acid). The biology governing the
tissue levels of these three compounds is fairly
well understood because both -threo-hex-2,3-
diulosono-1,4-lactone and -threo-hex-2,3-di-
ulosonic acid are unequivocally related to
ascorbic acid and can be assayed relatively
easily in biological specimens. However, in
selected tissues, it has been hypothesized that
Ascorbic acid (vitamin C) is one of the
major water-soluble antioxidants. Under
physiological conditions, most of the intracel-
lular ascorbic acid is in a reduced form and
L
L
less than 10% is present in form of
L-threo-
hex - 2,3 - diulosono - 1,4 - lactone (dehydro -
ascorbic acid) or its delactonized form,
L
-
-
L
fragmentation products of ascorbic acid,
threo-hex-2,3-diulosono-1,4-lactone and
L
-
-
L
* Corresponding author. Fax: +1-216-3680495.
threo-hex-2,3-diulosonic acid may participate
in Maillard reaction with proteins and form
advanced glycation products, which are in
E-mail address: vmm3@po.cwru.edu (V.M. Monnier).
1 Present address. Department of Chemistry, University of
Gdansk, Poland.
0008-6215/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0008-6215(00)00183-X

478
J. Madaj et al. / Carbohydrate Research 329 (2000) 477–485
detecting several oxidation products with 19F
NMR spectroscopy. It should be noted that a
synthesis for 6-deoxy-6-18F-ascorbic acid was
also proposed, which, however, did not lead
to the crystalline product [18].
part responsible for browning and crosslink-
ing of proteins in aging and in diseases of
accelerated aging, such as diabetes and end
stage renal disease [2,3].
Extensive studies have been carried out by
various groups on the nature of ascorbate
degradation products formed in vitro. A
dozen or so degradation compounds have
Chemical synthesis of 6-deoxy-6-fluoro- -
L
ascorbic acid (6).—The synthetic scheme is
outlined in Scheme 1.
been found, which include
threo-pentos-2-ulose ( -xylosone) [5], 3-deoxy-
-glycero-pentos-2-ulose (3-deoxy- -xylosone)
[6], -lyxonic and -xylonic acid [7–9],
threose [10], -threonic, glyceric acid and ox-
alic acid [8,11]. Recently, while this work was
in progress, -erythrulose and not -threose
L
-xylose [4],
L
-
In contrast to the method of Kiss et al.,
L
methyl 2,3-O-isopropylidene- -xylo-hex-2-
L
L
L
ulofuranosonate (2) was obtained in a one-pot
L
L
L
-
reaction from commercially available 2,3:4,6-
L
di-O-isopropylidene- -xylo-hex-2-ulofurano-
L
sonic acid (2,3:4,6-di-O-isopropylidene-2-keto-
-gulonic acid) (1). During heating of the
L
L
L
was found to be the major degradation
product of ascorbic acid under non-oxidative
conditions [12].
substrate in methanol in the presence of sulfu-
ric acid, the carboxyl group was converted
into methyl ester and the 4,6-O-isopropylidene
protecting group was selectively cleaved. The
crude by-product was, without any purifica-
tion, directly O-tosylated with p-toluenesul-
fonyl chloride in pyridine. Crystals of the
mono O-tosyl derivative (3) were obtained
with good yield. The next step, in which the
O-tosyl group is replaced with a fluoride ion,
is the bottleneck of this method. Because the
fluoride anion is a weak nucleophile, the re-
placement was carried out with KF in dry
N,N-dimethylformamide at 150 °C. In order
to obtain methyl 6-deoxy-6-fluoro-2,3-O-iso-
On the other hand, various advanced ascor-
bylation products have been characterized
both in vitro and in vivo and found to accu-
mulate in the aging eye lens, or to be present
in plasma proteins. Examples include car-
boxymethyl-lysine [13], pentosidine [14], oxalic
acid monoamide (HO2CCONH2), vesperlysine
A [15] to name just a few. In spite of these
accomplishments little is yet known of the in
vivo nature and catabolic pathways of ascor-
bate degradation products beyond 2,3-diketo-
L
-gulonic acid because many of the
degradation products do not substantially dif-
fer from reducing sugars. Similarly, most
ascorbylation products found so far in vivo,
such as carboxymethyl-lysine and pentosidine
are also advanced glycation products, and
even the recently described oxalic acid
monoamide is not entirely specific for ascor-
bate, as it can also form from reducing sugars
[16,17].
propylidene- -xylo-hex-2-ulofuranosonate (4)
L
we tested CsF and KF from many sources,
but satisfactory results were obtained only
with dry 99.99% KF packed under argon in
vials (Aldrich). In addition to the fluoro
derivative, methyl 4,6-anhydro-2,3-O-iso-
propylidene- -xylo-hex-2-ulofuranosonate (5)
L
was formed as the side-product of intramolec-
ular substitution during heating at relatively
high temperatures (Scheme 1). Both the weak
nucleophilicity of the fluoride anion and the
formation of this product caused the yield to
drop. In the last step of the synthesis, the
2,3-O-isopropylidene group was cleaved and
the product underwent isomerization to 6-
Searching for ways to label ascorbate at a
position that would allow detection of most of
its degradation products, we hypothesized that
ascorbic acid, when labeled with fluorine in
position 6, should be suitable for identification
of ascorbate catabolites in conjunction with
19F NMR spectroscopy. Here we describe the
synthesis for 6-deoxy-6-fluoroascorbic acid
(6), which was modified from the procedure of
Kiss et al. [1], its crystal structure and oxida-
tion kinetics in comparison with native ascor-
bic acid, and demonstrate the feasibility of
deoxy-6-fluoro- -ascorbic acid (6). For this
L
step we attempted to use ion-exchange resin
(H+ form) according to Kiss et al., but
better results were obtained by refluxing 4
in 40% sulfuric acid aqueous solution,
with acetonitrile as a co-solvent (1:1 ratio).

J. Madaj et al. / Carbohydrate Research 329 (2000) 477–485
479
The crystalline product was obtained from
nitromethane.
a fluorine atom in the 6-deoxy position might
introduce some slight difference in both chem-
ical characteristics and biological functions.
This assumption needs to be confirmed.
It should be noted that Kiss et al. deter-
mined the crystal structure of 6-deoxy-6-
chloroascorbic acid. Their R value was only
0.187 and could not be improved through
anisotropic refinement. Thus, the precise posi-
tion of the atoms in the crystal could not be
determined except for the chloride atom,
which could be assigned to the end of the side
chain. The question of whether 6 was present
as dienol-g-lactone or g-keto-enol acetal could
not be resolved.
UV spectroscopic and oxidant properties of
(6).—As expected based on the X-ray struc-
ture, the UV spectrum of 6 showed a maxi-
mum at 260 nm, which was indistinguishable
from that of ascorbic acid (Fig. 2). When
0.216 mM solutions were incubated with 10
mM CuCl2 in 0.1 mM sodium phosphate at
various pH values, initial oxidation rates for
Determination of the X-ray structure of 6-
deoxy-6-fluoro-ascorbic acid (6).—The crystal
and X-ray diffraction analyses data of 6 are
shown in Table 1. The molecular structure of
6-deoxy-6-fluoroascorbic acid is shown in Fig.
1. All bond lengths and angles are illustrated
in Table 2, which includes ascorbic acid data
from Hvoslef [19]. The crystal and molecular
structures of 6 are very close to those of
ascorbic acid. In particular, the planar struc-
ture on the lactone ring and the structure of
the enediol group on 6-deoxy-6-fluoro- -
L
ascorbic acid was almost equal to that of
ascorbic acid. This characteristic structure is
considered most important in the reducing
activity of ascorbic acid, thus 6 is also ex-
pected to have almost the same reducing activ-
ity. Although the planar structure in the
lactone ring and the enediol group of both
structures are similar, the replacement of the
OH group in the side chain of ascorbic acid by
Scheme 1.

480
J. Madaj et al. / Carbohydrate Research 329 (2000) 477–485
Fig. 1. X-ray structure of 6-deoxy-6-fluoro-
structure.
L-ascorbic acid. Below: the arrangement of 6-deoxy-6-fluoroascorbic acid in the crystal

J. Madaj et al. / Carbohydrate Research 329 (2000) 477–485
481
Table 1
ment is shown in Fig. 4 (top). A triplet split
into a doublet of triplets by the protons at C-6
was observed at −227.64 ppm relative to
Crystal data and structure refinement parameters for 6 a
Crystal data
Molecular formula
Molecular weight
Crystal system
Space group
Z
C6H7FO5
178.12
monoclinic
P21
Table 2
,
Bond lengths (A) and bond angles (°) for 6-deoxy-6-fluoro-
ascorbic acid and ascorbic acid (non-dissociated form19)
2
Cell dimensions
6-Deoxy-6-fluoro- Ascorbic acid
ascorbic acid
,
a (A)
5.547(2)
6.769(3)
9.302(2)
90.00
91.80(3)
90.00
non-dissociated form19
,
b (A)
,
c (A)
FꢀC-6
1.398
1.201
1.348
1.337
1.373
1.466
1.415
1.461
1.337
1.495
1.535
1.511
1.431 (O-6ꢀC-6)
1.216
1.361
1.326
1.355
1.444
1.427
1.452
1.338
h (°)
i (°)
k (°)
O-1ꢀC-1
O-2ꢀC-2
O-3ꢀC-3
O-4ꢀC-1
O-4ꢀC-4
O-5ꢀC-5
C-1ꢀC-2
C-2ꢀC-3
C-3ꢀC-4
C-4ꢀC-5
C-5ꢀC-6
3
,
V (A )
349.1(2)
1.695
Dcalcd (mg cm−3
)
Structure determination and
refinement data
Crystal dimensions (mm)
Temperature (K)
0.52×0.32×0.12
293
1.493
1.521
1.521
Radiation
Cu Ka (v=14.7 cm−1
,
,
u=1.5418 A)
689
675 (I\2|)
0BhB6, 0BkB8,
−11BlB11
67.47
No. of reflections measured
No. of unique reflections
Range of h, k, l
C-1ꢀO-4ꢀC-4
O-1ꢀC-1–C-2
O-4ꢀC-1–C-2
C-3–C-2ꢀO-2
C-3ꢀC-2ꢀC-1
O-2ꢀC-2ꢀC-1
O-3ꢀC-3ꢀC-2
O-3ꢀC-3ꢀC-4
C-2ꢀC-3ꢀC-4
O-4ꢀC-4ꢀC-3
O-4ꢀC-4ꢀC-5
C-3ꢀC4ꢀC-5
O-5ꢀC-5ꢀC-6
O-5ꢀC-5ꢀC-4
C-6ꢀC-5ꢀC-4
FꢀC-6ꢀC-5
108.6
128.8
109.0
129.5
108.5
122.2
132.0
118.4
109.9
103.9
106.8
115.1
108.0
110.3
111.8
109.2
109.1
129.1
109.5
127.5
107.8
124.6
133.5
117.1
109.5
104.0
110.4
114.8
106.9
111.7
112.7
108.0 (O-6ꢀC-6ꢀC5)
qmax (°)
Largest residual electron
density peaks (e A )
Final agreement factors
Function minimized
0.488 and −0.398
3
,
R=0.0519, Rw=0.2152
R=[w(F2o−F2c)2, using
least B square weights,
w=1/[|2(F2o)+0.1000P2],
where P=(F2o+2Fc2)/3,
109 parameters were
refined
a Intensity data were processed using the program TE
XSAN.
A few cycles of refinement were carried out using SHELXL-93.
both compounds were virtually identical (Fig.
3). Slower and more complex kinetics were
apparent at higher pH (pH 9) with 6, suggest-
ing that enolization of ascorbate hydroxyls in
positions 5 and 6 can influence the delac-
tonization rate. However, it is unlikely that
this phenomenon would be of importance for
the in vivo application of 6 since its initial
oxidation rate and half-life was not different
from that of ascorbic acid at physiological pH
values.
19F NMR properties of oxidation products of
6.—The fluorine NMR of pure 6 in deuterium
oxide recorded with a 750 MHz Varian instru-
Fig. 2. UV spectrum of ascorbic acid and 6-deoxy-6-fluoro-L-
ascorbic acid. Concentrations were arbitrarily chosen to fit on
the absorbance scale. For both compounds a 15.6 mM solu-
tion of the compound had 0.95 absorbance units at 264 nm.

482
J. Madaj et al. / Carbohydrate Research 329 (2000) 477–485
2. Experimental
General methods
NMR spectroscopy. H 400 MHz and 13C
100 MHz NMR spectra were determined on
a Varian Mercury 400 spectrometer with
chloroform-d as solvent and (CH3)4Si as the
internal standard. 19F NMR spectra were ob-
tained with a Varian UnityPlus-750 MHz
spectrometer equipped with a Nalorac 5 mm
1H/13C/19F PFG triple resonance probe (opti-
mized for 19F detection). The spectra were
obtained from samples in D2O solution with
the temperature regulated at 2590.1 °C. All
data processing was performed on a Sun
SPARCstation-10 using Varian’s Vnmr soft-
ware. The one-dimensional 19F spectra were
obtained with the following parameters: a 5.0
ms (ꢀ30°) 19F pulse width, 25,000 Hz spec-
tral window, 0.5 s acquisition time, 250 tran-
sients and no relaxation delay. The data were
weighted with 0.20 Hz exponential line
broadening and zero-filled before Fourier
transformation.
1
Fig. 3. Comparative oxidative degradation kinetics at various
pH values measured by absorbance at 260 nm of ascorbic
acid and 6-deoxy-6-fluoro- -ascorbic acid (each about 10 mM)
in presence of 10 mM CuCl2 in 0.1 M sodium phosphate
buffer at pH 6.0, 7.4 and 9.0. (Ascorbic acid: pH 6.0 --"--,
pH 7.4 --ꢀ--, pH 9.0 -- --; compound 6: pH 6.0 –2–, pH
7.4 –ꢁ–, pH 9.0 –ꢂ–).
L
CFCl3 as the internal standard. Further
splitting due to long-range coupling with
H-5 is also evident, with additional fine
splittings most likely from H-4. Following 30
min of oxidation at room temperature with
10 mM CuCl2, the ascorbic acid peak at
−227.64 ppm completely disappeared (not
shown). The signal shown in Fig. 4 (bottom)
for 6 was pasted into the Figure in order to
illustrate the position of unoxidized 6 relative
to the five new signals (U-1–U-5). Impor-
tantly, similar results were obtained when
fresh 6 was added to an already oxidized
mixture of ascorbic acid (not shown). The
structures of the compounds corresponding
to the new signals remain to be elucidated.
In summary, we have presented here a par-
tially modified synthesis of 6 and demon-
strated that its structure, UV properties, and
behavior toward pro-oxidant conditions do
not differ from those of native ascorbic acid.
Furthermore, we have shown that sufficiently
well resolved new signals are detectable in the
19F NMR spectrum recorded with a 750
MHz instrument, suggesting that 6-deoxy-
X-ray measurement.—Single crystals of 6-
deoxy-6-fluoroascorbic acid crystallized from
nitromethane having a size about 0.52×
0.32×0.12 mm were mounted on a glass fiber.
Intensity data for 6-deoxy-6-fluoroascorbic
acid were collected with graphite-monochro-
mated Cu Ka radiation (v=14.7 cm−1, u=
,
1.5418 A) on a Rigaku AFC-7R four-circle
diffractometer at 293 K. The initial positions
of 12 atoms without hydrogen atoms were
obtained by the direct method with the
SHELXS-86 [20] program for space group P21.
These atomic positions, together with an-
isotropic temperature factors, were refined by
the full-matrix least-squares procedure after
several refinement cycles with isotropic tem-
perature factors. Computations were per-
formed on an Iris Indy workstation with the
help of TEXSAN [21] and the SHELXL-93 [22]
package program. Further refinement of the
atomic parameters were carried out using the
package SHELXL-93.
Methyl 2,3-O-isopropylidene-6-O-tosyl-
xylo-hex-2-ulofuranosonate (3).—2,3:4,6-Di-
O-isopropylidene- -xylo-hex-2-ulofuranosonic
L
-
6-fluoro- -ascorbic acid will be a powerful
L
L
tool to monitor ascorbic acid catabolism in
vivo and identify its major degradation prod-
ucts.
acid (1) (10 g) was dissolved in MeOH (20
mL) and carbon tetrachloride (20 mL) and
five drops of concd H2SO4 were added. The

J. Madaj et al. / Carbohydrate Research 329 (2000) 477–485
483
After filtration, inorganic salt solvents were
mixture was refluxed for 30 min. After cooling
to room temperature (rt), CaCO3 (1.5 g) was
added and the mixture was stirred for 20 min.
evaporated off and methyl 2,3-O-isopropyli-
dene- -xylo-hex-2-ulofuranosonate (2) was
L
Fig. 4. 750 MHz 19F NMR spectra of 6 before (top) and after oxidation with 100 mM CuCl2 (bottom). In the bottom spectrum,
the signal corresponding to 6 at −227.64 ppm was absent in the original spectrum. It was pasted from the spectrum to illustrate
the position of 6 relative to its degradation products U1–U5.

484
J. Madaj et al. / Carbohydrate Research 329 (2000) 477–485
mg) was refluxed under nitrogen in a mixture
of 40% aq soln H2SO4 (25 mL) and MeCN (25
mL) for 1.5 h. After cooling to rt, MeCN was
evaporated off and the residue was extracted
with EtOAc (4×20 mL). The organic layers
were combined, washed with cold water (B10
mL), dried with Na2SO4 and concentrated by
evaporation to a solvent-free yellow oil (290
mg). The oil was dissolved in MeNO2 (3 mL)
and after conditioning overnight at −18 °C
obtained as a thick yellow oil (9.5 g). The oil
was dissolved in freshly distilled pyridine (135
mL), and TsCl (8.1 g) was added at 0 °C.
After 16 h at rt, pyridine was evaporated and
the residue was dissolved in 1 L of cold water.
After 1 h white crystals were removed by
filtration. The product, recrystallized from
EtOH and water, yielded methyl 2,3-O-iso-
propylidene-6-O-tosyl- -xylo-hex-2-ulofura-
L
1
nosonate (3) (7.5 g, yield 38%); H NMR
(CDCl3): 7.70–7.35 (m, 4 H, aromatic ring);
4.70 (s, 1 H, H-3); 4.48 (m, 1 H, J5,6 6.1 Hz,
H-5); 4.38 (dd, 1 H, J6,6% 10.4 Hz, H-6); 4.28
(d, 1 H, J4,5 2.2 Hz, H-4); 4.18 (dd, 1 H, H-6%);
3.80 (s, 3 H, OCH3); 2.45 (s, 3 H, PhCH3);
1.50–1.35 (2×s, 6 H, C(CH3)2); 13C NMR
(CDCl3): 145 (C1); 140–160 (aromatic ring);
115 (C(CH3)2); 110 (C-2); 87 (OCH3); 81 (C-
6); 74 (C-3); 67 (C-4); 54 (C-5); 27–26
(C(CH3)2); 22 (PhCH3).
white crystals of 6-deoxy-6-fluoro- -ascorbic
L
acid (6) were isolated (130 mg, yield 36%, mp
139–141 °C, [h]D +19.5° (0.5, water); 1H
NMR (D2O): 4.89 (d, 1 H, H-4); 4.65 (m, 1 H,
H-6); 4.56 (m, 1 H, H-6%); 4.25 (m, 1 H, H-5);
13C NMR (Me2SO): 170.39 (C-1), 152.06 (C-
3), 118.26 (C-2); 85.56–82.20 (d, JCF 168.7 Hz,
C-6); 74.65–74.51 (d, JCF 7.0 Hz, C-4); 66.51–
66.11 (d, JCF 20.2 Hz, C-5); 19F NMR (D2O,
CFCl3 as a reference peak): dt −227.64 ppm.
Oxidati6e degradation of 6 with CuCl2.—
Compound 6 (1.9 mg, 10.7 mmol) was dis-
solved in 1 mL of water (concentration 10.7
mM) and 20 mL of this solution was added to
1 mL of 10 mM aq CuCl2. The solution con-
taining 0.216 mM of 6 was kept at rt and the
spectrum was recorded 30 min later.
Methyl 6-deoxy-6-fluoro-2,3-O-isopropyli-
dene- -xylo-hex-2-ulofuranosonate (4).—To a
L
stirred suspension of KF (5 g, 99.99%, Aldrich
Co.) in 100 mL dry of N,N-dimethylfor-
mamide (distilled over P2O5) was added
methyl 2,3 - O - isopropylidene - 6 - O - tosyl - -
L
xylo-hex-2-ulofuranosonate (3) (5 g) and the
mixture was stirred and heated for 5 h at
150 °C (glycerol bath). After cooling to rt,
inorganic salts were filtered off and solvents
were evaporated (bath temperature 70 °C). To
the residue, EtOAc (100 mL) was added and
undissolved salts were filtered off. The mixture
was concentrated to a yellow oil, which was
purified twice by column chromatography
(Silica Gel, E. Merck 70–230 mesh, 3:7 ace-
tone–hexane) yielding pure crystals of methyl
Acknowledgements
This work was supported by grants from
the National Eye Institute (EY 07099 and
P30-EY11373) and in part by the Polish Re-
search Council (KBN) (grant BW/8000-8-
0307-9). We thank Mr Dale Ray (Dept. of
Chemistry, University of Akron, Akron, OH)
for assistance with NMR spectroscopy.
6-deoxy-6-fluoro-2,3-O-isopropylidene- -xylo-
L
hex-2-ulofuranosonate (4) (650 mg, yield
1
21%); H NMR (CDCl3): 4.82 (m, 1 H, H-6);
4.70 (d, 1 H, H-3); 4.59 (m, 2 H, H-5 and
H-6%); 4.30 (d, 1 H, H-4); 3.87 (s, 3 H, OCH3);
1.54–1.38 (2×s, 6 H, C(CH3)2); 13C NMR
(CDCl3): 168.89 (C1); 114.83 (C(CH3)2);
109.82 (C-2); 87.53 (C-3); 81.58–82.02 (d, JCF
22.1 Hz, C-5); 79.57–82.90 (d, JCF 167.4 Hz,
C-6); 74.85–74.97 (d, JCF 5.9 Hz, C-4); 53.66
(OCH3); 26.92–25.76 (C(CH3)2).
References
[1] J. Kiss, K.P. Berg, A. Dirscherl, W.E. Oberha¨nsli, W.
Arnold, Hel6. Chim. Acta, 63 (1980) 1728–1739.
[2] R. Ikan, The Maillard Reaction: Consequences for the
Chemical and Life Sciences, Wiley, New York, 1996, pp.
1–214.
[3] T. Miyata, C. van Ypersele de Strihou, K. Kurokawa,
J.W. Baynes, Kidney Int., 55 (1999) 389–399.
[4] P.C. Chan, R.R. Becker, C.G. King, J. Biol. Chem., 231
(1957) 231–240.
6-Deoxy-6-fluoro-
Methyl 6-deoxy-6-fluoro-2,3-O-isopropyli-
dene- -xylo-hex-2-ulofuranosonate (4) (500
L
-ascorbic acid (6).—
[5] S.O. Kang, H. Sapper, W. Lohman, Z. Naturforsch., Teil
C, 73 (1982) 1063–1069.
L

J. Madaj et al. / Carbohydrate Research 329 (2000) 477–485
485
[6] D.B. Shin, M.S. Feather, Carbohydr. Res., 208 (1990)
246–250.
[7] J. Kafner, G. Ashwell, J.J. Burns, J. Biol. Chem., 231
(1960) 2518–2521.
[8] Y. Kagawa, H. Takiguchi, J. Biochem. (Tokyo), 51 (1962)
134–144.
[9] Y. Kagawa, H. Takiguchi, N. Shimazono, Biochem. Bio-
phys. Acta, 51 (1961) 413–415.
[10] M.G. Lopez, M.S. Feather, J. Carbohydr. Chem., 11
(1992) 799–806.
[11] T. Miyake, T. Shibamoto, J. Agric. Food Chem., 43
(1995) 1669–1672.
[12] G.L. Simpson, B.J. Ortwerth, Biochim. Biophys. Acta,
1501 (2000) 2–24.
[13] J.A. Dunn, M.U. Ahmed, M.H. Murtiashaw, J.M.
Richardson, M.D. Walla, S.R. Thorpe, J.W. Baynes,
Biochemistry, 29 (1990) 10964–10970.
[15] R.H. Nagaraj, D.R. Sell, M. Prabhakaram, B.J.
Ortwerth, V.M. Monnier, Proc. Natl. Acad. Sci. USA, 88
(1991) 10257–10261.
[16] M. Pischetsrieder, B. Larisch, W. Seidel, J. Agric. Food
Chem., 45 (1997) 2070–2075.
[17] R.H. Nagaraj, F.A. Shamsi, B. Huber, M. Pischetsrieder,
FEBS Lett., 453 (1999) 327–330.
[18] F. Yamamoto, S. Sasaki, M. Maeda, Appl. Radiat. Isot.,
43 (1992) 633–639.
[19] J. Hvoslef, Acta Crystallogr., Sect. B, 24 (1968)
23.
[20] G.M. Sheldrick, Program for Crystal Structure Determi-
nation, University of Go¨ttingen, Germany, 1986.
[21] TEXSAN, Single Crystal Structure Analysis Software, ver-
sion 1.6, Molecular Structure Corporation, The Eood-
lands, TX 77381, 1993.
[22] G.M. Sheldrick, Program for Crystal Structure Refine-
ment, University of Go¨ttingen, Germany, 1993.
[14] F. Tessier, M. Obrenovich, V.M. Monnier, J. Biol.
Chem., 274 (1999) 20796–20804.
.