Novel human lens metabolites from normal and cataractous human lenses

Article abstract of DOI:10.1016/j.tet.2007.03.133

4-(2-Aminophenyl)-4-oxobutanoic acid, 4-(2-amino-3-hydroxyphenyl)-4-oxobutanoic acid and glutathionyl-kynurenine have been identified as novel metabolites in normal and cataractous human lenses following total synthesis and comparison with authentic human

Full text of DOI:10.1016/j.tet.2007.03.133

Tetrahedron 63 (2007) 4990–4999
Novel human lens metabolites from normal and cataractous
human lenses
Jasminka Mizdrak,a,y Peter G. Hains,b,y Danuta Kalinowski,a Roger J. W. Truscott,b
Michael J. Daviesc,d and Joanne F. Jamiea,
*
aDepartment of Chemistry and Biomolecular Sciences, Macquarie University, Sydney, NSW 2109, Australia
bThe Save Sight Institute, Sydney, NSW 2001, Australia
cThe Heart Research Institute, Sydney, NSW 2050, Australia
dFaculty of Medicine, University of Sydney, Sydney, NSW 2006, Australia
Received 14 November 2006; revised 7 March 2007; accepted 22 March 2007
Available online 27 March 2007
Abstract—4-(2-Aminophenyl)-4-oxobutanoic acid, 4-(2-amino-3-hydroxyphenyl)-4-oxobutanoic acid and glutathionyl-kynurenine have
been identified as novel metabolites in normal and cataractous human lenses following total synthesis and comparison with authentic human
lens samples. Their structures are consistent with those derived from the major human lens UV filters kynurenine and 3-hydroxykynurenine,
and it is proposed that these compounds also play a role as UV filters. These metabolites were isolated in pmol/mg levels (dry mass) in lenses.
4-(2-Amino-3-hydroxyphenyl)-4-oxobutanoic acid and glutathionyl-kynurenine were found to be unstable at physiological pH. Other poten-
tial metabolites, glutathionyl-3-hydroxykynurenine, kynurenine yellow and 3-hydroxykynurenine yellow, were not detected in either normal
or cataractous lenses.
Ó 2007 Elsevier Ltd. All rights reserved.
1. Introduction
acid O-b-D-glucoside (2), kynurenine (3) and 3-hydroxy-
kynurenine (4) (Fig. 1).5
The primate lens contains fluorescent tryptophan-derived
compounds known as UV filters. These compounds absorb
UV light (295–400 nm) and are believed to protect the
lens and retina from UV-induced photo-damage.1–4 Major
UV filters detected to date in primate lenses, in decreasing
order of abundance, include 3-hydroxykynurenine O-b-D-
glucoside (1), 4-(2-amino-3-hydroxyphenyl)-4-oxobutanoic
Studies in our laboratory have shown that 1, 3 and 4 undergo
non-enzymatic deamination at physiological pH to form a,b-
unsaturated carbonyl compounds (Scheme 1).6–8 We have
found that these unsaturated molecules undergo covalent
binding (Michael addition) in vivo with amine and thiol
nucleophiles including lysine, cysteine and histidine residues
COOH
O
S
H
O
GSH
COOH
NH2
O
COOH
H
R2
COOH
NH2
NH2
NH2
OGlu
OGlu
5
R1
6
1: R1=OGlu, R2=NH2
2: R1=OGlu, R2=H
3: R1=H, R2=NH2
4: R1=OH, R2=NH2
O
COOH
H
NH2
OH
H
N
O
OOC
GSH =
H
N
H
OGlu =
HO
HO
O
CH2
S
O
OH
Figure 1. Major human lens UV filter compounds.
* Corresponding author. Tel.: +612 98508283; fax: +612 98508313; e-mail: joanne.jamie@mq.edu.au
J.M. and P.G.H. contributed equally to this manuscript.
y
0040–4020/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2007.03.133

J. Mizdrak et al. / Tetrahedron 63 (2007) 4990–4999
4991
O
COOH
8: R=H
9: R=OH
NH2
R
O
GSH
COOH
NAD(P)H
O
GSH
COOH
O
7: R=H
12: R=OH
COOH
H
NH2
NH2
O
NH3
R
R
NH2
NH2
3: R=H
4: R=OH
10: R=H
11: R=OH
R
R
N
H
COOH
α,β-unsaturated
intermediate
Scheme 1. Schematic diagram illustrating the degradation pathways of the human lens UV filter compounds 3 and 4.
on human lens proteins,9–11,7 and have also identified
glutathionyl-3-hydroxykynurenine O-b-D-glucoside (5)12
and cysteine-3-hydroxykynurenine O-b-D-glucoside (6)13
as novel human lens UV filter compounds (Fig. 1). In addi-
tion, deaminated 3 readily reacts in vitro with glutathione
(GSH) to give glutathionyl-kynurenine (7),10 while deami-
nated 1, 3 and 4 can be reduced by NAD(P)H to give 2,
4-(2-aminophenyl)-4-oxobutanoic acid (8) and 4-(2-amino-
3-hydroxyphenyl)-4-oxobutanoic acid (9), respectively.8
The deamination and subsequent reduction of 1 is the likely
metabolic process for formation of 2 in the lens.6 In the
absence of NAD(P)H, deaminated 1, 3 and 4 undergo slow
intramolecular Michael addition in vitro to give 3-hydroxy-
kynurenine O-b-D-glucoside yellow, kynurenine yellow
(10) and 3-hydroxykynurenine yellow (11), respectively
(Scheme 1).8
cataractous lenses. In addition, the stability of 7, 8, 9 and
12 is described.
2. Results and discussion
2.1. Synthesis and spectral analysis of the proposed
lens metabolites
In order to investigate if the proposed lens metabolites 7–12
were present in human lenses, authentic standards of each
were synthesised (Schemes 2 and 3). 3-Hydroxyacetophe-
none was nitrated according to the method of Butenandt
et al.14 to afford 2-nitro-3-hydroxyacetophenone (13). 2-Ni-
troacetophenone (14) and 13 were separately condensed
with glyoxylic acid according to a modified method of
Bianchi et al.15 to afford 4-(2-nitrophenyl)-4-oxo-2-butenoic
acid (15) and 4-(3-hydroxy-2-nitrophenyl)-4-oxo-2-but-
enoic acid (16) in 45 and 55% yield, respectively. Hydroge-
nation of 15 and 16 at ca. 1 mg/mL in ethyl acetate with
acetic acid (0.1–0.7%), catalysed by PtO2, afforded 8 in
41% yield and 9 in 47% yield, respectively. Acetic acid
was employed to assist protonation of the amino group pro-
duced and to prevent the formation of Michael adducts.
Treatment of 3 and 4 with NaHCO3 (ca. 0.5 M), following
Despite the similar structures and reactivities of 3 and 4 to 1,
no previous studies have examined human lenses for the
presence of metabolites derived from the deamination of 3
or 4. In this paper we report the synthesis of the reduced
compounds 8 and 9, the GSH adducts 7 and glutathionyl-
3-hydroxykynurenine (12), the cyclised compounds 10 and
11 and the subsequent identification and quantification of
7, 8 and 9 as novel human lens compounds in normal and
O
O
O
O
a
b
c
COOH
COOH
NO2
NH2
NO2
R
R
OH
R
8: R=H
9: R=OH
15: R=H
16: R=OH
14: R=H
13: R=OH
Scheme 2. Synthesis of 8 and 9. Conditions: (a) Cu(NO3)2$2.5H2O, AcOH, Ac2O,w16 h, 10–15 ꢀC, 25% (13); (b) HCOCOOH, 24 h, 110 ꢀC, 45% (15), 55%
(16); (c) H2/PtO2, EtOAc, AcOH, rt, 41% (8), 47% (9).
O
GSH
COOH
a
O
NH2
COOH
O
7: R=H
12: R=OH
NH2
O
COOH
R
R
NH2
NH2
10: R=H
b
11: R=OH
R
R
3: R=H
4: R=OH
N
H
COOH
Scheme 3. (a) Synthesis of 7 and 12. Conditions: GSH, Na2CO3/NaHCO3, pH 9.5, 48 h, 37 ꢀC, 51% (7), 44% (12); (b) synthesis of 10 and 11. Conditions:
NaHCO3, pH 9.5, reflux, 24 h, 26% (10), 22% (11).

4992
J. Mizdrak et al. / Tetrahedron 63 (2007) 4990–4999
the method of Tokuyama et al.,16 promoted deamination and
intramolecular Michael addition to produce 10 in 26% yield
and 11 in 22% yield, respectively. In a similar manner to the
synthesis of the GSH adduct 5,12 reaction of 3 and 4 in
Na2CO3/NaHCO3 buffer at pH 9.5 in the presence of GSH
afforded diastereomeric mixtures (w1:1) of 7 in 51% yield
and 12 in 44% yield, respectively.
revealed the presence of four and three adjacent aromatic
protons, respectively. Three additional protons from the
methylene (d 2.80–2.93) and methine (d 4.20–4.35) groups
were seen in both compounds. Their spectral data were in
agreement with the literature.21,16
The absorbance and fluorescence spectral characteristics of
7, 8, 9 and 12 were similar to those observed by the major
UV filters under the same conditions, i.e., 1: lmax at 263
and 365 nm and maximum fluorescence at lex 360 nm/lem
500 nm; 2: lmax at 259 and 358 nm and maximum fluores-
cence at lex 357 nm/lem 495 nm; 3: lmax at 257 and
359 nm and maximum fluorescence at lex 355 nm/lem
485 nm; 4: lmax at 268 and 370 nm and maximum fluores-
cence at lex 370 nm/lem 460 nm. Compounds 10 and 11
showed two fluorescence maxima at lex 310/392 nm/lem
400/513 nm and lex 370/392 nm/lem 457/547 nm, respec-
tively. This was consistent with the literature.12,19,16,22
Compounds 7, 9 and 12 are novel compounds. The synthesis
of 8 has been recently reported,17 but no spectral data were
provided. The mass spectrum (ES-MS/MS positive mode)
of 8 showed a molecular ion at m/z 194 and major fragment
ions at m/z 176 and 148. These are consistent with loss of wa-
ter and formic acid, respectively. Using the same conditions,
the mass spectrum of 9 revealed a molecular ion at m/z 210
with accompanying fragment ions at m/z 192 and 164 due to
water and formic acid loss, respectively. Furthermore, the
mass spectra of 7 and 12 showed the presence of a molecular
ion at m/z 499 and 515, respectively, and fragment ions due
to loss of glutamic acid and water. These were consistent
with the mass spectra of 5.12
2.2. Identification and quantification of the proposed
lens metabolites in human lenses
1
The aromatic region of the H NMR spectra of 7 and 8 re-
Eight normal human lenses ranging in age from 24 to 88 and
two cataractous lenses of ages 60 and 70 were examined for
the presence of the proposed lens metabolites. The nuclear
and cortical regions were separately extracted with ethanol
to allow isolation of kynurenine-based metabolites.23
HPLC analysis was consistent with previous studies, show-
ing the presence of the major UV filters 1, 2, 3 and 5 at
360 nm.13 A typical lens profile is shown in Figure 2. Com-
pound 4 was not detected in the investigated lenses. This
may be due to the instability of 4, which is an ortho-amino-
phenol and can readily oxidise.24,25 Nuclear and cortical
concentrations of 1 and 2 were found to be 0.1–3.6 nmol/
mg lens (dry mass), while 3 and 5 were found in
0–580 pmol/mg lens (dry mass) (Tables 1 and 2). A marked
decline in the levels of 1, 2 and 3 was noted after the seventh
decade of life, whilst the levels of 5 were generally higher
after middle age. These findings are consistent with previous
vealed four adjacent aromatic resonances with chemical
shifts and coupling patterns consistent with the aromatic
ring of 3.11 Three aromatic protons were found for 9 and
12 with chemical shifts and coupling patterns characteristic
of the aromatic ring of 4.18 Both 8 and 9 contained two iso-
lated triplets at d 2.65–3.26, consistent with the CH2–CH2
side chain.19 The aliphatic side chains of 7 and 12 contained
characteristic signals for a CH2–CH moiety with diastereo-
topic methylene protons at d w3.3 and w3.5 (2H, m) and
a methine proton at d w3.8 (1H, m).11 The distinctive de-
shielded diastereotopic protons are indicative of covalent
attachment of the cysteine of GSH at C-2 of the 3 or 4
side chain.11,12 Further analysis of 7 and 12 by COSY,
HSQC and HMBC confirmed that the GSH moiety was
intact and chemical shifts and coupling constants agreed
1
with the literature.12,20 The H NMR spectra of 10 and 11
0.03
0.02
0.01
0
1
5
2
3
6
0
5
10
15
20
25
30
Time (min)
35
40
45
50
55
60
Figure 2. RP-HPLC profile of UV filter extract from a normal lens nucleus (88 years old) as described in Section 4.13. Detection was at 360 nm. 3-Hydroxy-
kynurenine O-b-D-glucoside (1); 4-(2-amino-3-hydroxyphenyl)-4-oxobutanoic acid O-b-D-glucoside (2); kynurenine (3); glutathionyl-3-hydroxykynurenine
O-b-D-glucoside (5); cysteine-3-hydroxykynurenine O-b-D-glucoside (6).

J. Mizdrak et al. / Tetrahedron 63 (2007) 4990–4999
4993
studies.5 The RP-HPLC absorbance traces of normal and
cataractous lenses did not show the presence of the proposed
lens metabolites, therefore, the eluent from the HPLC re-
gions corresponding to the retention times of the synthetic
compounds were collected, concentrated and investigated
by ES-MS and MS/MS. MS analyses and spiking experi-
ments with the authentic synthetic standards confirmed the
presence of the reduced compounds 8 and 9 and the GSH
adduct of 3 (7). The intramolecular Michael adducts 10
and 11 and the GSH adduct of 4 (12) were not detected in any
of the lenses, suggesting that if present, they would be invery
low levels (the estimated MS sensitivity wasw50 fmol).
concentration of 7 was generally higher than in the cortex
of the same lens. This is consistent with 5, which has been
shown in this and related studies to generally be present in
higher levels in the lens nucleus than in cortex.5,26
Due to the mode of lens removal in surgery, only the nuclear
concentrations of the metabolites could be obtained for
the cataractous lenses. The concentrations were found to
be similar to those found in the normal nucleus (Table 1).
Taylor et al.8 have investigated decomposition rates of
the major UV filters at pH 7.0. They concluded that the
deaminated UV filters cyclise very slowly to the yellow com-
pounds and were more prone to react with lens components,
such as GSH,12 cysteine,13 proteins11,27,28,7 or NADPH29
than to undergo intramolecular Michael addition. Therefore,
the absence of 10 and 11 in the investigated lenses was not
surprising.
The concentration of each lens metabolite was determined
using a standard curve constructed with authentic synthetic
standards. Compounds 8 and 9 were present in both the nu-
cleus and cortex of all normal human lenses in 0–8.2 pmol/
mg lens (dry mass) (Tables 1 and 2). Similar to the lenticular
levels of 2,5 8 and 9 did not show any clear age correlation in
the first four decades, however, a steady decrease in 8 and 9
concentrations was observed after ca. 40 years of age in both
the nucleus and cortex. This is in accordance with the decline
of their metabolic precursors 3 (Tables 1 and 2) and 4.5 Inter-
estingly, while 2 is typically present at 20–40% of the levels
of 1 in normal human lenses,5 8 was only present at 0.5–2%
of the levels of 3. Levels of 9 could not be correlated to levels
of its precursor since 4 was not detected in the investigated
lenses.
2.3. Stability of the novel lens metabolites
The finding of the novel lens metabolites in only low pmol
levels, particularly the reduced compound 8 and GSH adduct
7 (both derived from 3), was surprising, especially given the
much greater concentrations of 2 and 5 typically present in
lenses relative to their precursor (1). Compound 1 has been
shown to deaminate faster at pH 7.0 than 3, but the difference
is not significant enough to account for the lower levels of 7
and 8.8 The stabilities of 7, 8 and 9 were therefore examined
under extraction conditions (used to recover the UV filters
from the human lenses) and HPLC conditions (used to ana-
lyse the extracts). They were also examined in phosphate-
buffered saline (PBS, pH 7.0) under oxygen free conditions
to mimic younger lenses, which have a relatively low oxygen
environment and greater levels of antioxidants,30,31 and in
the presence of atmospheric oxygen to resemble older
lenses, which are typically depleted of antioxidants.31–33
In addition, the stability of 12 was investigated to confirm
if the lack of its precursor (4) or stability of 12 may have
contributed to its absence from the investigated lenses.
Similar to 8 and 9, the concentration of 7 was found to be low
(0.3–28 pmol/mg lens (dry mass)) in both the nucleus and
cortex. Even though no clear age-related correlation was
seen for the concentrations of 7, the normal nuclear
Table 1. Quantification of lens compounds in nuclei of normal and catarac-
tous lenses
Nucleus age 1b
2b
3b
5b
7a
8a
9a
24
27
42
47
65
66
83
88
60c
70c
3600 692
3630 690
1840 408
2560 897
2220 527
658 268
586
385
135
261
188
0.00
50.4
39.5
282
562
89.9
41.1
14.9 210
1.80 4.00 2.67
1.05 4.39 5.97
0.63 0.86 1.14
4.76 1.87 5.20
28.0
1.83 2.40
3.09 1.37 0.69
2.00 0.70 0.00
7.60 0.32 0.63
0.11 1.83 0.99
0.10 1.21 1.31
Known quantities of 7, 8, 9 and 12 in the presence of bovine
lens tissue were separately extracted with 80% ethanol. Re-
covery was 57–79% (data not shown). This was comparable
to the literature.34 Additionally, 7, 8, 9 and 12 were stable
over 24 h under extraction conditions and up to 5 h under
HPLC conditions (data not shown). This confirms that the
low concentrations measured for 7, 8 and 9, and the lack
of detection of 12, were not due to significant loss/break-
down during extraction and analysis of the lens extracts
and are representative of the concentrations in the investi-
gated lenses.
73.1
61.4
426 72.0
452 127
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
pmol/mg lens (dry mass). N/A not determined.
Determined by LC–MS.
a
b
Determined by HPLC.
Cataractous lenses.
c
Table 2. Quantification of lens compounds in cortices of normal lenses
Cortex age 1b
2b
3b
5b
7a
8a
9a
Compound 8 proved to be stable under physiological condi-
tions in the presence of atmospheric oxygen for >200 h
(data not shown). This suggests that formation of 8 may pro-
tect the lens from modification due to its greater stability
compared to its precursor (3).8 By contrast, 9 decomposed
steadily under the same conditions (Fig. 3) resulting in its
total disappearance after 7 days of incubation. The decom-
position of 9 was accompanied by the generation of high
molecular mass compounds (molecular masses >400) and
predominantly less polar peaks, which were probably
24
27
42
47
65
66
83
88
3100 558
2300 328
1500 365
3850 891
1900 407
1490 311
732 147
317
226
114
361
177
80.3
93.8
10.8
0.00
14.2
27.5
56.9
0.34 2.24 1.76
0.30 1.39 0.79
2.21 8.20 6.04
1.23 2.82 3.03
173
18.4
1.40 2.79
2.36 1.14 1.11
0.52 0.85 0.58
2.58 0.21 0.37
71.7
44.8
48.5
418
90.7
pmol/mg lens (dry mass).
Determined by LC–MS.
Determined by HPLC.
a
b

4994
J. Mizdrak et al. / Tetrahedron 63 (2007) 4990–4999
0.3
0.2
0.1
0
0
5
0
100
150
200
Time (h)
Figure 3. Compound 9 (0.25 mM) was incubated in PBS (pH 7.0) in the presence of atmospheric oxygen or in oxygen free PBS (pH 7.0) in the presence
of ascorbic acid (2.0 mM) and butylated hydroxytoluene (100 mM). Aliquots of the reaction mixtures were taken at the indicated time points and analysed
by RP-HPLC as described in Section 4.2. Detection was at 254 nm. A—aerobic; -—anaerobic.
dimeric aggregates arising from oxidation of this ortho-ami-
nophenol.35–37 The initially colourless incubation mixtures
also became yellow/tanned over time, exhibiting lmax at
240 and 420 nm, which are characteristic absorption peaks
of xanthommatin based compounds.38,39 A significantly
greater stability of 9 was seen under anaerobic conditions
in the presence of the potent radical scavengers, butylated
hydroxytoluene and ascorbic acid, at pH 7.0. Only 27% of
9 was decomposed after 9 days of incubation in the presence
of these agents (Fig. 3). Compound 9 is therefore expected to
be unstable in the oxidising environment of cataractous
lenses and may result in lens colouration and possibly
contribute towards age-related nuclear cataract.9,40
Compound 7 decreased in concentration by 60–64% under
physiological conditions, independent of the level of oxygen
(Fig. 4). Similar to 7, a significant decrease in the concentra-
tion of 12 was noted both in the aerobic (74% loss) and
anaerobic conditions (80% loss) (Fig. 5). The breakdown
products were identified by LC–MS and UV–vis as the
500
400
300
200
100
0
500
400
300
200
100
0
Aerobic
Anaerobic
0
50
100
Time (h)
150
200
0
5
0
100
Time (h)
150
200
Figure 4. Compound 7 (0.4 mM) was incubated in PBS (pH 7.0) in the presence of atmospheric oxygen or in oxygen free PBS (pH 7.0) in the presence of
ascorbic acid (0.2 mM) and butylated hydroxytoluene (100 mM). Aliquots of the reaction mixtures were taken at the indicated time points and analysed by
RP-HPLC as described in Section 4.2. Detection was at 254 nm. A—7; -—deaminated 3; :—10.

J. Mizdrak et al. / Tetrahedron 63 (2007) 4990–4999
4995
200
150
100
50
200
150
100
50
Anaerobic
Aerobic
0
0
50
100
Time (h)
150
200
0
0
50
100
Time (h)
150
200
Figure 5. Compound 12 (0.1 mM) was incubated in PBS (pH 7.0) in the presence of atmospheric oxygen or in oxygen free PBS (pH 7.0) in the presence
of ascorbic acid (2.0 mM) and butylated hydroxytoluene (100 mM). Aliquots of the reaction mixtures were taken at the indicated time points and analysed
by RP-HPLC as described in Section 4.2. Detection was at 254 nm. A—12; -—deaminated 4; :—11.
a,b-unsaturated ketones (via elimination of GSH) and yel-
low compounds, 10 and 11. Similar instability of 3 and 4
was observed by Taylor et al.8 The dynamic nature of GSH
adduct formation and breakdown suggests that the GSH
adducts may delay binding of deaminated 1, 3 and 4 to lens
proteins in younger lenses that have high levels of GSH
(w4.5–6 mM),5 but to a lesser degree in aged (GSH depleted)
lenses.
DL-Kynurenine sulfate salt (ꢁ95%), 3-hydroxy-DL-kynure-
nine, GSH (99%), glyoxylic acid monohydrate (98%), Cu-
(NO3)2$2.5H2O (99%), 2-nitroacetophenone (14) (95%)
and trifluoroacetic acid (TFA) (>99%) were from Sigma–Al-
drich. Glacial acetic acid (AcOH, >99.9%), acetic anhydride
(99.0%) and ascorbic acid (99.7%) were purchased from
BDH. Chelex resin (100–200 mesh) was purchased from
BioRad and butylated hydroxytoluene from CalBiochem.
CD3OD (99.8%), CD3COCD3 (99.9%) and CDCl3 (99.8%)
were from Cambridge Isotope Laboratories. MilliQÒ H2O
(purified to 18.2 MU/cm2) was used in the preparation of
all solutions. Dulbecco’s phosphate-buffered saline (PBS),
without calcium and magnesium, consisted of KCl
2.7 mM, KH2PO4 1.4 mM, NaCl 137 mM and Na2HPO4
7.68 mM.41 The pH was adjusted to 7.0. Pre-washed chelex
resin was added to the PBS buffer (w2 g/L) and left for
24 h prior to use. Normal human lenses were obtained post-
mortem from donor eyes at the Sydney Eye Bank (Sydney,
Australia) and cataract lenses were obtained from K. T.
Seth Eye Hospital (Rajkot, Gujarat, India) with ethical
approval from the University of Wollongong Human Ethics
Committee (HE99/001). After removal, lenses were immedi-
ately placed into sterile plastic screw-capped vials and kept at
ꢂ80 ꢀC until analysed. Thin-layer chromatography (TLC)
plates were of normal phase 60F254 (Merck, Germany) and
developed using a mobile phase of n-butanol/AcOH/H2O
(BAW, 12:3:5, v/v/v), unless otherwise stated, and reversed
phase 18F254 using a mobile phase of 20% CH3CN/H2O (v/
v). TLC plates were visualised under UV light (254 and
365 nm) and sprayed with ninhydrin. C18 reversed phase
Sep-PackÒ cartridges were purchased from Waters. Normal
phase silica gel (230–400 mesh) was from Merck (Germany).
3. Conclusion
This study describes the identification, via total synthesis
and spectral analysis, and quantification of three novel
human lens metabolites, 4-(2-aminophenyl)-4-oxobutanoic
acid (8), 4-(2-amino-3-hydroxyphenyl)-4-oxobutanoic acid
(9) and glutathionyl-kynurenine (7). These metabolites
were isolated in pmol/mg levels (dry mass) in normal and
cataractous lenses. Their spectral characteristics suggest
that these compounds may have the capability to act as
UV filters and protect the lens from UV-induced photo-
damage. Compound 8 was found to be very stable under
conditions mimicking the lens environment. In contrast, both
7 and 9 were not stable. In particular, 9 was unstable under
aerobic conditions mimicking older lenses and produced
yellow-tanned products that may contribute to lens coloura-
tion observed with lens aging and age-related nuclear cata-
ract. Compounds 7 and 12 showed similar decomposition
rates to their precursors 3 and 4, respectively.
4. Experimental
4.1. Materials
4.2. Instruments
Acetonitrile (CH3CN) was of HPLC grade. All other organic
solvents were of AR grade and distilled prior to use.
Melting points were determined on a SMP 10 Stuart scien-
tific (UK) apparatus and are uncorrected. Infrared spectra

4996
J. Mizdrak et al. / Tetrahedron 63 (2007) 4990–4999
were recorded on a Perkin–Elmer Paragon 1000 PC FTIR
spectrometer.
phase gradient as follows: 0–10 min (5% buffer B), 10–
40 min (5–55% buffer B), 40–45 min (55% buffer B), 45–
50 min (55–5% buffer B) and 50–60 min (5% buffer B).
Compounds 10 and 11 were purified by the above prepara-
tive method while the mobile phase gradient was: 0–
10 min (20% buffer B), 10–30 min (20–80% buffer B),
30–35 min (80% buffer B), 35–40 min (80–20% buffer B)
and 40–50 min (20% buffer B).
1H, 13C, COSY (1H–1H correlation spectroscopy), HSQC
(1H–13C heteronuclear single quantum correlation) and
HMBC (1H–13C heteronuclear multiple bond correlation)
NMR experiments were acquired on a Bruker Avance 400
13
spectrometer (1H, 400 MHz; C, 100 MHz) at 25 ꢀC.
NMR spectra were run in CD3OD unless otherwise stated.
The solvent signal was used as the internal reference. Reso-
nances are quoted in parts per million and coupling constants
(J) are given in Hertz.
4.3. Synthesis of 2-nitro-3-hydroxyacetophenone (13)
3-Hydroxyacetophenone (20.0 g, 0.15 mol) was dissolved
in a mixture of glacial acetic acid (110 mL, 1.92 mol) and
acetic anhydride (12.0 mL, 0.13 mol). This colourless solu-
tion was stirred and cooled down to 10–15 ꢀC in a water/ice
bath. Ground Cu(NO3)2$2.5H2O (40.0 g, 0.17 mol) was
added to the reaction mixture over 3 h. The reaction mixture
changed in colour to dark blue-green. The reaction was
monitored by TLC (DCM, Rf 0.60) over a period of 8 h
and left in the fridge O/N. The next day the reaction was
continued at 10–15 ꢀC for 7–8 h, with initial addition of
ground Cu(NO3)2$2.5H2O (20.0 g, 0.08 mol) over a period
of 1 h. Water (w200 mL) was added to the green solution
and the reaction was stirred at 10–15 ꢀC for w1 h. The
yellow precipitate was vacuum filtered, washed with cold
water and air dried. The filtrate was extracted with DCM,
washed with saturated brine, dried with MgSO4 and decol-
ourised with activated carbon. The yellow dried solid was
purified by two sequential normal phase chromatography
steps (DCM, followed by toluene/EtOAc, 5:1, v/v) to afford
13 (6.65 g, 25%, mp 134–136 ꢀC, lit.14 131–132 ꢀC) as
a yellow solid. nmax (KBr disc): 3100 (br), 1668, 1531,
1376, 1291, 798 cmꢂ1; dH (CDCl3): 10.50 (1H, OH, s),
7.59 (1H, dd, J 7.3, 8.5, ArH-5), 7.23 (1H, dd, J 1.2, 8.5,
ArH-6), 6.85 (1H, dd, J 1.2, 7.3, ArH-4), 2.50 (3H, s,
CH3); dC: 199.4 (CO-2), 155.0 (ArC-3), 140.6 (ArC-2,
ArC-1), 136.9 (ArC-5), 121.0 (ArC-6), 118.1 (ArC-4),
30.3 (C-1).
Electrospray mass spectrometry (ES-MS) data were ob-
tained in positive ion mode using a Micromass Q-TOF2
equipped with a nanospray source. MS settings were as fol-
lows: cone voltage 25 V, LM/HM 12/12, MCP 2300 V, mass
range 50–600 m/z. For tandem MS (MS/MS) analysis, ions
were subjected to a range of collisions energy settings (typ-
ically between 10–25 eV). High-resolution mass spectro-
metry (HRMS) was performed on a Q-TOF Ultima with
lock spray. Leucine-encephalin (555.2692 Da) was used as
the reference compound.
UV–vis absorbance spectra were obtained using a Varian
DMS 90 UV–vis spectrometer. Fluorescence spectra were
recorded on a Perkin–Elmer LS55 luminescence spectro-
meter. Slit widths were 10 nm for excitation and emission.
PBS was used as solvent for all measurements.
Liquid chromatography–mass spectrometry (LC–MS) was
run on a Shimadzu LC–MS-2010EVunit attached to an elec-
trospray ionisation mass spectrometer (ES-MS). An Alltech
˚
(Prevail, 100 A, 5 mm, 2.1ꢃ150 mm, C18) column was used
at 27 ꢀC. Mobile phase/gradient conditions consisted of
buffer A (H2O/0.1% formic acid v/v) and buffer B
(CH3CN/0.1% formic acid); 0–3 min (20% buffer B), 3–
20 min (20–100% buffer B), 20–25 min (100% buffer B),
25–27 min (100–20% buffer B) and 27–35 min (20% buffer
B). The flow rate was 0.2 mL/min and the source tempera-
ture was maintained at 170 ꢀC. All spectra were acquired
in continuum mode.
4.4. Synthesis of 4-(2-nitrophenyl)-4-oxobut-2-enoic acid
(15)
2-Nitroacetophenone (14) (0.50 mL, 3.03 mmol) was com-
bined with melted glyoxylic acid monohydrate (2.78 g,
30.2 mmol) at 60 ꢀC. The temperature was increased to
110 ꢀC and the yellow reaction was left under vacuum.
The reaction progress was monitored by TLC (EtOAc/1%
AcOH, Rf 0.46). After 5 h another portion of glyoxylic
acid monohydrate (0.20 g, 2.17 mmol) was added. After
14 h, the brown viscous reaction mixture was mixed with
normal phase silica (w3 g) while still warm and set aside
to cool down to rt. The mixture was purified by normal phase
chromatography (DCM/1% AcOH) to yield 15 (0.3 g, 45%,
mp 171–172 ꢀC, lit.15 170–173 ꢀC) as a white solid. nmax
(KBr disc): 3500–2300 (br), 1707, 1679, 1519, 1343,
1317, 1297, 740 cmꢂ1; dH (acetone-d6): 8.25 (1H, d, J 8.0,
ArH-3), 7.95 (1H, ddd, J 1.1, 7.5, 7.5, ArH-5), 7.85 (1H,
ddd, J 1.3, 7.5, 8.0, ArH-4), 7.69 (1H, dd, J 1.3, 7.5, ArH-
6), 7.26 (1H, d, J 16.1, H-3), 6.45 (1H, dd, J 16.1, H-2);
dC: 192.8 (CO-4), 166.2 (CO-1), 147.6 (ArC-2), 139.9
(C-3), 135.8 (ArC-1), 135.5 (ArC-5), 134.4 (C-2), 132.5
(ArC-4), 129.8 (ArC-6), 125.4 (ArC-3); ES-MS m/z: 222
(MH+, 89%).
Reversed phase high performance liquid chromatography
(RP-HPLC) was performed on a Shimadzu HPLC. Standard
curves and stability sample analyses were performed on
˚
a Phenomenex (Luna, 100 A, 5 mm, 4.6ꢃ250 mm, C18) col-
umn with the following mobile phase system: buffer A
(H2O/0.05% TFA v/v) and buffer B (80% CH3CN). The
flow rate of 1 mL/min was kept constant with a mobile phase
gradient as follows: 0–3 min (20% buffer B), 3–15 min (20–
90% buffer B), 15–18 min (90% buffer B), 18–22 min (90–
20% buffer B) and 22–28 min (20% buffer B). Detection was
at 254 and 360 nm. Preparative separations were performed
on the same instrument and using the same mobile phase
system as for the analytical separations. Compounds 8 and
˚
9 were purified using a Phenomenex (Luna, 100 A, 10 mm,
15ꢃ250 mm, C18) column. The flow rate of 7 mL/min
was kept constant with a mobile phase gradient as follows:
0–10 min (10% buffer B), 10–40 min (10–70% buffer B),
40–50 min (70% buffer B), 50–55 min (70–10% buffer B)
and 55–65 min (10% buffer B). Compounds 7 and 12 were
purified by the above preparative method with a mobile

J. Mizdrak et al. / Tetrahedron 63 (2007) 4990–4999
4997
4.5. Synthesis of 4-(3-hydroxy-2-nitrophenyl)-4-oxobut-
2-enoic acid (16)
reversed phase, Rf 0.57) when visualised by UV light and
sprayed with ninhydrin. The yellow reaction mixture was
worked up similarly to 8 and further purification was
achieved by using a C18 reversed phase Sep-pack column
to yield 9 in w80–85% purity. Subsequent purification by
preparative RP-HPLC (tR 23.8 min) yielded 9 as a light
brown solid (59 mg, 47%, mp 134–135 ꢀC) in 99% purity.
Found: M, 209.0701. Calculated for C10H11NO4: M,
209.0688; nmax (KBr disc): 3500–2300 (br), 3370, 1709,
1681, 1652, 1195, 1149, 788, 719 cmꢂ1; dH: 7.82 (1H,
dd, J 1.3, 8.3, ArH-6), 7.25 (1H, dd, J 1.3, 8.0, ArH-4),
6.57 (1H, dd, J 8.0, 8.3, ArH-5), 3.26 (2H, t, J 6.4, H-2),
2.66 (2H, t, J 6.4, H-3); dC: 202.0 (CO-4), 201.8 (CO-4),
176.9 (CO-1), 147.1 (ArC-3), 139.0 (ArC-2), 122.5
(ArC-6), 120.4 (ArC-1), 118.3 (ArC-4), 117.6 (ArC-5),
35.0 (C-2), 29.1 (C-3); ES-MS/MS of m/z: 210.08 (MH+,
100%), 192.07 (28%), 164.07 (54%). lmax at 267.4 and
369.4 nm and maximum fluorescence at lex 346.5 nm/lem
435.0 nm.
3-Hydroxy-2-nitroacetophenone (13) (0.4 g, 2.21 mmol)
was combined with melted glyoxylic acid monohydrate
(2.0 g, 21.7 mmol) at 60 ꢀC. The yellow reaction mixture
was further heated to 120 ꢀC under vacuum for 24 h. The re-
action progress was monitored by TLC (EtOAc/1% AcOH,
Rf 0.48). After both 7 and 14 h, glyoxylic acid monohydrate
(0.20 g, 2.17 mmol) was added. The brown viscous reaction
mixture was mixed with normal phase silica (w3 g) while
warm and set aside to cool down to rt. The mixture was pu-
rified by normal phase chromatography (DCM/1% AcOH)
to yield 16 (0.28 mg, 55%, 158–159 ꢀC, lit.14 158 ꢀC) as
a yellow solid. nmax (KBr disc): 3500–2300 (br), 3300,
1704, 1676, 1601, 1518, 1431, 1273, 1168 cmꢂ1; dH
(CDCl3): 10.51 (1H, OH, s), 7.66 (1H, dd, J 7.3, 8.5, ArH-
5), 7.31 (1H, dd, J 1.2, 8.5, ArH-6), 7.31 (1H, d, J 16.2,
H-3), 6.90 (1H, dd, J 1.2, 7.3, ArH-4), 6.40 (1H, d, J 16.1,
H-2); dC: 190.8 (CO-4), 166.3 (CO-1), 152.9 (ArC-3),
138.9 (C-3), 136.6 (ArC-2), 134.9 (ArC-1), 134.8 (ArC-5),
134.5 (C-2), 123.0 (ArC-6), 121.1 (ArC-4); ES-MS m/z:
238 (MH+, 67%).
4.8. Synthesis of glutathionyl-kynurenine (7)
DL-Kynurenine sulfate salt (100 mg, 0.32 mmol) was dis-
solved in argon-bubbled (w20 min) Na2CO3/NaHCO3
buffer (20 mL, 25 mM, pH 9.5). GSH (500 mg,
1.62 mmol) was added and the light yellow solution was
bubbled with argon for w20 min, sealed with parafilm and
incubated in the dark at 37 ꢀC with shaking. The progress
of the reaction was monitored by normal phase TLC
(Rf 0.25) and reversed phase TLC (Rf 0.86). After 24 h,
another portion of GSH (100 mg, 0.32 mmol) was added
to the reaction mixture and the pH was readjusted to
9.5. After 72 h, the light yellow reaction mixture was
acidified to pH 2 by dropwise addition of 25% HCl (v/v)
and lyophilised. The crude solid was purified by preparative
RP-HPLC (tR 34.6 min), as described above, to obtain 7
(121 mg, 51%) as aw1:1 mixture of diastereomers. Found:
MH+, 499.1534. Calculated for C20H26N4O9S: MH+,
499.1499; nmax (KBr disc): 3500–2300 (br), 1722, 1650
(br), 1545, 1200 (br) cmꢂ1; dH: 7.77 (1H, br d, J 8.2,
ArH-6), 7.24 (1H, ddd, J 1.0, 7.6, 8.3, ArH-4), 6.73 (1H,
d, J 8.3, ArH-3), 6.60 (1H, dd, J 7.6, 8.2, ArH-5), 4.74
(w0.5H, dd, J 4.8, 9.2, SCH2CH), 4.68 (w0.5H, dd, J
5.7, 7.9, SCH2CH), 4.00 (1H, t, J 6.2, CH2CHCOOH),
3.94 (2H, s, CH2COOH), 3.87 (1H, dd, J 3.8, 9.9,
H-2), 3.62 (1H, m, H-3), 3.34 (1H, m, H-3), 3.31 (w0.5H,
dd, J 4.8, 14.0, SCH2), 3.22 (w0.5H, dd, J 5.7, 13.9,
SCH2), 3.07 (w0.5H, dd, J 7.9, 13.9, SCH2), 2.92 (w0.5H,
dd, J 9.2, 14.0, SCH2), 2.58 (2H, t, J 6.9, COCH2CH2),
2.20 (2H, m, COCH2CH2); dC: 200.1 (CO-4), 200.0
(CO-4), 175.9 (CO-1), 175.8 (CO-1), 174.5 (NHCOCH2),
174.4 (NHCOCH2), 172.8 (CHCONH), 172.8 (CHCONH),
172.6 (CH2COOH), 171.7 (CHCOOH), 152.6 (ArC-2),
135.6 (ArC-4), 132.0 (ArC-6), 118.3 (ArC-3), 117.9
(ArC-1), 116.2 (ArC-5), 54.5 (SCH2CH), 53.8 (SCH2CH),
53.7 (CH2CHCOOH), 43.3 (C-2), 42.6 (C-2), 42.6 (C-
3), 42.2 (C-3), 41.8 (CH2COOH), 34.5 (SCH2), 34.4
(SCH2), 32.4 (COCH2CH2), 27.1 (COCH2CH2), 27.0
(COCH2CH2); ES-MS/MS of m/z: 499.0 (MH+, 11%),
424.0 (30%), 370.0 (21%), 352.0 (62%), 259.0 (100%),
192.0 (21%), 179.0 (76%), 174.0 (30%). lmax 256 and
356 nm and maximum fluorescence at lex 350.5 nm/lem
475.0 nm.
4.6. Synthesis of 4-(2-aminophenyl)-4-oxobutanoic acid
(8)
A mixture of 1515 (255 mg, 1.15 mmol), EtOAc (280 mL),
AcOH (450 mL) and PtO2 (40 mg, 0.17 mM) was treated
with H2 gas at rt in the dark for 22 h. TLC analysis revealed
multiple spots including 8 (normal phase, EtOAc/1% AcOH,
Rf 0.88 and reversed phase, Rf 0.37) when visualised by UV
light and sprayed with ninhydrin. The yellow reaction mix-
ture was gravity filtered through a plug of Celite and the sol-
vent was removed under vacuum. Crude 8 was dissolved in
10% CH3CN/H2O and loaded onto a preconditioned C18
reversed phase Sep-pack column and eluted with an increas-
ing gradient of CH3CN (10–40%). Fractions containing the
product were pooled and lyophilised to yield 8 inw80–85%
purity. Further purification was achieved by preparative
RP-HPLC as described above. The fraction containing 8 (tR
31.3 min) was collected and lyophilised to yield an off-white
solid (43 mg, 41%, mp 92–94 ꢀC) in 99% purity. Found: M,
193.0745. Calculated for C10H11NO3: M, 193.0739; nmax
(KBr disc): 3500–2300 (br), 3486, 3465, 3338, 1709,
1650, 763 cmꢂ1; dH: 7.82; (1H, dd, J 1.5, 8.2, ArH-6),
7.25 (1H, ddd, J 1.5, 7.0, 8.0, ArH-4), 6.75 (1H, dd, J 1.0,
8.0, ArH-3), 6.65 (1H, ddd, J 1.0, 7.0, 8.2, ArH-5), 3.25
(2H, t, J 6.4, H-2), 2.65 (2H, t, J 6.4, H-3); dC: 201.7
(CO-4), 177.0 (CO-1), 152.1 (ArC-2), 135.3 (ArC-4),
132.0 (ArC-6), 118.6 (ArC-1), 118.4 (ArC-3), 116.4 (ArC-
5), 35.7 (C-2), 29.1 (C-3); ES-MS/MS of m/z: 194.08
(MH+, 37%), 176.06 (100%), 148.03 (62%). lmax 254 and
365 nm and maximum fluorescence at lex 346.5 nm/lem
480.0 nm.
4.7. Synthesis of 4-(2-amino-3-hydroxyphenyl)-4-oxo-
butanoic acid (9)
A
mixture of 1614,15 (143 mg, 0.60 mmol), EtOAc
(185 mL), AcOH (250 mL) and PtO2 (35 mg, 0.15 mmol)
was treated with H2 gas at rt in the dark for 3 h. TLC analysis
revealed 9 (normal phase, EtOAc/1% AcOH, Rf 0.65 and

4998
J. Mizdrak et al. / Tetrahedron 63 (2007) 4990–4999
4.9. Synthesis of glutathionyl-3-hydroxykynurenine (12)
4.11. Synthesis of 3-hydroxykynurenine yellow
(8-hydroxy-4-oxo-1,2,3,4-tetrahydroquinoline-2-
carboxylic acid, 11)
3-Hydroxy-DL-kynurenine (100 mg, 0.44 mmol) was dis-
solved in argon-bubbled (w20 min) Na2CO3/NaHCO3 buffer
(100 mL, 25 mM, pH 9.5). GSH (500 mg, 1.62 mmol) was
added to the light yellow solution and the solution was bub-
bled with argon (w20 min), sealed with parafilm and incu-
bated in the dark at 37 ꢀC with shaking. The progress of
the reaction was monitored by normal phase TLC (Rf 0.28)
and reversed phase TLC (Rf 0.87). Another portion of
GSH (100 mg, 0.32 mmol) was added after 24 h and the
pH was readjusted to 9.5. Similar to 7, after 72 h the light
yellow reaction mixture was acidified to pH 2 and purified
by RP-HPLC (tR 24.2 min), as described above, to obtain
12 (102 mg, 44%) as a w1:1 mixture of diastereomers.
Found: MH+, 515.1520. Calculated for C20H26N4O10S:
MH+, 515.1448; nmax (KBr disc): 3500–2300 (br), 1721,
1670 (br), 1545, 1200 (br) cmꢂ1; dH: 7.20 (w0.5H, dd, J
1.3, 8.4, ArH-6), 7.19 (w0.5H, dd, J 1.3, 8.4, ArH-6), 6.70
(1H, br d, J w7.5, ArH-4), 6.40 (1H, br dd, J w7.5, 8.4,
ArH-5), 4.63 (w0.5H, dd, J 4.7, 9.2, SCH2CH), 4.57
(w0.5H, dd, J 5.9, 8.0 SCH2CH), 3.90 (1H, t, J 6.7,
CH2CHCOOH), 3.82 (1H, s, CH2COOH), 3.81 (1H, s,
CH2COOH), 3.76 (1H, m, H-2), 3.50 (1H, m, H-3), 3.25
(1H, m, H-3), 3.20 (w0.5H, dd, J 4.7, 14.2, SCH2), 3.10
(w0.5H, dd, J 5.9, 14.1, SCH2), 2.97 (w0.5H, dd, J 8.0,
14.1, SCH2), 2.80 (w0.5H, dd, J 9.2, 14.2, SCH2), 2.47
(2H, t, J 7.0, COCH2CH2), 2.10 (2H, m, COCH2CH2); dC:
200.2 (CO-4), 200.1 (CO-4), 175.9 (CO-1), 175.9 (C-1),
174.5 (NHCOCH2), 174.4 (NHCOCH2), 172.9 (CHCONH),
172.8 (CHCONH), 172.6 (CH2COOH), 171.7 (CHCOOH),
146.3 (ArC-3), 142.0 (ArC-2), 122.4 (ArC-6), 118.4 (ArC-
1), 118.0 (ArC-4), 115.9 (ArC-5), 54.6 (SCH2CH), 53.8
(SCH2CH), 53.7 (CH2CHCOOH), 43.3 (C-2), 42.8 (C-2),
42.6 (C-3), 42.4 (C-3), 41.8 (CH2COOH), 34.5 (SCH2),
34.5 (SCH2), 32.4 (COCH2CH2), 27.1 (COCH2CH2), 27.0
(COCH2CH2); ES-MS/MS of m/z: 515.1 (MH+, 32%),
440.1 (45%), 386.1 (32%), 368.1 (100%), 259.0 (97%),
208.1 (20%), 179.1 (41%). lmax 256.5 and 356.5 nm and
maximum fluorescence at lex 350.5 nm/lem 475.0 nm.
Compound 11 was synthesised by the procedure of To-
kuyama et al.16 The progress of the reaction was monitored
by normal phase TLC (Rf 0.78). Purification by preparative
RP-HPLC (tR 15.5 min), as described above, afforded 11
(22%) as a yellow solid. The spectral data were in agreement
with the literature.16,21 Found: MH+, 207.0386. Calculated
for C10H9NO4: MH+, 207.0532; nmax (KBr disc): 3500–
2400 (br), 3400 (br), 1635, 1605, 1510, 1400, 1269, 1223,
787, 732 cmꢂ1; dH (acetone-d6): 7.21 (1H, dd, J 1.3, 7.9,
H-5), 6.82 (1H, dd, J 1.3, 7.9, H-7), 6.50 (1H, dd, J 7.9,
7.9, H-6), 4.20 (1H, dd, J 5.0, 10.5, H-2), 2.93 (1H, dd, J
5.0, 16.5, H-3), 2.80 (1H, dd, J 10.5, 16.5, H-3), dC: 193.4
(CO-4), 173.6 (COOH), 143.8 (C-8), 141.1 (C-8a), 117.3
(C-4a), 116.7 (C-7), 115.7 (C-5), 115.1 (C-6), 54.3 (C-2),
39.1 (C-3); ES-MS/MS of m/z: 208.0 (MH+, 9%), 190.0
(27%), 162.0 (100%), 148.0 (5%), 120.0 (20%). lmax 277
and 383 nm and maximum fluorescence at lex 370/392 nm
and lem 457/547 nm.
4.12. Stability studies
4.12.1. Stability under extraction, HPLC and pH 7.0
(aerobic) conditions. Compound 8 was dissolved in 80%
ethanol/MilliQÒ H2O (v/v) to give a final concentration of
1.6 mM. The solutions (3ꢃ4 mL) were sealed and incubated
in the dark at 37 ꢀC with gentle shaking. Single aliquots
were taken from each solution over 24 h and analysed by
RP-HPLC. Separate incubations were repeated with 9
(0.2 mM), 7 (0.4 mM) and 12 (0.1 mM). Similar stability
studies were carried out in MilliQÒ H2O/0.01% TFA (v/v)
over 24 h and chelex treated PBS (pH 7.0) in the presence
of atmospheric oxygen over 9 days.
4.12.2. Stability at pH 7.0 (anaerobic). Solutions of 9
(0.25 mM) with butylated hydroxytoluene (100 mM) and as-
corbic acid (2.0 mM) in argon degassed chelex treated PBS
(3ꢃ4 mL) were incubated in the dark at 37 ꢀC with gentle
shaking. Single aliquots were taken over 9 days from each
solution and analysed by HPLC. Separate stability studies
were repeated with 7 and 12 under similar conditions.
4.10. Synthesis of kynurenine yellow (4-oxo-1,2,3,4-
tetrahydroquinoline-2-carboxylic acid, 10)
Compound 10 was synthesised by the procedure of To-
kuyama et al.16 The progress of the reaction was monitored
by normal phase TLC (Rf 0.85) and reversed phase TLC (Rf
0.61). Purification by preparative RP-HPLC (tR 19.2 min),
as described above, afforded 10 (26%) as a yellow-orange
solid. The spectral data were in agreement with the litera-
ture.16 Found: MH+, 191.0571. Calculated for C10H9NO3:
MH+, 191.0582; nmax (KBr disc): 3500–2300 (br), 3341,
1695, 1655, 1618, 1283, 760 cmꢂ1; dH (acetone-d6): 7.66
(1H, dd, J 1.5, 7.9, H-5), 7.30 (1H, ddd, J 1.5, 7.4, 8.0, H-
7), 6.96 (1H, br d, J 8.0, H-8), 6.67 (1H, ddd, J 0.7, 7.4,
7.9, H-6), 6.20 (1H, s, N–H), 4.35 (1H, dd, J 5.5, 8.7, H-
2), 2.90 (1H, dd, J 5.5, 16.2, H-3), 2.80 (1H, dd, J 8.7,
16.2, H-3), 1.40 (1H, s, COOH); dC: 191.6 (CO-4), 173.0
(COOH), 152.1 (C-8a), 135.8 (C-7), 127.4 (C-5), 119.5
(C-4a), 118.1 (C-6), 117.1 (C-8), 55.1 (C-2), 40.58 (C-3);
ES-MS/MS of m/z: 192.2 (MH+, 24%), 174.1 (32%), 146.1
(100%), 132.1 (14%). lmax 260 and 378 nm and maximum
fluorescence at lex 310/392 nm and lem 400/513 nm.
4.13. Lens preparation and RP-HPLC purification
Eight normal human lenses of ages 24, 27, 42, 47, 65, 66, 83
and 88 and two cataractous human lenses, assigned as Type
III–IVaccording to the Pirie classification system,42 of ages
60 and 70, were separated into cortex and nucleus using
a 5 mm cork borer. The ends of each nuclear core were re-
moved and added to the cortex. UV filters were extracted
as described previously.23 The average extraction recovery
was measured by separately adding 7, 8, 9 and 12 in
500 pmol/mg lens protein. Subsequent extraction and
HPLC analysis was as described. The retention times on
RP-HPLC and ES-MS/MS spectra of synthetic 7–12 were
determined. Lens extracts were separated by RP-HPLC
˚
using a Microsorb (MV, 100 A, 4.6ꢃ250 mm, C18) column.
The column was equilibrated in buffer A (H2O/0.1% TFA
(v/v)) at a flow rate of 0.5 mL/min. The following gradient
was used: buffer B (80% CH3CN/0.1% TFA, v/v), 0–5 min

J. Mizdrak et al. / Tetrahedron 63 (2007) 4990–4999
4999
11. Vazquez, S.; Aquilina, J. A.; Jamie, J. F.; Sheil, M. M.;
Truscott, R. J. W. J. Biol. Chem. 2002, 277, 4867–4873.
12. Garner, B.; Vazquez, S.; Griffith, R.; Lindner, R. A.; Carver,
J. A.; Truscott, R. J. W. J. Biol. Chem. 1999, 274, 20847–20854.
13. Hains, P. G.; Mizdrak, J.; Streete, I. M.; Jamie, J. F.; Truscott,
R. J. W. FEBS Lett. 2006, 580, 5071–5076.
(0–50% buffer B), 5–50 min (50% buffer B), 50–55 min,
(50–0% buffer B) and 55–60 min (0% buffer B). The eluent
was monitored at 360 nm. Fractions were collected at the
known elution times of 8 (26.5% CH3CN/0.1% TFA), 9
(20% CH3CN/0.1% TFA), 7 (26.5% CH3CN/0.1% TFA),
12 (20% CH3CN/0.1% TFA), 10 (32.5% CH3CN/0.1%
TFA) and 11 (28% CH3CN/0.1% TFA).
14. Butenandt, A.; Hallman, G.; Beckmann, R. Chem. Ber. 1957,
90, 1120–1124.
15. Bianchi, M.; Butti, A.; Christidis, Y.; Perronnet, J.; Barzaghi,
F.; Cesana, R.; Nencioni, A. Eur. J. Med. Chem. 1988, 23,
45–52.
16. Tokuyama, T.; Senoh, S.; Sakan, T.; Brown, K. S., Jr.; Witkop,
B. J. Am. Chem. Soc. 1967, 89, 1017–1022.
17. Rossi, F.; Garavaglia, S.; Giovenzana, G. B.; Arca, B.; Li, J.;
Rizzi, M. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 5711–5716.
18. Wilson, T. J. G.; Thomsen, K. K.; Petersen, B. O.; Duus, J. O.;
Oliver, R. P. Biochem. J. 2003, 371, 783–788.
19. Truscott, R. J. W.; Wood, A. M.; Carver, J. A.; Sheil, M. M.;
Stutchbury, G. M.; Zhu, J.; Kilby, G. W. FEBS Lett. 1994,
348, 173–176.
20. Thornalley, P. K. Biochem. J. 1991, 275, 535–539.
21. Brown, K. S., Jr. J. Am. Chem. Soc. 1965, 87, 4202–4203.
22. Matin, A.; Streete, I. M.; Jamie, I. M.; Truscott, R. J. W.; Jamie,
J. F. Anal. Biochem. 2006, 349, 96–102.
23. Hains, P. G.; Gao, L.; Truscott, R. J. W. Exp. Eye Res. 2003, 77,
547–553.
4.14. LC–MS quantification
Compounds 7–9 were analysed by LC–MS and detected in
single-ion recording (SIR) mode on a Micromass Quattro mi-
cro in ESI positive mode.43 Settings were as follows: LM1
16, HM1 15, cone 20 V, capillary 3.5 kV. Compounds
˚
were separated on a Phenomenex (Luna, 100 A, 3 mm,
2.0ꢃ150 mm, C18(2)) column. The column was equilibrated
in 95% buffer A (0.1% formic acid/H2O, v/v) and 5% buffer
B (80% CH3CN/0.1% formic acid, v/v) at a flow rate of
0.1 mL/min. The following gradient was used: 0–5 min
(5% buffer B), 5–50 min (5–50% buffer B), 50–55 min
(50% buffer B), 55–60 min (50–5% buffer B) and 60–
75 min (5% buffer B). The following masses were used for
SIR data acquisition; 194.1 [M+H]+ (8), 210.1 [M+H]+ (9),
499.19 [M+H]+ (7), 515.19 [M+H]+ (12), 192.1 [M+H]+
(10)and 208.1 [M+H]+ (11). Lens compound was quantitated
using standard curves of the synthetic samples.
24. Aquilina, J. A.; Carver, J. A.; Truscott, R. J. W. Biochemistry
1999, 38, 11455–11464.
4.15. HPLC Quantification
25. Iwahashi, H.; Ishii, T. J. Chromatogr., A 1997, 773, 23–31.
26. Dickerson, J. E., Jr.; Lou, M. F. Exp. Eye Res. 1997, 65,
451–454.
27. Aquilina, J. A.; Truscott, R. J. W. Biochim. Biophys. Acta 2002,
1596, 6–15.
28. Garner, B.; Shaw, D. C.; Lindner, R. A.; Carver, J. A.; Truscott,
R. J. W. Biochim. Biophys. Acta 2000, 1476, 265–278.
29. Stewart, A.; Augusteyn, R. C. Exp. Eye Res. 1984, 39, 307–315.
30. McNulty, R.; Wang, H.; Mathias, R. T.; Ortwerth, B. J.;
Truscott, R. J. W.; Bassnett, S. J. Physiol. (London) 2004,
559, 883–898.
31. Harding, J. J. Biochem. J. 1970, 117, 957–960.
32. Reddy, V. N.; Giblin, F. J. Ciba Found. Symp. 1984, 106, 65–87.
33. Kamei, A. Biol. Pharm. Bull. 1993, 16, 870–875.
34. Streete, I. M.; Jamie, J. F.; Truscott, R. J. W. Invest.
Ophthalmol. Vis. Sci. 2004, 45, 4091–4098.
Compounds 1, 2, 3 and 5 were quantified using RP-HPLC at
360 nm. Compounds were separated on a Microsorb (MV,
˚
100 A, 4.6ꢃ250 mm, C18) column as described above.
Standards were used as follows: synthetic 1 for the quanti-
fication of 1 and 5, commercially available 3 for 3 and
synthetic 2 for 2.
Acknowledgements
The authors are grateful to the National Health and Medical
Research Council and the Australian Research Council
(through the ARC Centres of Excellence program) for finan-
cial support.
References and notes
35. Vazquez, S.; Garner, B.; Sheil, M. M.; Truscott, R. J. W. Free
Radical Res. 2000, 32, 11–23.
36. Bolognese, A.; Liberatore, R.; Riente, G.; Scherillo, G.
J. Heterocycl. Chem. 1988, 25, 1247–1250.
1. Van Heyningen, R. Ciba Found. Symp. 1973, 151–171.
2. Dillon, J.; Skonieczna, M.; Mandal, K.; Paik, D. Photochem.
Photobiol. 1999, 69, 248–253.
37. Bolognese, A.; Bonomo, R. P.; Chillemi, R.; Sciuto, S.
J. Heterocycl. Chem. 1990, 27, 2207–2208.
3. Collier, R.; Zigman, S. Prog. Clin. Biol. Res. 1987, 247, 571–
585.
38. Butenandt, A.; Schiedt, U.; Biekert, E. Liebigs Ann. 1954, 588,
106–116.
39. Ishii, T.; Iwahashi, H.; Kido, R.; Nishioka, S. Int. Congr. Ser.
1992, 998, 159–162.
40. Aquilina, J. A.; Carver, J. A.; Truscott, R. J. W. Exp. Eye Res.
1997, 64, 727–735.
41. Goldstein, L. E.; Leopold, M. C.; Huang, X.; Atwood, C. S.;
Saunders, A. J.; Hartshorn, M.; Lim, J. T.; Faget, K. Y.;
Muffat, J. A.; Scarpa, R. C.; Chylack, L. T., Jr.; Bowden,
E. F.; Tanzi, R. E.; Bush, A. I. Biochemistry 2000, 39, 7266–
7275.
4. Zigman, S.; Paxhia, T. Exp. Eye Res. 1988, 47, 819–824.
5. Bova, L. M.; Sweeney, M. H.; Jamie, J. F.; Truscott, R. J. Invest.
Ophthalmol. Vis. Sci. 2001, 42, 200–205.
6. Bova, L. M.; Wood, A. M.; Jamie, J. F.; Truscott, R. J. Invest.
Ophthalmol. Vis. Sci. 1999, 40, 3237–3244.
7. Hood, B. D.; Garner, B.; Truscott, R. J. W. J. Biol. Chem. 1999,
274, 32547–32550.
8. Taylor, L. M.; Aquilina, J. A.; Jamie, J. F.; Truscott, R. J. W.
Exp. Eye Res. 2002, 75, 165–175.
9. Korlimbinis, A.; Truscott, R. J. W. Biochemistry 2006, 45,
1950–1960.
42. Pirie, A. Invest. Ophthalmol. 1968, 7, 634–650.
43. Ito, S.; Tsukada, K. J. Chromatogr., A 2002, 943, 39–46.
10. Taylor, L. M.; Aquilina, J. A.; Jamie, J. F.; Truscott, R. J. W.
Exp. Eye Res. 2002, 74, 503–511.