December 1999
Volume 40, Issue 13
Free
Lens  |   December 1999
UV Filter Compounds in Human Lenses: the Origin of 4-(2-Amino-3-hydroxyphenyl)-4-oxobutanoic Acid O-β-d-Glucoside
Author Affiliations
  • Lisa M. Bova
    From the Australian Cataract Research Foundation, Department of Chemistry, University of Wollongong, Wollongong, New South Wales, Australia.
  • Andrew M. Wood
    From the Australian Cataract Research Foundation, Department of Chemistry, University of Wollongong, Wollongong, New South Wales, Australia.
  • Joanne F. Jamie
    From the Australian Cataract Research Foundation, Department of Chemistry, University of Wollongong, Wollongong, New South Wales, Australia.
  • Roger J. W. Truscott
    From the Australian Cataract Research Foundation, Department of Chemistry, University of Wollongong, Wollongong, New South Wales, Australia.
Investigative Ophthalmology & Visual Science December 1999, Vol.40, 3237-3244. doi:
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Lisa M. Bova, Andrew M. Wood, Joanne F. Jamie, Roger J. W. Truscott; UV Filter Compounds in Human Lenses: the Origin of 4-(2-Amino-3-hydroxyphenyl)-4-oxobutanoic Acid O-β-d-Glucoside. Invest. Ophthalmol. Vis. Sci. 1999;40(13):3237-3244.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

purpose. To investigate UV filter synthesis in the human lens, in particular the biosynthetic origin of the second most abundant UV filter compound, 4-(2-amino-3-hydroxyphenyl)-4-oxobutanoic acid O-β-d-glucoside.

methods. Human lenses were analyzed by high-performance liquid chromatography (HPLC) after separate incubation with 3H-tryptophan (3H-Trp), β-benzoylacrylic acid, d,l-α-amino-β-benzoylpropionic acid, or d,l-3-hydroxykynurenine O-β-d-glucoside. The effect of pH on the model compound d,l-α-amino-β-benzoylpropionic acid and d,l-3-hydroxykynurenine O-β-d-glucoside was also investigated.

results. UV filters were not detected in fetal lenses, despite a 5-month postnatal lens displaying measurable levels of UV filters. In adults no radiolabel was incorporated into 4-(2-amino-3-hydroxyphenyl)-4-oxobutanoic acid O-β-d-glucoside after 3H-Trp incubations. β-Benzoylacrylic acid was readily reduced in lenses. d,l-α-Amino-β-benzoylpropionic acid and d,l-3-hydroxykynurenine O-β-d-glucoside slowly deaminated at physiological pH and were converted to β-benzoylpropionic acid and 4-(2-amino-3-hydroxyphenyl)-4-oxobutanoic acid O-β-d-glucoside, respectively, after lens incubations.

conclusions. UV filter biosynthesis appears to be activated at or near birth. Compounds containing the kynurenine side chain slowly deaminate, and in the lens, the newly formed double bond is rapidly reduced. These findings suggest that 4-(2-amino-3-hydroxyphenyl)-4-oxobutanoic acid O-β-d-glucoside is derived from l-3-hydroxykynurenine O-β-d-glucoside through this deamination-reduction process. The slowness of the deamination presumably accounts for the absence of incorporation of radiolabel from 3H-Trp into 4-(2-amino-3-hydroxyphenyl)-4-oxobutanoic acid O-β-d-glucoside.

The human lens contains a number of low-molecular-weight, fluorescent compounds, which are UV filters. They absorb most of the UV light transmitted by the cornea between 295 and 400 nm with a maximum absorption between 360 and 370 nm. The most abundant of the UV filter compounds is l-3-hydroxykynurenine O-β-d-glucoside (3OHKG). It was first identified in the early 1970s as a lens-specific product of humans and other primates 1 2 3 4 and is the major product of human lens tryptophan (Trp) metabolism, being present in higher quantities than free Trp itself. 5 It is formed through the normal catabolism pathway of Trp to kynurenine (Kyn) and 3-hydroxykynurenine (3OHKyn), which is then glucosylated to 3OHKG, 6 a conjugation unusual in mammalian systems. The unique formation of 3OHKG has inspired studies into the metabolism of Trp in human lenses. 
More recently, another UV-absorbing glucoside has been identified in human lenses. 7 8 This compound, 4-(2-amino-3-hydroxyphenyl)-4-oxobutanoic acid O-β-d-glucoside (AHBG) is the second most abundant UV filter compound of the human lens. It is similar to 3OHKG in UV absorbance profile 8 and structure, the only difference being the absence of an α-amino group. The biosynthetic pathway for the formation of AHBG, however, has not been determined. 
Although no definite function has been established for these lens-specific compounds, their UV-filtering properties 7 8 indicate a role in the protection of ocular tissue from long-wave UV radiation and/or as an aid to visual acuity by decreasing chromatic aberration. It has been proposed that the UV filter compounds may also be involved in the normal age-dependent coloration of the human lens and in crystallin modification during the development of senile nuclear cataract. 9 For proper evaluation of the role of the UV filter compounds and their involvement in the cause of aging and cataract, a greater understanding of the endogenous UV protection pathway is needed. In this study, we investigated aspects of UV filter synthesis, in particular the biosynthetic origin of AHBG within the human lens. 
Methods
Materials
l-3-Hydroxykynurenine, HEPES, trifluoroacetic acid (TFA), β-glucosidase, indole-3-propionic acid, l-Trp, and l-glutamine were purchased from Sigma (St. Louis, MO). [5-3H]-l-Trp (specific activity, 32 Ci/millimole), was purchased from Amersham Australia (Sydney). β-Benzoylacrylic acid was purchased from Acros Scientific (Geel, Belgium). d,l-α-Amino-β-benzoylpropionic acid was synthesized according to the method previously described by Cerani and Tarzia. 10 d,l-3-Hydroxykynurenine O-β-d-glucoside and 4-(2-amino-3-hydroxyphenyl)-4-oxobutanoic acid O-β-d-glucoside were synthesized by the method of Manthey et al. 11  
Normal lenses were obtained from donor eyes at The Sydney Eye Hospital Lions Eye Bank. Fetal lenses were obtained from abortions performed at Westmead Hospital, Sydney. After removal, lenses were immediately placed into 5-ml artificial aqueous humor (AAH) in 30-ml sterile plastic screw-capped vials. The vials were kept at 35°C until transported to the laboratory, usually within 24 hours. Lenses not incubated in AAH were preserved in liquid nitrogen. The AAH 12 consisted of Eagle’s minimum essential medium (EMEM; Auto-Pow version; ICN Biomedicals, Sydney, Australia) supplemented with 50 μM l-Trp, 10 mM HEPES, 2.0 mM l-glutamine, and 200 mg/ml streptomycin sulfate/200 IU penicillin G (Boehringer–Mannheim, Sydney, Australia). The pH was adjusted to 7.4 with 1 M NaOH. Lenses were incubated individually in 35-mm sterile plastic petri dishes (Corning, NY). 
Instrumental Conditions
The high-performance liquid chromatography (HPLC) system consisted of a pump (model K35D; ICI, GBC, Sydney, Australia), a sample injector (model 7125; Rheodyne, Cotati, CA) fitted with a 100-μl sample loop and a Knauer (Berlin-Zehlendorf, Germany) variable wavelength UV detector. Chromatograms were recorded and peak areas integrated on an integrator (Chromatopac CR6A; Shimadzu, Columbia, MD). The UV filter compounds from the AAH and protein-free lens extracts and the d,l-3-hydroxykynurenine O-β-d-glucoside pH studies were analyzed on either a 250 × 4.6-mm C18 reversed-phase column (Spherisorb S5ODS2; Activon, Sydney, Australia) or on a 250 × 4.6-mm C18 reversed-phase column (Microsorb; Varian, Sunnyvale, CA), using the same solvent conditions as previously described. 8 A radiochromatography detector (Flo-One Beta A-100; Radiomatic, Tampa, FL) with a 500-μl flow cell (TR-LSC; Canberra Packard, Canberra, Australia) was used to detect and integrate peaks of radioactivity after tritiation experiments. A fluorescence detector (LC1250; ICI) was used to detect peaks after the d,l-3-hydroxykynurenine O-β-d-glucoside incubation studies. Analyses of the β-benzoylacrylic acid and d,l-α-amino-β-benzoylpropionic acid experiments were performed on the Varian Microsorb C18 reversed-phase column. A mobile phase of 25 mM phosphate buffer (pH 6.8) in 10% acetonitrile was used, with a flow rate of 0.6 ml/min for the β-benzoylacrylic acid lens incubation experiments and a flow rate of 1.0 ml/min for the d,l-α-amino-β-benzoylpropionic acid experiments. The peaks were detected at 250 nm. Electrospray ionization mass spectrometry was performed on a mass spectrometer with a hexapole collision cell (Quattro; VG Biotech, Altrincham, UK). UV spectra were recorded on a recording spectrophotometer (UV-265; Shimadzu). 
Lens Experiments
Analysis of UV Filters.
Lenses were removed from the AAH or thawed if preserved in liquid nitrogen, and the protein-free lens extracts were obtained and analyzed by high-performance liquid chromatography (HPLC), following the conditions previously described. 8  
Tritiated Trp Experiments.
Lenses were removed from the AAH in which they had been transported and stabilized separately overnight in fresh AAH (4.5 ml) at 35°C. The AAH was changed again at the start of the experiment, and[ 5-3H]-l-Trp was added to each medium to obtain a final activity of 2.0 μCi/ml. After 24 hours, lenses were removed from the culture and rinsed twice for 10 seconds in unlabeled AAH to remove any label adhering to the capsule. The protein-free lens extracts were obtained for each lens and analyzed by HPLC. 8 Lenses not extracted immediately were frozen at− 20°C until needed. 
Model Compounds.
The lenses from a 64-year-old donor were incubated separately at 35°C in AAH (10 ml) containing 1 mM β-benzoylacrylic acid and 0.05% ethanol (to aid dissolution of β-benzoylacrylic acid). One lens was incubated for 24 hours and the other for 48 hours. The lenses and AAH were separated, and both the AAH and protein-free extracts of each lens were examined by HPLC. The peaks at 24 minutes (retention time ofβ -benzoylpropionic acid) for both protein-free lens extracts were collected, acidified with dilute HCl, extracted with ether, and evaporated to dryness under argon. As a control, a solution of the AAH (10 ml) containing 1 mM β-benzoylacrylic acid and 0.05% ethanol was incubated at 35°C for 48 hours and analyzed by HPLC. 
A 27-year-old lens was incubated at 35°C in a 1-mM solution of d,l-α-amino-β-benzoylpropionic acid in AAH (5 ml) for 48 hours, and the AAH and protein-free lens extract were examined by HPLC. The peak at 17 minutes (retention time of β-benzoylpropionic acid) in the lens extract and peaks at 17 and 20 minutes (retention time of β-benzoylacrylic acid) in the AAH were collected and extracted as described. As a control, a 1-mM solution of d,l-α-amino-β-benzoylpropionic acid in AAH (5 ml) was incubated for 48 hours at 37°C and analyzed by HPLC. 
d,l-3-Hydroxykynurenine O-β-d-Glucoside Lens Incubation.
A 26-year-old lens was incubated at 35°C in AAH (6 ml), and aliquots (100 μl) were taken every 2 hours and snap frozen in liquid nitrogen. After 8 hours, d,l-3-hydroxykynurenine O-β-d-glucoside was added to the AAH (final concentration, 5 mM), and the incubation was continued for another 40 hours with aliquots collected as described approximately every 2 to 4 hours. The lens was separated from the AAH and the protein-free extract obtained. The protein-free extract and AAH aliquots were analyzed by HPLC for the presence of d,l-3-hydroxykynurenine O-β-d-glucoside and 4-(2-amino-3-hydroxyphenyl)-4-oxobutanoic acid O-β-d-glucoside. As a control, the contralateral lens was incubated at 35°C in AAH (6 ml) for 48 hours and the protein-free lens extract analyzed by HPLC. 
pH Stability of d,l-α-Amino-β-benzoylpropionic Acid
d,l-α-Amino-β-benzoylpropionic acid (2 mg/ml) was incubated at 37°C in 25 mM phosphate buffer (pH 6, 7, and 8) and 25 mM carbonate buffer (pH 9 and 10). Duplicate aliquots (100 μl) were taken at various intervals over a period of 25 hours. The aliquots were diluted (1:1) with 25 mM phosphate buffer (pH 6.8) and analyzed by HPLC. For confirmation of structure, the peak at approximately 20 minutes (retention time of β-benzoylacrylic acid) was collected from the pH 7 experiment, extracted, and dried as described. 
pH Stability of d,l-3-Hydroxykynurenine O-β-d-Glucoside
d,l-3-Hydroxykynurenine O-β-d-glucoside (0.22 mg/ml) was incubated at 37°C in 25 mM carbonate buffer (pH 9) for 24 hours. Duplicate aliquots (100 μl) were taken at t = 0 minutes, 2 hours, and 24 hours, diluted (1:1) with 25 mM phosphate buffer (pH 6.8) and analyzed by HPLC. The peak at 47 minutes’ retention was collected from the 24-hour aliquot, freeze dried, desalted with a C18 Sep-Pak (Waters, Milford, MA), and lyophilized. d,l-3-Hydroxykynurenine O-β-d-glucoside (1 mg/ml) was also incubated at 37°C with 25 mM phosphate buffer (pH 7) for 7 days. Duplicate aliquots (100 μl) were taken once a day and analyzed by HPLC as described. 
Results
Fetal Lenses
Fetal lenses were examined to determine whether the pathway leading to the formation of the UV filter compounds is active before birth. Seven fetal lenses, ranging in age from 13 to 20 weeks, were pooled and the protein-free extracts analyzed by HPLC. No Trp metabolites were detected, despite the presence of Trp itself. In contrast, a 5-month-old postnatal lens was found to contain 1.7 micromoles 3OHKG/g lens wet weight—a concentration in the upper range based on previous studies of this compound. 5 This lens also contained high levels of Kyn and 3OHKyn. 
Radiolabeled Trp incubation experiments performed over 24 hours on individual fetal lenses also showed no incorporation of radiolabel into UV filters. Because there was concern that the apparent absence of incorporation may have been due to the small size of fetal lenses, three fetal lenses aged 12, 13, and 17 weeks were incubated together in AAH containing tritiated Trp (3H-Trp). The incubation was maintained for 72 hours to maximize incorporation of label. Analysis of the protein-free extract of the lenses showed that, although radiolabeled Trp entered the lenses, it was not metabolized into the UV filter compounds in any detectable level. Integration of the 3H-Trp peak produced a value of 16,000 disintegrations per minute, a quantity of radiolabel more than adequate to detect metabolism to 3OHKG at typical incorporation levels found in adult lenses. 
Incorporation of Radiolabel from 3H-Trp into UV Filters
Eleven adult human lenses were incubated separately with 3H-Trp for 24 hours at 35°C, and the UV filters present in the protein-free extracts were analyzed by HPLC. A typical HPLC chromatogram is shown in Figures 1A and 1B for a 21-year-old lens. Both online radiochemical detection and UV absorption were used and the profiles plotted simultaneously. At the commencement of the analysis, the detection wavelength was set at 365 nm. The sharp line at 30 minutes on the UV trace of Figure 1A (top) indicates the alteration of the detection wavelength to 278 nm, to detect Trp, which eluted between 60 and 70 minutes. Using elution with acetate buffer, three 365-nm absorbing peaks were detected. Each of these UV peaks also corresponded to peaks of radioactivity (Fig. 1A , bottom). The small peaks running at 13 minutes and 27 minutes have been previously identified as 3OHKyn and Kyn, respectively, whereas the large peak at 15 minutes has been identified as 3OHKG. 5 Electrospray mass spectroscopy was used to confirm the identities of these peaks. 
Integration of the radiolabeled peaks (Fig. 1A) showed that 74% of the total radioactivity was due to the substrate, 3H-Trp. The largest incorporation of label (as a percentage of total radioactivity detected in the protein-free lens extracts) was into 3OHKG (14%), followed by Kyn (6%), and 3OHKyn (1%). The remaining 5% of label was due to uncharacterized products, including possible autolytic breakdown products that chromatograph with the solvent front. The 10 other lens experiments showed incorporation of radiolabel into 3OHKG, with levels ranging from 5% to 21%. 
After elution of Trp (Fig. 1A) , a return in the detection wavelength to 365 nm and replacement of the acetate buffer mobile phase with 20% methanol resulted in the detection of two additional long-wave UV-absorbing peaks, with retention times of 21 and 25 minutes (Fig. 1B) . The earlier eluting compound has previously been identified as AHBG, 8 whereas the structure of the latter peak is not yet known. No radiolabel coincided with the AHBG peak (or the unknown peak) in the analysis shown in Figure 1B , nor was any label observed in the other 10 lens experiments performed over 24 hours, or after longer periods of incubation with 3H-Trp (up to 65 hours) in two further lens experiments. A peak with radiolabel was observed at 19 minutes (Fig. 1B) , which did not correspond to the elution position of AHBG. This peak was not consistently observed in the other lens experiments. Incubation of lenses with tritiated 3OHKyn, as described previously, 6 also demonstrated incorporation of label into 3OHKG, but not into AHBG. 
The integrated UV peak areas of 3OHKG and AHBG from the lenses used in the radiolabeling studies and from 18 other lenses were used to estimate quantities of these two UV filter compounds. The lenses varied in age ranging from 21 to 84 years. Lens pairs showed little variation in 3OHKG and AHBG levels and in any given lens, the quantity of 3OHKG was always greater than that of AHBG (Fig. 2) . Four lenses with very low levels of 3OHKG (one each 71- and 83-year-old lenses and a pair of 84-year-old lenses), also had the lowest levels of AHBG. One 64-year-old lens, with an exceptionally high level of 3OHKG, also possessed the highest quantity of AHBG. No correlation was observed between the levels of AHBG and age. 
Origin of AHBG
Given the absence of 3H-Trp incorporation into AHBG, a parallel biosynthetic pathway to that of the other known UV filters, using indole-3-propionic acid rather than Trp as the biosynthetic precursor, was investigated. The protein-free extract of a 70-year-old lens was examined for the presence of this acid; however, none was observed by HPLC. In contrast, the precursors of 3OHKG, that is, Trp, Kyn, and 3OHKyn, were all detected, suggesting that the indole-3-propionic acid pathway is not operating in the human lens. 
Because the structures of AHBG and 3OHKG are very similar and there was a correlation between the levels of these UV filters in the human lens, it seemed feasible that AHBG could in fact be derived from 3OHKG despite the absence of 3H-Trp incorporation. A possible pathway for this involves the elimination of ammonia (deamination) from 3OHKG to produce the correspondingα ,β-ketoalkene, followed by reduction, possibly enzymatically, of the α,β-ketoalkene to AHBG (Fig. 3) . The feasibility of this mechanism was therefore investigated using the model compounds d,l-α-amino-β-benzoylpropionic acid, β-benzoylacrylic acid, and β-benzoylpropionic acid (Fig. 4)
Model Studies for Reduction
To examine the proposed reduction step, a pair of 64-year-old lenses were incubated with β-benzoylacrylic acid (1 mM) in AAH at 35°C. One lens was incubated for 24 hours and the other for 48 hours. HPLC analysis of the AAH showed peaks with retention times consistent with authentic samples of the reduced compound β-benzoylpropionic acid (approximately 24 minutes) and β-benzoylacrylic acid (approximately 26 minutes; Fig. 5A ). HPLC analysis of the protein-free extract from the 24-hour lens experiment (Fig. 5B) also showed a peak consistent withβ -benzoylpropionic acid along with β-benzoylacrylic acid, whereas the extract from the 48-hour lens experiment (Fig. 5C) showedβ -benzoylpropionic acid with no detectable β-benzoylacrylic acid. Analysis by electrospray mass spectrometry of the 24-minute peak from each lens extract confirmed that reduction had taken place by displaying M + 1 (m/z 179) and M − 1 (m/z 177) molecular ion peaks for β-benzoylpropionic acid. 
In contrast, no peak was observed for β-benzoylpropionic acid in a control experiment in which β-benzoylacrylic (1 mM) was incubated with AAH for 48 hours at 35°C in the absence of a lens (Fig. 5D) . Quantification studies showed that the amount of reduction ofβ -benzoylacrylic acid to β-benzoylpropionic acid was approximately 70% in the media of the 24-hour lens incubation study and approximately 60% in the media of the 48-hour incubation study. In the lenses, approximately 80% of β-benzoylacrylic acid had been reduced after 24 hours, and virtually 100% reduction had occurred after 48 hours. 
Model Studies for Deamination
d,l-α-Amino-β-benzoylpropionic acid was incubated at pH 6, 7, 8, 9, and 10 at 37°C and monitored by HPLC over a period of 25 hours for the formation of β-benzoylacrylic acid. The identity of β-benzoylacrylic acid was confirmed by mass spectrometry of the HPLC peak. The deamination was optimal at pH 9, with approximately 74% conversion of the amino acid to β-benzoylacrylic acid, whereas a 12% conversion was observed at pH 7 and 33% conversion at pH 8 (Fig. 6)
To determine whether this deamination also occurred in human lenses, a 27-year-old lens was incubated with d,l-α-amino-β-benzoylpropionic acid (1 mM) in AAH for 48 hours at 35°C. HPLC analysis of the AAH displayed peaks at approximately 10, 17, and 20 minutes, consistent with the retention times of d,l-α-amino-β-benzoylpropionic acid,β -benzoylpropionic acid, and β-benzoylacrylic acid, respectively (Fig. 7A ). Analysis of the protein-free lens extract displayed peaks consistent with d,l-α-amino-β-benzoylpropionic acid and the reduced compound β-benzoylpropionic acid (Fig. 7B) . Noβ -benzoylacrylic acid was detected. Electrospray mass spectrometry of the collected peaks confirmed the identity of β-benzoylpropionic acid in the AAH and protein-free lens extract and the identity ofβ -benzoylacrylic acid in the lens media. A control in which a 1-mM solution of d,l-α-amino-β-benzoylpropionic acid in AAH was incubated at 35°C for 48 hours also displayed a peak forβ -benzoylacrylic acid; however, no peak was observed for the reduced compound β-benzoylpropionic acid (Fig. 7C) . Quantification studies showed that in both the lens AAH and the control AAH approximately 35% of d,l-α-amino-β-benzoylpropionic acid had undergone deamination and that in the lens AAH approximately 85% ofβ -benzoylacrylic acid was reduced after 48 hours’ incubation, whereas the amount of reduction within the lens itself was virtually 100%. 
Deamination of 3OHKG
Now that the feasibility of the deamination process at physiological pH had been established with the model compound, deamination of d,l-3OHKG 11 was investigated. d,l-3OHKG was not examined earlier as, unlike the model compound, it is only available in small quantities. To maximize deamination and allow characterization, 3OHKG was initially incubated at pH 9 at 37°C for 24 hours. A new peak was detected after 2 hours by HPLC. After 24 hours this peak had doubled in size, with an integrated peak area approximately 5% that of 3OHKG. Analysis of the peak by electrospray mass spectrometry confirmed that deamination had taken place by displaying a peak at m/z 370, consistent with the M + 1 molecular ion peak of the deaminated product of 3OHKG. 3OHKG was then incubated at physiological pH (pH 7) for 7 days at 37°C and levels of 3OHKG and the deaminated product monitored every 24 hours by HPLC (Fig. 8) . 3OHKG levels were found to decrease uniformly over the 7 days, with an overall change of approximately 2.5%. Over the first 24 hours none of the deaminated product was detected; however, after 55 hours it was observed with an integrated peak area approximately 1% that of 3OHKG. The level of the deaminated product did not change significantly subsequent to this, but a number of new peaks with lower retention times were observed after 75 hours. The formation of other species was not surprising, because the newly formed double bond of the deaminated product would be susceptible to nucleophilic attack. Compared with that of the model compound d,l-α-amino-β-benzoylpropionic acid, deamination of 3OHKG was significantly slower. 
Conversion of d,l-3OHKG to AHBG
Because d,l-3OHKG can deaminate at physiological pH, and the lens is capable of reducing α,β-ketoalkenes, the lenticular conversion of 3OHKG to AHBG was examined directly. A 26-year-old lens was incubated in AAH at 35°C. After 8 hours the AAH was spiked with d,l-3OHKG to achieve a final concentration of 5 mM, and the incubation was continued for another 40 hours. Aliquots from the AAH were taken at regular time intervals and examined by HPLC. Previous research has shown that UV filters accumulate in the AAH when lenses are incubated. 5 Analysis of human vitreous indicates that this efflux also occurs in vivo. 5 During the first 8 hours, 3OHKG was observed at very low levels in the AAH, whereas none of the other UV filters, including AHBG or the deaminated product, was detected. AHBG was observed 2 hours after the spiking, with an integrated peak area approximately 7% that of 3OHKG. A corresponding decrease in 3OHKG was also observed. The level of AHBG appeared to increase slowly over the remainder of the experiment, as did the level of 3OHKG (Fig. 9) . As a control, the contralateral lens was incubated in AAH in the absence of added 3OHKG at 35°C for 48 hours. Analysis of the protein-free lens extract of both the spiked lens and the contralateral lens showed that the spiked lens had significantly higher levels of 3OHKG (approximately 1.6 times higher), whereas AHBG levels were comparable. 
Discussion
A key role of Trp in the human lens, apart from incorporation into lens proteins, is to form UV filter compounds. There is, however, still much to understand about the mechanism of formation and regulation of the Trp metabolic pathway in the human lens. None of the UV filter compounds was detected in fetal lenses aged up to 20 weeks, nor could their biosynthesis be demonstrated in these lenses. High levels of Kyn, 3OHKyn, and 3OHKG, however, were detected in a 5-month postnatal lens. This suggests that the UV filter biosynthetic pathway in the human lens is activated sometime between late pregnancy and birth. Such a finding seems reasonable in relation to the probable role of these compounds as protective UV filter species or as aids to visual acuity. The biochemical trigger for activation of the kynurenine pathway is unknown but may be linked to the induction of the first enzyme in this pathway, indoleamine 2,3-dioxygenase, by a compound such as γ-interferon, as described for other cells. 13  
The second most abundant UV filter compound, AHBG, differs from 3OHKG only in the absence of an α-amino group. This similarity suggests that AHBG may be a further metabolic product of 3OHKG, yet no incorporation of radiolabel into AHBG using either radiolabeled Trp or 3OHKyn 6 could be shown. In these experiments significant amounts of radiolabel were incorporated into 3OHKG, thus providing evidence that AHBG was not formed artifactually during isolation or extraction. AHBG is almost certainly the unidentified glucoside isolated by Van Heyningen, 4 who also found no incorporation of radiolabel from Trp into the unknown glucoside, although label was detected in Kyn and 3OHKG. 
Indole-3-propionic acid was not detected in a lens extract, whereas the precursors of 3OHKG (i.e., Trp, Kyn, and 3OHKyn) were all detected. This suggests that a parallel pathway to the kynurenine pathway, using a precursor without the α-amino group, was not operating. There did, however, appear to be a correlation between the levels of 3OHKG and AHBG in the lens. All this evidence led us to speculate that AHBG was derived from 3OHKG. 
Deamination of 3OHKG to produce an α,β-ketoalkene, followed by reduction, was proposed as a possible route to AHBG. Deaminations from systems similar to that of 3OHKG have been reported to occur in a facile manner at high pH, 14 whereas enzymatic elimination of ammonia from phenylalanine to produce trans-cinnamic acid is known to occur in higher plants. 15 No previous studies have examined the reduction of α,β-ketoalkenes by lenses. Studies on bovine ocular tissues, however, have confirmed the presence of reductases capable of reducing the α,β-ketoalkene trans-phenyl-1-propenyl ketone in the presence of reduced nicotinamide adenine dinucleotide phosphate (NADPH) or reduced nicotinamide adenine dinucleotide (NADH). 16 Alkene reductase enzymes of this type are also present in humans. For example, the transformation of 7-dehydrocholesterol to cholesterol is catalyzed by 7-dehydrocholesterol-delta7-reductase. 17  
The model compound β-benzoylacrylic acid was readily reduced toβ -benzoylpropionic acid when incubated with lenses in AAH. Because reduction was not observed in the absence of a lens, this reduction was clearly a result of lens activity and not due to any species within the AAH. Given the similarity of β-benzoylacrylic acid with the proposed elimination product of 3OHKG, it can be concluded that the reduction of this alkene to form AHBG within the lens is feasible. The details of the reduction were not examined further. 
Because only small amounts of 3OHKG are available, we initially investigated deamination of the model compound d,l-α-amino-β-benzoylpropionic acid, which has the same amino acid side chain as 3OHKG. The model compound was found to deaminate over a range of pHs, albeit slowly at physiological pH and maximally at pH 9. Lens incubation studies with the amino acid showed that the rate of deamination at physiological pH in the presence and absence of the lens was similar, confirming the nonenzymatic nature of the deamination. Furthermore, extension of these studies to d,l-3OHKG showed that the proposed deamination can occur at physiological pH, although this process was slower than that for d,l-α-amino-β-benzoylpropionic acid. The rate of reduction was found to be significantly faster than the rate of deamination, suggesting that the elimination of ammonia from 3OHKG may be the rate-limiting step in the formation of AHBG. 
d,l-3OHKG was found to yield AHBG after incubation of a lens with d,l-3OHKG; however, the amount of conversion was low. This presumably reflects a low rate of deamination as well as other processes that may be occurring in the lens, such as conjugation of the double bond with glutathione. 18 It may well be that AHBG formation occurs largely in the lens nucleus where glutathione is low and the effective residence time of 3OHKG is longer, allowing deamination to take place. If so, these factors make lens experiments of AHBG formation difficult to undertake. 
The 3OHKG lens incubation study and the model studies provide strong support for the notion that AHBG is derived from the major UV filter compound, 3OHKG, through a slow nonenzymatic elimination of ammonia to produce the corresponding α,β-ketoalkene, followed by reduction of the side-chain double bond. The slowness of the deamination presumably accounts for the absence of radiolabeling of AHBG after the lens incubations with 3H-Trp. These studies also show the intrinsic instability of compounds with the kynurenine side chain, which may have important implications for the human lens. 
 
Figure 1.
 
(A) HPLC analysis using acetate buffer (pH 4.5) of the protein-free extract of a 21-year-old human lens, after incubation of the lens in 3H-Trp for 24 hours. The UV trace and radiolabel detector traces are shown. UV detection was at 365 nm until 30 minutes, when the wavelength was altered to 278 nm. The large radioactive peak running between 61 and 67 minutes is 3H-Trp. (B) Continued isocratic HPLC analysis of the same lens as in (A) after changing the mobile phase to 20% methanol. The UV detection wavelength was 365 nm. The two peaks detected at 21 and 25 minutes are AHBG and an unknown compound, respectively.
Figure 1.
 
(A) HPLC analysis using acetate buffer (pH 4.5) of the protein-free extract of a 21-year-old human lens, after incubation of the lens in 3H-Trp for 24 hours. The UV trace and radiolabel detector traces are shown. UV detection was at 365 nm until 30 minutes, when the wavelength was altered to 278 nm. The large radioactive peak running between 61 and 67 minutes is 3H-Trp. (B) Continued isocratic HPLC analysis of the same lens as in (A) after changing the mobile phase to 20% methanol. The UV detection wavelength was 365 nm. The two peaks detected at 21 and 25 minutes are AHBG and an unknown compound, respectively.
Figure 2.
 
The quantity of 3OHKG plotted against the quantity of AHBG in 29 human lenses ranging in age from 21 to 84 years.
Figure 2.
 
The quantity of 3OHKG plotted against the quantity of AHBG in 29 human lenses ranging in age from 21 to 84 years.
Figure 3.
 
Proposed formation of AHBG through deamination of 3OHKG to produceα ,β-ketoalkene, followed by reduction of the side-chain double bond.
Figure 3.
 
Proposed formation of AHBG through deamination of 3OHKG to produceα ,β-ketoalkene, followed by reduction of the side-chain double bond.
Figure 4.
 
d,l-α-Amino-β-benzoylpropionic acid, β-benzoylacrylic acid, and β-benzoylpropionic acid, the model compounds used to test the deamination and reduction processes proposed for the formation of AHBG.
Figure 4.
 
d,l-α-Amino-β-benzoylpropionic acid, β-benzoylacrylic acid, and β-benzoylpropionic acid, the model compounds used to test the deamination and reduction processes proposed for the formation of AHBG.
Figure 5.
 
HPLC analysis of a pair of 64-year-old human lenses after separate incubation with 1 mM β-benzoylacrylic acid. UV detection at 250 nm. The peak at approximately 24 minutes is the reduced compoundβ -benzoylpropionic acid, and the peak at approximately 26 minutes isβ -benzoylacrylic acid. The peaks between 5 and 15 minutes were due to compounds present in the original AAH. (A) AAH after 48 hours lens incubation in the presence of β-benzoylacrylic acid. (B) Protein-free extract after 24 hours lens incubation in the presence of β-benzoylacrylic acid. (C) Protein-free extract after 48 hours lens incubation in the presence ofβ -benzoylacrylic acid. (D) Control. AAH after 48 hours’ incubation of 1 mM β-benzoylacrylic acid in the absence of a lens.
Figure 5.
 
HPLC analysis of a pair of 64-year-old human lenses after separate incubation with 1 mM β-benzoylacrylic acid. UV detection at 250 nm. The peak at approximately 24 minutes is the reduced compoundβ -benzoylpropionic acid, and the peak at approximately 26 minutes isβ -benzoylacrylic acid. The peaks between 5 and 15 minutes were due to compounds present in the original AAH. (A) AAH after 48 hours lens incubation in the presence of β-benzoylacrylic acid. (B) Protein-free extract after 24 hours lens incubation in the presence of β-benzoylacrylic acid. (C) Protein-free extract after 48 hours lens incubation in the presence ofβ -benzoylacrylic acid. (D) Control. AAH after 48 hours’ incubation of 1 mM β-benzoylacrylic acid in the absence of a lens.
Figure 6.
 
The percentage conversion of d,l-α-amino-β-benzoylpropionic acid to the deaminated product β-benzoylacrylic acid over 25 hours at pH 6, 7, 8, 9, and 10 at 37°C.
Figure 6.
 
The percentage conversion of d,l-α-amino-β-benzoylpropionic acid to the deaminated product β-benzoylacrylic acid over 25 hours at pH 6, 7, 8, 9, and 10 at 37°C.
Figure 7.
 
HPLC analysis of a 27-year-old human lens after incubation with 1 mM d,l-α-amino-β-benzoylpropionic acid for 48 hours. UV detection at 250 nm. The peak at approximately 10 minutes is d,l-α-amino-β-benzoylpropionic acid, at approximately 17 minutes is β-benzoylpropionic acid, and at approximately 20 minutes is β-benzoylacrylic acid. (A) AAH after incubation of lens in the presence of d,l-α-amino-β-benzoylpropionic acid. (B) Protein-free extract after incubation of lens in the presence of d,l-α-amino-β-benzoylpropionic acid. (C) Control. AAH after 48 hours’ incubation of 1 mM d,l-α-amino-β-benzoylpropionic acid in the absence of a lens.
Figure 7.
 
HPLC analysis of a 27-year-old human lens after incubation with 1 mM d,l-α-amino-β-benzoylpropionic acid for 48 hours. UV detection at 250 nm. The peak at approximately 10 minutes is d,l-α-amino-β-benzoylpropionic acid, at approximately 17 minutes is β-benzoylpropionic acid, and at approximately 20 minutes is β-benzoylacrylic acid. (A) AAH after incubation of lens in the presence of d,l-α-amino-β-benzoylpropionic acid. (B) Protein-free extract after incubation of lens in the presence of d,l-α-amino-β-benzoylpropionic acid. (C) Control. AAH after 48 hours’ incubation of 1 mM d,l-α-amino-β-benzoylpropionic acid in the absence of a lens.
Figure 8.
 
Quantity of 3OHKG and the deaminated product plotted against time (hours) after incubation of 3OHKG (1 mg/ml) at pH 7 and 37°C for a period of 7 days. HPLC peaks detected using fluorescence. Concentrations of 3OHKG and the deaminated product were determined using a 3OHKG standard curve.
Figure 8.
 
Quantity of 3OHKG and the deaminated product plotted against time (hours) after incubation of 3OHKG (1 mg/ml) at pH 7 and 37°C for a period of 7 days. HPLC peaks detected using fluorescence. Concentrations of 3OHKG and the deaminated product were determined using a 3OHKG standard curve.
Figure 9.
 
Quantity of 3OHKG and AHBG found in the AAH plotted against time (hours) after incubation of a 26-year-old lens in AAH at 35°C and subsequent spiking of the AAH at 8 hours with 3OHKG (final concentration 5 mM). HPLC peaks detected using fluorescence.
Figure 9.
 
Quantity of 3OHKG and AHBG found in the AAH plotted against time (hours) after incubation of a 26-year-old lens in AAH at 35°C and subsequent spiking of the AAH at 8 hours with 3OHKG (final concentration 5 mM). HPLC peaks detected using fluorescence.
The authors thank Philip Penfold, Jan Provis, Michelle Madigan, Tania Balind, Claudia Diaz–Araya, Raj Devasahayam, and Frank Billson of The Sydney University Department of Ophthalmology for providing lenses from the Sydney Eye Hospital Lions Eye Bank and fetal lenses from Westmead Hospital, Sydney. 
van Heyningen R. Fluorescent derivatives of 3-hydroxy-l-kynurenine in the lens of man, the baboon and the grey squirrel. Biochem J. 1971a;123:30P–31P.
van Heyningen R. Fluorescent glucoside in the human lens. Nature. 1971b;230:393–394. [CrossRef]
van Heyningen R. Assay of fluorescent glucosides in the human lens. Exp Eye Res. 1973a;15:121–126. [CrossRef]
van Heyningen R. The glucoside of 3-hydroxykynurenine and other fluorescent compounds in the human lens. CIBA Found Symp. 1973;19:151–171.
Wood A, Truscott R. UV filters in human lenses: tryptophan catabolism. Exp Eye Res. 1993;56:317–325. [CrossRef] [PubMed]
Wood AM, Truscott RJW. Ultraviolet filter compounds in human lenses: 3-hydroxykynurenine glucoside formation. Vision Res. 1994;34:1369–1374. [CrossRef] [PubMed]
Inoue A, Satoh K. Identification of a fluorescent glucoside isolated from the protein-free extract of human lens. Bio Med Chem Lett. 1994;4:2303–2306. [CrossRef]
Truscott RJW, Wood AM, Carver JA, et al. A new UV-filter compound in human lenses. FEBS Lett. 1994;348:173–176. [CrossRef] [PubMed]
Stutchbury GM, Truscott RJW. The modification of proteins by 3-hydroxykynurenine. Exp Eye Res. 1993;57:149–155. [CrossRef] [PubMed]
Ceriani S, Tarzia G. Simple synthesis of Dl-β-amino-β-aroylpropionic acids. Ann Chim. 1973;63:457–466.
Manthey MK, Jamie JF, Truscott RJW. Synthesis of human ultraviolet filter compounds: O-β-d-glucopyranosides of 3-hydroxykynurenine and 2-amino-3-hydroxy-γ-oxobenzenebutanoic acid. J Org Chem.. 1999;64:3930–3933. [CrossRef]
Marcantonio J, Duncan G. Amino acid transport and protein synthesis in human normal and cataractous lenses. Curr Eye Res. 1987;6:1299–1308. [CrossRef] [PubMed]
Takikawa O, Kuroiwa T, Yamazaki F, et al. Mechanism of interferon gamma action. J Biol Chem. 1988;263:2041–2048. [PubMed]
Tokuyama T, Senoh S, Hirace YX, et al. Kynurenine yellow. J Chem Soc. Jpn.. 1958;179:752–761.
Metzler DE. Biochemistry the Chemical Reactions of Living Cells. 1977;463–464. Academic Press New York.
Takeda Y, Mishima HK, Kitamura S, et al. Purification and characterisation of alpha,beta-ketoalkene double bond reductases from bovine eyes. Curr Eye Res. 1997;16:327–332. [CrossRef] [PubMed]
Shefer S, Salen G, Batta AK, et al. Markedly inhibited 7-dehydrocholesterol-delta7-reductase activity in liver microsomes from Smith-Lemli-Opitz homozygotes. J Clin Invest. 1995;96:1779–1785. [CrossRef] [PubMed]
Garner B, Vazquez S, Griffith R, et al. Identification of glutathionyl-3-hydroxykynurenine glucoside as a novel fluorophore associated with aging of the human lens. J Biol Chem. 1999;274:20847–20854. [CrossRef] [PubMed]
Figure 1.
 
(A) HPLC analysis using acetate buffer (pH 4.5) of the protein-free extract of a 21-year-old human lens, after incubation of the lens in 3H-Trp for 24 hours. The UV trace and radiolabel detector traces are shown. UV detection was at 365 nm until 30 minutes, when the wavelength was altered to 278 nm. The large radioactive peak running between 61 and 67 minutes is 3H-Trp. (B) Continued isocratic HPLC analysis of the same lens as in (A) after changing the mobile phase to 20% methanol. The UV detection wavelength was 365 nm. The two peaks detected at 21 and 25 minutes are AHBG and an unknown compound, respectively.
Figure 1.
 
(A) HPLC analysis using acetate buffer (pH 4.5) of the protein-free extract of a 21-year-old human lens, after incubation of the lens in 3H-Trp for 24 hours. The UV trace and radiolabel detector traces are shown. UV detection was at 365 nm until 30 minutes, when the wavelength was altered to 278 nm. The large radioactive peak running between 61 and 67 minutes is 3H-Trp. (B) Continued isocratic HPLC analysis of the same lens as in (A) after changing the mobile phase to 20% methanol. The UV detection wavelength was 365 nm. The two peaks detected at 21 and 25 minutes are AHBG and an unknown compound, respectively.
Figure 2.
 
The quantity of 3OHKG plotted against the quantity of AHBG in 29 human lenses ranging in age from 21 to 84 years.
Figure 2.
 
The quantity of 3OHKG plotted against the quantity of AHBG in 29 human lenses ranging in age from 21 to 84 years.
Figure 3.
 
Proposed formation of AHBG through deamination of 3OHKG to produceα ,β-ketoalkene, followed by reduction of the side-chain double bond.
Figure 3.
 
Proposed formation of AHBG through deamination of 3OHKG to produceα ,β-ketoalkene, followed by reduction of the side-chain double bond.
Figure 4.
 
d,l-α-Amino-β-benzoylpropionic acid, β-benzoylacrylic acid, and β-benzoylpropionic acid, the model compounds used to test the deamination and reduction processes proposed for the formation of AHBG.
Figure 4.
 
d,l-α-Amino-β-benzoylpropionic acid, β-benzoylacrylic acid, and β-benzoylpropionic acid, the model compounds used to test the deamination and reduction processes proposed for the formation of AHBG.
Figure 5.
 
HPLC analysis of a pair of 64-year-old human lenses after separate incubation with 1 mM β-benzoylacrylic acid. UV detection at 250 nm. The peak at approximately 24 minutes is the reduced compoundβ -benzoylpropionic acid, and the peak at approximately 26 minutes isβ -benzoylacrylic acid. The peaks between 5 and 15 minutes were due to compounds present in the original AAH. (A) AAH after 48 hours lens incubation in the presence of β-benzoylacrylic acid. (B) Protein-free extract after 24 hours lens incubation in the presence of β-benzoylacrylic acid. (C) Protein-free extract after 48 hours lens incubation in the presence ofβ -benzoylacrylic acid. (D) Control. AAH after 48 hours’ incubation of 1 mM β-benzoylacrylic acid in the absence of a lens.
Figure 5.
 
HPLC analysis of a pair of 64-year-old human lenses after separate incubation with 1 mM β-benzoylacrylic acid. UV detection at 250 nm. The peak at approximately 24 minutes is the reduced compoundβ -benzoylpropionic acid, and the peak at approximately 26 minutes isβ -benzoylacrylic acid. The peaks between 5 and 15 minutes were due to compounds present in the original AAH. (A) AAH after 48 hours lens incubation in the presence of β-benzoylacrylic acid. (B) Protein-free extract after 24 hours lens incubation in the presence of β-benzoylacrylic acid. (C) Protein-free extract after 48 hours lens incubation in the presence ofβ -benzoylacrylic acid. (D) Control. AAH after 48 hours’ incubation of 1 mM β-benzoylacrylic acid in the absence of a lens.
Figure 6.
 
The percentage conversion of d,l-α-amino-β-benzoylpropionic acid to the deaminated product β-benzoylacrylic acid over 25 hours at pH 6, 7, 8, 9, and 10 at 37°C.
Figure 6.
 
The percentage conversion of d,l-α-amino-β-benzoylpropionic acid to the deaminated product β-benzoylacrylic acid over 25 hours at pH 6, 7, 8, 9, and 10 at 37°C.
Figure 7.
 
HPLC analysis of a 27-year-old human lens after incubation with 1 mM d,l-α-amino-β-benzoylpropionic acid for 48 hours. UV detection at 250 nm. The peak at approximately 10 minutes is d,l-α-amino-β-benzoylpropionic acid, at approximately 17 minutes is β-benzoylpropionic acid, and at approximately 20 minutes is β-benzoylacrylic acid. (A) AAH after incubation of lens in the presence of d,l-α-amino-β-benzoylpropionic acid. (B) Protein-free extract after incubation of lens in the presence of d,l-α-amino-β-benzoylpropionic acid. (C) Control. AAH after 48 hours’ incubation of 1 mM d,l-α-amino-β-benzoylpropionic acid in the absence of a lens.
Figure 7.
 
HPLC analysis of a 27-year-old human lens after incubation with 1 mM d,l-α-amino-β-benzoylpropionic acid for 48 hours. UV detection at 250 nm. The peak at approximately 10 minutes is d,l-α-amino-β-benzoylpropionic acid, at approximately 17 minutes is β-benzoylpropionic acid, and at approximately 20 minutes is β-benzoylacrylic acid. (A) AAH after incubation of lens in the presence of d,l-α-amino-β-benzoylpropionic acid. (B) Protein-free extract after incubation of lens in the presence of d,l-α-amino-β-benzoylpropionic acid. (C) Control. AAH after 48 hours’ incubation of 1 mM d,l-α-amino-β-benzoylpropionic acid in the absence of a lens.
Figure 8.
 
Quantity of 3OHKG and the deaminated product plotted against time (hours) after incubation of 3OHKG (1 mg/ml) at pH 7 and 37°C for a period of 7 days. HPLC peaks detected using fluorescence. Concentrations of 3OHKG and the deaminated product were determined using a 3OHKG standard curve.
Figure 8.
 
Quantity of 3OHKG and the deaminated product plotted against time (hours) after incubation of 3OHKG (1 mg/ml) at pH 7 and 37°C for a period of 7 days. HPLC peaks detected using fluorescence. Concentrations of 3OHKG and the deaminated product were determined using a 3OHKG standard curve.
Figure 9.
 
Quantity of 3OHKG and AHBG found in the AAH plotted against time (hours) after incubation of a 26-year-old lens in AAH at 35°C and subsequent spiking of the AAH at 8 hours with 3OHKG (final concentration 5 mM). HPLC peaks detected using fluorescence.
Figure 9.
 
Quantity of 3OHKG and AHBG found in the AAH plotted against time (hours) after incubation of a 26-year-old lens in AAH at 35°C and subsequent spiking of the AAH at 8 hours with 3OHKG (final concentration 5 mM). HPLC peaks detected using fluorescence.
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×