Received: 10 September 2002Revised: 28 November 2002Accepted: 4 December 2002Published online: 16 April 2003© Springer-Verlag 2003

Abstract Background: Advancedglycation endproduct (AGE) forma-tion is thought to contribute to agingand cataract formation in the lens. Inthis study, we evaluated AGE immu-noreactivity in human diabetic(n=14) and nondiabetic (n=31) cata-ractous lenses in relation to high-mo-lecular-weight (HMW) protein con-tent, which is believed to contributeto the onset of cataract. Methods:AGE immunoreactivity was detectedin alkali-soluble individual lens sam-ples. Competitive ELISA with poly-clonal anti-AGE antibody was per-formed to estimate AGEs. SDS-PAGE was used to detect changes inlens protein composition on the basisof molecular size. Results: Regres-sion analysis of data from nondiabet-ic lenses showed a significant corre-lation between lens AGE content andpatient age (r=0.665, P<0.001). Thecurve exhibited exponential regres-sion (y=0.272·e0.025x). The level ofnonspecified AGEs measured in dia-betic lenses showed an overall in-

crease compared with nondiabeticlenses (4.03±1.85 vs1.78±0.71 AU/mg protein,P<0.0078). SDS-PAGE showed theoccurrence of HMW proteins in bothdiabetic and nondiabetic lens sam-ples. However, in diabetic patients,who had a higher level of AGEs, asignificantly higher proportion ofHMW proteins was also observed.The levels of AGE and percent ofHMW aggregates showed a very sig-nificant correlation (r=0.68,P<0.007) in the diabetic group,whereas in nondiabetics the correla-tion, although positive, did not reachstatistical significance. Conclusion:The AGE distribution, with a higherproportion in the samples of lensesrich in HMW aggregates, corrobo-rates the hypothesis that the ad-vanced glycation process might havea role in degenerative changes in eyelens, which in diabetic patients occurvigorously and much earlier than inthose without diabetes.

Graefe’s Arch Clin Exp Ophthalmol(2003) 241:378–384

DOI 10.1007/s00417-002-0616-2

L A B O R AT O RY I N V E S T I G AT I O N

Rajko PokupecMiro KalauzNikša TurkZdenka Turk

Advanced glycation endproducts in human diabetic and non-diabetic cataractous lenses

Introduction

Cataract is the major cause of impaired vision and blind-ness worldwide, and is often called senile cataract to in-dicate that it is by far more common in advanced age.Diabetes is considered a major risk factor for the devel-opment of cataract and retinopathy [8, 26], and there is amounting body of evidence that cataract reaches maturi-ty almost 10 years earlier in the presence of diabetes [9].The majority of studies have identified hyperglycemia

and disease duration as the leading risk factors. Severaldifferent pathogenetic mechanisms have been proposedto explain the accelerated cataractogenesis of diabetes[6]. These mechanisms include increased polyol pathwayflux [23], elevated oxygen free radical formation [22],and the advanced glycation process [24].

Cataract is associated with conformational changesand unfolding of proteins in the lens, which can arise di-rectly as the result of post-translational modifications.Nonenzymatic glycation leading to advanced glycation

R. Pokupec · M. KalauzDepartment of Ophthalmology,Zagreb University Hospital Center,Zagreb, Croatia

N. TurkMerkur University Hospital,Zagreb, Croatia

Z. Turk (✉)Vuk Vrhovac University Clinic for Diabetes, Endocrinology and Metabolic Diseases,Dugi dol 4A, 10000 Zagreb, Croatiae-mail: zturk@idb.hrTel.: +385-1-2332222Fax: +385-1-2331515

379

endproduct (AGE) formation is the key to such post-translational changes. Glycation occurs when free aminogroups of proteins react nonenzymatically with the acy-clic form of glucose via an Amadori rearrangement.With time, the Amadori product undergoes dehydrationand cleavage reactions to give rise to AGEs with brownor fluorescent property [7]. This process has been knownas the Maillard reaction. Accumulation of these ad-vanced glycated products in long-lived proteins such ascollagen, lens crystallins and elastin participates in thepathogenesis of the complications of diabetes mellitusand aging [31]. Lens proteins are particularly vulnerableto the advanced glycation process for two main reasons.First, the lens is not dependent on insulin for glucose up-take and is thus constantly exposed to high concentra-tions of this sugar. Second, lens crystallins have little orno turnover and are thus particularly susceptible to post-translational changes. The lens has a very high contentof proteins, most of which are represented by crystallins,i.e. structural proteins. They undergo post-translationalchanges in the normal aging process and appear to un-dergo them to a greater extent in pathology. It has beenhypothesized that lens proteins become water-insolubleand aggregate to form large high-molecular-weight(HMW) particles that scatter light, producing lens opaci-ties, which is characteristic of cataracts [12].

The aim of this study was to evaluate AGE immuno-reactivity in human diabetic and nondiabetic cataractouslenses in relation to HMW-protein content, which is be-lieved to contribute to the onset of cataract.

Material and methods

Patients

The study included 45 patients undergoing routine extracapsularcataract extraction at the Zagreb University Hospital Center. Sam-ples were not targeted, i.e., 45 of them were recruited in consecu-tive order from patients who had undergone surgery (M/F=14/31).Mean age was 67.2 (27–88) years in men and 73.1 (51–85) yearsin women. Nineteen (42%) patients had never had medically con-firmed hypertension or diabetes; the remaining 26 (58%) patientshad been treated for either hypertension (n=10) or diabetes (n=13),and three of them had been treated for both conditions. The meanrecorded diabetes duration in diabetic patients was 17.1 (2–36)years. Nonproliferative diabetic retinopathy, proliferative retinopa-thy, and a history of laser photocoagulation were recorded in fourpatients each. Eight diabetic patients were on insulin therapy andfour on oral hypoglycemic agents, whereas in one patient diabetesmellitus was diagnosed at his visit to eye clinic for cataract sur-gery. Patients with hypertension (n=14) were treated for a mean of8.2 (3–14) years. The experimental procedure was designed andcarried out in accordance with ethical standards for human sub-jects, as formulated in the Declaration of Helsinki. All personsgave their informed consent prior to their inclusion in the study.

Lens preparation

The lens from each individual was rinsed with 0.05 mol/l phos-phate-buffered saline (PBS, pH 7.5), decapsulated and homoge-

nized (Biotech International, Braun) in 1.5 ml of 0.1 mol/l NaOHat 1,200–1,300 revolutions/min. The resulting suspension wascentrifuged at 10,000 g for 20 min at 4°C. The alkali-soluble su-pernatant was collected and the pH was immediately adjusted to~7.5 with HCl, and samples were stored at −80°C. The proteinconcentration of alkali-soluble samples was determined by themethod of Bradford [5] using bovine serum albumin as a standard.

ELISA measurements of AGE proteins

The enzyme-linked immunosorbent assay (ELISA) was performedas a competitive test with polyclonal anti-AGE antibody [30]. Theimmunoplate was coated with AGE-albumin, 0.5 µg/well in0.05 mol/l carbonate buffer, pH 9.6, and incubated for 3 h at roomtemperature and overnight at 4°C. This was followed by triplicatewashing with saline containing Tween-20, and the unbound siteswere blocked with SuperBlock blocking buffer in PBS (Pierce).To each well, 0.05 ml of competing antigen (lens sample) dilutedwith PBS-Tween 20 and 0.05 ml of rabbit antiserum against AGEs(dilution 1:1000) were added. The plate was incubated for 3 h atroom temperature and overnight at 4°C. After incubation, the platewas rinsed, and the wells were treated with other enzyme-labeledantibodies. To each well, 0.1 ml of alkaline phosphatase-linked an-ti-rabbit IgG was added. Following incubation for 1 h at 37°C, theplate was extensively washed and developed utilizing p-nitrophe-nyl phosphate as a colorimetric substrate in 2-amino-2-methyl-propanol buffer. After a convenient time, an alkali stopping solu-tion was added and the degree of enzymatic substrate turnover wasdetermined by dual-wavelength absorbance measurement at 405and 630 nm by an ELISA reader. Competitors were added to thewells at concentrations ranging from 10−3 µg/well to 103 µg/well.The difference between sample absorbance and blank absorbancewas calculated. Results were expressed as B/B0, where B wascompetitor absorbance, B0 absorbance in the absence of competi-tor along with correction for nonspecific binding (NSB), and thecalculation followed the relationship: B/B0=(B–NSB):(B0–NSB).

Gel electrophoresis

Alkali-soluble lens samples were investigated by sodium dodecylsulfate–polyacrylamide gel electrophoresis (SDS-PAGE) accord-ing to the procedure of Laemmli [11] with 7% gel 1 mm thickness,imidazole buffer 0.05 mol/l, SDS 1%, and constant current of15 mA per gel. After Coomassie blue staining and destaining,electrophoretic bands were scanned by a 5301-Sebia densitometer.Gamma globulins and immunoglobulin M were used as proteinsize markers (molecular masses 0.15–1.0×106 Da).

Statistical analysis

Data are expressed as mean±SD unless stated otherwise. The dis-tribution of particular variable was tested for normality. Differ-ences between independent groups were examined by Mann–Whitney U test. The relationship between two variables was as-sessed by means of Spearman’s rank correlation coefficient. Re-gression analysis was used as a simultaneous test for relationshipbetween multiple variables. The level of significance was less thanP<0.05.

Results

A total of 45 cataractous human lenses were analyzed.Thirty-one of these were from patients without diabetes

380

(age range 51–88 years, mean 72.8 years, median 75years) and 14 from patients with diabetes (age range27–85 years, mean 67.5 years, median 71 years). AGEimmunoreactivity was detected in alkali-soluble sampleswhich were also sonified. We chose an alkali milieu be-

cause alkali treatment has been proved to efficiently ex-tract insoluble AGE-containing proteins [29]. Sonificat-ion was used to effectively solubilize water-insolubleproteins. Competitive ELISA with polyclonal anti-AGEantibody was performed to estimate AGE content in in-dividual lens samples.

Nondiabetic patients were divided into three sub-groups according to age, and advanced glycation end-products were compared as shown in Fig. 1. There wasno evident difference in AGE lens protein content be-tween the nondiabetic 50–60 and 61–70 age groups(P<0.83). However, the accumulated AGE level was sig-nificantly higher in the >70 age group (P<0.008). Multi-ple regression analysis of data from nondiabetic lensesyielded a significant correlation between AGE lens con-tent and patient age (r=0.665, P<0.001). The curveshowed exponential regression (y=0.272·e0.025x), asshown in Fig. 2. There was no difference in AGE lenscontent between the subjects with and without hyperten-sion in any of the patient groups. The level of nonspeci-fied AGEs measured in alkali-soluble diabetic catarac-tous lenses showed an overall increase compared withnondiabetic lenses (4.03±1.85 vs 1.78±0.71 U/mg pro-tein, P<0.0012) (Fig. 3).

SDS-PAGE was used to detect changes in lens proteincomposition on the basis of molecular size (Fig. 4). Theelectrophoretic pattern of alkali-soluble lens proteins

Fig. 2 Correlation of aging with advanced glycation endproductsin nondiabetic cataractous lenses (AGE AU/mg lens protein). Thecurve shows exponential regression (y=0.272·e0.025x) (r=0.665,P<0.001), pointing to increased accumulation in older patients

Fig. 1 AGE content in nondiabetic cataractous lenses. AGEs werequantified by ELISA AGE as described in Material and methods.Patients were divided into three subgroups according to age. TheAGE content was significantly higher in the group of patientsaged >70 years (P<0.008)

Fig. 3 Comparison of AGE immunoreactivity and proportion of al-kali-soluble HMW aggregates in cataractous lenses from patientswith (DM) and without (nonDM) diabetes. There was a significantincrease in both AGEs (DM vs nonDM P<0.0012) and HMW (DMvs non-DM P<0.0078) in diabetic alkali-soluble lens samples

Table 1 Alkali-soluble lensprotein components (%) fromdiabetic (DM) and nondiabetic(non-DM) patients analyzed atindividual level by SDS-PAGEelectrophoresis on the basis ofmolecular size

Patients Peak 1 Peak 2 Heterogeneous HMW protein size proteins

Non-DM 45.6±5.0 33.2±5.3 13.8±5.8 7.3±2.7DM 43.0±18.5* 34.6±6.9 15.0±7.7 11.6±4.3**

The values are electrophoretic bands scanned by densitometer and each peak is expressed as a per-centage of the total. Gamma globulins were used as a high molecular weight (HMW) size marker. DMvs non-DM: *p<0.025, **p<0.0078

381

showed two main peaks. Since the identity of electro-phoretic bands had not been established, the peaks weremarked with Arabic numerals as peak 1 and peak 2. Sev-eral minor peaks of heterogeneous molecular size ap-peared between peak 2 and slot position. Finally, HMWwhich failed to penetrate the gel were detected. Resultsof SDS-PAGE densitometric analysis of individual lensproteins are summarized in Table 1. In diabetic lens sam-ples, a significantly higher percent of HMW protein ag-gregates was detected as compared with samples fromnondiabetic individuals (P<0.0078) (Fig. 3). Concerningother electrophoretic bands, a difference in the propor-

Fig. 4 SDS-PAGE pattern of alkali-soluble cataractous lenses ob-tained from a 27-year-old diabetic patient (curve A) and an 80-year-old nondiabetic patient (curve B). Electrophoretic bands werequantitated after staining by densitometer scanning. Since theidentity of electrophoretic bands had not been established, theywere marked as peak 1, peak 2, heterogeneous size proteins be-tween peak 2 and slot position, and finally HMW aggregates that

Fig. 5A, B Relationship of AGE content in cataractous lenses andmoiety of HMW aggregates detected by means of SDS-PAGE. Di-abetic patients (A) showed a high degree of significant correlationbetween AGEs and HMW proteins (r=0.68, P<0.007), but in thenondiabetic group (B), although positive, the correlation did notreach statistical significance (r=0.34, P<0.059)

failed to penetrate to the gel. There was a significantly higher pro-portion of HMW in the lens sample of the younger diabetic patientthan in that of the 80-year-old nondiabetic patient (DM HMW vsnon-DM HMW: 11% vs 2.7% of total)

ship between AGE level and age was not linear but wascharacterized by exponential regression, which may bethe result of a cumulative effect (Fig. 2). In the diabeticgroup we observed a significantly higher level of AGEsthan in nondiabetic individuals, which is consistent withprevious findings [1, 16, 25, 28, 32]. Increased AGE lev-el in serum and tissues is well documented; however,there have been many more studies in experimental thanhuman diabetes, which can be explained by the fact thathuman tissue is not easily available for analyses. Therole of AGEs in diabetic and nondiabetic cataractogene-sis is potentially important, especially because this pro-cess induces structural and functional implications [14].Because various AGEs can form by a variety of chemi-cal reactions according to environmental conditions, thereasons for the formation of such structures in nondia-betic conditions are difficult to explain. The role of oxi-dative stress has been suggested [18, 21]. Therefore, itmight be postulated that reactive oxygen intermediatesmay accelerate the rate of AGE formation via reactiveoxoaldehydes, and that, vice versa, AGEs might induceoxidative stress through chemical and cellular mecha-nisms. Nε-(carboxymethyl) lysine (CML) is a typical ex-ample of a glyco-oxidation product, which can accountfor up to 50% of total AGE structures [2, 20]. However,as reported elsewhere, there was no evident difference inCML level in senile and diabetic lenses [15, 32]. In con-trast, total AGE lens content in diabetic state was foundto be increased in most studies [1,16, 25, 28, 32], our re-sults also being in accordance with these reports. Thisresult of AGE analysis in human lens is consistent withour earlier study carried out in experimental diabetes[30], which demonstrated an increase in fluorescing AG-Es in diabetic state. We can therefore speculate that AGEstructures resulting from persisting hyperglycemia aremore profusely formed in diabetes. Such observationssuggest that ocular permeability barriers to glucose maybe compromised in diabetic patients. A case from thisstudy may serve as an illustration to corroborate thisstatement. A 27-year-old patient suffering from diabetesmellitus for 25 years, had a total AGE lens content of7.5 AU/mg lens protein. On the other hand, in all elderlynondiabetic individuals (median age 75 years, range51–88) the highest individually measured AGE valuewas 3.11 AU/mg. In the present study, we used polyclo-nal anti-AGE antibodies, their epitope(s) being unidenti-fied. That is why we could not identify which of theAGE lens structures was predominant in the lenses of di-abetic versus nondiabetic subjects.

Previous studies have shown that by in vitro glycationlens proteins reticulate, forming dimers, trimers andpolymers as well as aggregates of large molecular mass-es [13]. Many AGE structures (pentosidine, crossline,imidazolone), which either reticulate or form cyclicstructures, modify lens proteins and in doing so causechanges in conformation [4, 17, 27]. All these may be

382

tion of protein fraction constituting densitometric peak 1was also observed, whereas in other electrophoretic frac-tions there was no significant difference in lens proteinsbetween diabetic and nondiabetic patients. Furthermore,we analyzed the relationship between glycated proteincontent and moiety of HMW aggregates for all investi-gated lenses. We observed a positive and statistically sig-nificant correlation between the level of AGEs and pro-portion of HMW proteins. However, separate analyses ofthe correlation of these parameters for the groups of dia-betic and nondiabetic individuals showed different re-sults. While diabetic patients showed a high degree ofsignificant correlation between AGEs and HMW pro-teins (r=0.68, P<0.007) (Fig. 5A), the correlation in thegroup of nondiabetic individuals, although positive, didnot reach statistical significance (r=0.34, P<0.059)(Fig. 5B).

Discussion

Massive increase in water insoluble protein fraction isknown to be characteristic of human and experimentalcataract. Protein fractions made of protein polymers andcharacterized by brown coloration and fluorescence ap-pear in such lenses [19]. These brown, fluorescent pro-tein polymers are important for human cataractogenesis.Nevertheless, despite numerous studies, their nature hasnot been fully investigated. Such proteins are known tobe better soluble in an alkaline medium. Alkalinity mayaffect post-translational inter- and intramolecular links,releasing a certain degree of reticulation characteristic ofadvanced glycated structures, which are then more easilydetected. In the present study, an AGE ELISA methodwith polyclonal anti-AGE antibody was developed toquantify AGE products in alkali-soluble lens proteins.The same AGE antibody as reported in our previous in-vestigations was used [29, 30]. In contrast to fluores-cence methods, which can only detect fluorescing AGEstructures, the advantage of AGE ELISA is that it candetect low levels of immunoreactive AGE moieties, re-gardless of whether the analyzed AGE structure containsfluorescent or nonfluorescent AGE compounds.

In the present study, we observed positive correlationof AGE lens content with increasing age in patientswithout diabetes. AGEs have previously been shown toaccumulate in many tissues with age, independently ofdiabetes [3, 10]. Advanced glycation reactions are knownto be most intensive in tissues with little or almost noturnover, and in tissues that do not require insulin forglucose utilization. Since the body does not contain anysingle enzyme capable of AGE structure degradation,AGEs accumulate during the biological life of proteinson which they had been formed. Lens proteins are long-lived and might therefore be the subject of accumulativeadvanced glycation reactions. We found that the relation-

383

factors responsible for the increased formation of proteinaggregates and insolubility, which results in light scatter-ing and cataract formation [12]. The presented resultsconfirm the relationship between AGE lens content andmoiety of HMW polymeric products. SDS-PAGEshowed the occurrence of HMW proteins in both diabet-ic and nondiabetic lens samples (Fig. 4). However, in di-abetic patients, who had higher levels of AGEs on an av-erage, a significantly higher proportion of HMW pro-teins was also observed. In addition, the levels of AGEsand the proportion of HMW aggregates showed a very

significant correlation (r=0.68, P<0.007) in the diabeticgroup, whereas in nondiabetic individuals the correla-tion, although positive, did not reach statistical signifi-cance.

The results of this study support the hypothesis thatadvanced glycation process may have an important rolein degenerative changes in eye lens, which in diabeticpatients occur much earlier than in those without diabe-tes. The AGE distribution, with a higher proportion inthe samples of lenses rich in HMW aggregates, corrobo-rates this hypothesis.

References1. Abiko T, Abiko A, Ishiko S, Takeda M,

Horiuchi S, Yoshida A (1999) Rela-tionship between autofluorescence andadvanced glycation end products in di-abetic lenses. Exp Eye Res 68:361–366

2. Ahmed MU, Brinkmann E, DegenhardtTP, Thorpe SR, Baynes JW (1997) Nε-(carboxymethyl)lysine, a product ofthe chemical modification of proteinsby methylglyoxal, increases with agein human lens proteins. Biochem J324:565–570

3. Araki N, Ueno N, Chakrabarti B, Morino Y, Horiuchi S (1992) Immuno-chemical evidence for the presence ofadvanced glycation end products in hu-man lens proteins and its positive cor-relation with aging. J Biol Chem267:10211–10214

4. Beswick HT, Harding JJ (1987) Con-formational changes induced in lens αand γ crystallins by modification withglucose-6-phosphate. Biochem J246:761–769

5. Bradford MM (1976) A rapid and sen-sitive method for the quantities of pro-tein utilizing the principle of pro-tein–dye binding. Anal Biochem72:248–254

6. Bron AJ, Brown NAP, Harding JJ,Ganea E (1998) The lens and cataractin diabetes. Int Ophthalmol Clin38:37–67

7. Bucala R, Vlassara H, Cerami A(1992) Advanced glycosylation end-product. In: Harding JJ, Crabbe MJC(eds) Post-translational modificationsof proteins, vol 2. CRC Press, BocaRaton, pp 53–59

8. Clayton RM, Cuthbert J, Seth J, Phillips CI, Bartholomew RS, Reid JM(1984) Epidemiological and other stud-ies in the assessment of factors contrib-uting to cataractogenesis. Ciba Founda-tion Symposium 106:25–47

9. Ederer F, Hiller R, Taylor HR. (1981)Senile lens changes and diabetes in twopopulation studies. Am J Ophthalmol91:381–395

10. Kasper M, Funk RHW (2001) Age-re-lated changes in cells and tissues due toadvanced glycation end products (AG-Es). Arch Gerontol Geriatr 32:233–243

11. Laemmli UK (1970) Cleavage of struc-tural protein during the assembly of thehead of bacteriophage. Nature227:680–685

12. Liang JN, Li XY (1991) Interactionand aggregation of lens crystallins. ExpEye Res 53:61–66

13. Liang JN, Rossi MT (1990) In vitronon-enzymatic glycation and formationof browning products in the bovinelens alpha-crystallin. Exp Eye Res50:367–371

14. Luthra M, Balasubramania D (1993)Nonenzymatic glycation alters proteinstructure and stability. A study of twoeye lens crystallins. J Biol Chem268:18119–18127

15. Lyons TJ, Silvestri G, Dunn JA, DyerDG, Baynes JW (1991) Role of glycat-ion in modification of lens crystallinsin diabetic and nondiabetic senile cata-racts. Diabetes 40:1010–1015

16. Matsumoto K, Ikeda K, Horiuchi S,Zhao H, Abraham EC (1997) Immuno-chemical evidence for increased forma-tion of advanced glycation end prod-ucts and inhibition by aminoguanidinein diabetic rat lenses. Biochem Bio-phys Res Commun 241:352–354

17. Nagaraj RH, Sady C (1996) The pres-ence of a glucose derived Maillard re-action product in human lens. FEBSLett 382:234–238

18. Nagaraj RH, Sell DR, Prabhakaram M,Ortwerth BJ, Monnier VM (1991) Highcorrelation between pentosidine proteincrosslinks and pigmentation implicatesascorbate oxidation in human lens se-nescence and cataractogenesis. ProcNatl Acad Sci USA 88:10257–10261

19. Pirie A (1968) Color and solubility ofthe proteins of human cataract. InvestOphthalmol Vis Sci 7:634–650

20. Reddy S, Bichler J, Wells-Knecht KJ,Thorpe SR, Baynes JW (1995) Nε-(car-boxymethyl)lysine is a dominant ad-vanced glycation end product (AGE)antigen in tissue proteins. Biochemistry34:10872–10878

21. Shamsi FA, Lin K, Sady C, NagarajRH (1998) Methylglyoxal derivedmodification in lens aging and cataractformation. Invest Ophthalmol Vis Sci39:2355–2364

22. Spector A (1995) Oxidative stress-in-duced cataract: mechanism of actionFASEBJ 9;1173–1182.

23. Stevens MJ, Obrosova I, Feldman EL,Green DA (2000) The sorbitol-osmoticand sorbitol-redox hypothesis. In: Le-Roith D, Taylor SI, Olevsky JM (eds)Diabetes mellitus: a fundamental andclinical text. Lippincott, Williams &Wilkins, Philadelphia, pp 972–982

24. Stitt AW (2001) Advanced glycation:an important pathological event in dia-betic and age related ocular disease BrJ Ophthalmol 85:746–753

25. Stitt AW, Moore JE, Sharkey JA, Murphy G, Simpson DA, Bucala R,Vlassara H, Archer DB (1998) Ad-vanced glycation end products in vitre-ous: structural and functional implica-tions for diabetes vitreopathy. InvestOphthalmol Vis Sci 39:2517–2523

26. Struck HG, Heider C, LautenschlagerC (2001) Is diabetes in the elderly pa-tient a risk factor for cataracts? Oph-thalmologe 98:952–955

27. Swamy MS, Tsai C, Abraham A, Abraham EC (1993) Glycation mediat-ed lens crystallin aggregation andcross-linking by various sugars andsugar phosphates in vitro. Exp Eye Res56:177–185

28. Turk Z, Mišur I, Turk N (1997) Tempo-ral association between lens proteinglycation and cataract development indiabetic rats. Acta Diabetol 34:49–54

384

29. Turk Z, Mesic R, Benko B (1998)Comparison of advanced glycationendproducts on haemoglobin (Hb-AGE) and haemoglobin A1c forthe assessment of diabetic control. ClinChim Acta 277:159–170

30. Turk Z, Ljubić S, Turk N, Benko B(2001) Detection of autoantibodiesagainst advanced glycation endprod-ucts and AGE-immune complexes inserum of patients with diabetes melli-tus. Clin Chim Acta 303:105–115

31. Vlassara H, Bucala R, Striker L (1994)Pathogenic effects of advanced glyco-sylation endproducts: biochemical, bio-logic, and clinical implications for dia-betes and aging. Lab Invest70:138–151

32. Zarina S, Zhao HR, Abraham EC(2000) Advanced glycation end prod-ucts in human senile and diabetic cata-ractous lenses. Mol Cell Biochem210:29–34