Skip to main content

Evaluation of N ε-(3-formyl-3,4-dehydropiperidino)lysine as a novel biomarker for the severity of diabetic retinopathy



Recent studies suggest that oxidative stress should be monitored alongside HbA1c to identify subgroups of diabetic patients at high risk of initiation or progression of retinopathy. The acrolein-derived advanced lipoxidation end-product (ALE), \(N^{\text{ $ \varepsilon $ }} \)-(3-formyl-3,4-dehydropiperidino)lysine (FDP-lysine), is a useful biomarker that reflects the cumulative burden of oxidative stress over long periods of time. The purpose of the present study was to investigate whether serum and haemoglobin levels of FDP-lysine are associated with the severity of diabetic retinopathy in type 1 and type 2 diabetic patients.


Serum and haemoglobin levels of FDP-lysine were measured by competitive ELISA in 59 type 1 and 76 type 2 diabetic patients with no retinopathy, non-proliferative retinopathy or proliferative retinopathy (mean age [±SEM] 54.3 ± 1.3 years), and in 47 non-diabetic control individuals (mean age 51.9 ± 2.1 years).


Serum and haemoglobin levels of FDP-lysine were significantly increased in diabetic patients compared with control individuals (p = 0.04 and p = 0.002, respectively). However, no significant association was found between levels of serum FDP-lysine and the severity of diabetic retinopathy (p = 0.97). In contrast, increased haemoglobin FDP-lysine levels were observed in patients with proliferative retinopathy compared with patients without retinopathy and with non-proliferative retinopathy (p = 0.04). The relationship of FDP-lysine with proliferative retinopathy was unaltered after adjustment for HbA1c, or other clinical parameters.


Our data suggest that haemoglobin FDP-lysine may provide a useful risk marker for the development of proliferative diabetic retinopathy independently of HbA1c, and that elevated intracellular ALE formation may be involved in the pathogenesis of this sight-threatening complication of diabetes.


Retinopathy is one of the most common microvascular complication of diabetes and the leading cause of acquired blindness in the working population of developed countries [1]. After 20 years of diabetes, nearly all patients with type 1 diabetes will have at least some retinopathy. Overall, diabetic retinopathy is slightly less common in type 2 diabetic patients, but is still the most frequent microvascular complication suffered by this group [2]. Although the pathogenic basis of diabetic retinopathy is not wholly understood at the cellular and molecular level, large prospective clinical trials have demonstrated a strong relationship between time-averaged mean levels of glycaemia (measured as HbA1c) and the rate of development of retinopathy in both type 1 and type 2 diabetes [3, 4].

Hyperglycaemia is believed to contribute to retinal microangiopathy through the induction of oxidative stress [5]. Tissue oxidative stress can be evoked in diabetes through several mechanisms such as overproduction of oxygen radicals by the mitochondrial electron transport chain [6] and impairment of antioxidant defences [7, 8]. Reactive oxygen species (ROS) may directly cause oxidative damage of lipids, proteins and nucleic acids, impairing cell functions and leading to cell death by apoptosis [9]. Moreover, ROS are considered to be a causal link between elevated glucose and activation of other known pathways of hyperglycaemic damage, i.e. the polyol, hexosamine, diacylglycerol–protein kinase C and AGE pathways [6]. Investigation of appropriate animal models strongly supports the premise that oxidative stress plays an important role in the pathogenesis of diabetic retinopathy. Superoxide levels are elevated in the diabetic retina [10, 11] and inhibition of superoxide accumulation with antioxidants [12, 13] or overproduction of mitochondrial superoxide dismutase (SOD) [14, 15] can prevent lesions of retinopathy in experimentally induced diabetes in rodents.

While there is no dispute that hyperglycaemia is a critical factor in the aetiology of diabetic retinopathy, there are subgroups of patients where time-averaged mean HbA1c values do not predict clinical prognoses [16]. This might be explained by the fact that HbA1c fails to reflect the magnitude of acute glycaemic fluctuations, which may foster a pro-oxidative state [17]. In addition, genetically determined factors appear to influence susceptibility to diabetic retinopathy [18], which possibly involves polymorphic variations in genes that govern the generation of, or protection from, oxidative damage [19, 20]. These findings have important clinical implications and suggest that it may be beneficial to monitor oxidative stress in concert with HbA1c to identify subgroups of diabetic patients at high risk of initiation and progression of retinopathy. Unfortunately, currently used markers of oxidative stress in vivo, such as thiobarbituric acid-reacting substances [21] and F2-isoprostanes [17], are relatively limited in their clinical usefulness as they only provide a ‘snapshot’ of ROS formation at the time of sampling and do not report integrated oxidative stress over a prolonged period.

Lipid peroxidation caused by ROS yields lipid hydroperoxides that decompose to reactive aldehydes such as 4-hydroxynonenal, malondialdehyde and acrolein. These compounds subsequently react with the nucleophilic sites of proteins generating relatively stable chemical adducts known as advanced lipoxidation end-products (ALEs) [22]. Acrolein is one of the most reactive products of lipid peroxidation, and this aldehyde has been shown to preferentially modify lysine residues of proteins [23]. Reaction of acrolein with the ɛ-amino group of lysine leads to the formation of \(N^{\text{ $ \varepsilon $ }} \)-(3-formyl-3,4-dehydropiperidino)lysine (FDP-lysine) [23]. The presence of FDP-lysine has been demonstrated in human serum and urine [2426] and represents a marker of cumulative oxidative stress [27].

As a chronic marker of oxidative stress, FDP-lysine might represent a useful prognostic biomarker for the development diabetic retinopathy independently of HbA1c. Therefore, in the present study, serum and haemoglobin levels of FDP-lysine have been quantified in well-characterised diabetic patients with differing stages of retinopathy and compared with age-matched non-diabetic controls.


Patients and healthy volunteers

The patients consisted of 59 type 1 and 76 type 2 diabetic patients of white ethnic background. All were outpatients in the Regional Centre for Endocrinology and Diabetes, Royal Victoria Hospital, Belfast. Forty-seven healthy volunteers who were matched for sex, age and ethnicity served as controls. To rule out any confounding effects of renal impairment in causing an elevation of FDP-lysine levels, diabetic patients with clinical evidence of proteinuria were excluded from the study. Blood specimens were analysed for HbA1c, cholesterol, HDL-cholesterol, LDL-cholesterol and triacylglycerol using National Health Service approved methods at the Clinical Chemistry, Regional Endocrine, Haematology and Coagulation Laboratories, Royal Victoria Hospital, Belfast. BP was measured at the right brachial artery using an Omron HEM-705CP automated sphygmomanometer (Omron Healthcare, Henfield, UK). Mean systolic and diastolic BPs were averages of the last two of three measurements. Hypertension was defined as a mean systolic BP ≥160 mmHg, and/or a mean diastolic BP of ≥95 mmHg or a history of antihypertensive medication [28]. Retinopathy was assessed by fundus photography and was graded according to the UK National Screening Committee protocol [29]. Diabetic retinopathy in each patient was classified as ‘none’ (level R0), ‘non-proliferative’ (levels R1/R2) or ‘proliferative’ (level R3). The study was performed in accordance with the declaration of Helsinki and was approved by the Research Ethics Committee of Queen’s University Belfast. All patients gave written informed consent to participate.

Preparation of serum and haemoglobin samples

Venous blood was drawn into two polypropylene tubes (5 ml per tube) containing clot activator or heparin for the isolation of serum and haemoglobin, respectively, and stored on ice. For the serum, blood was allowed to clot for ∼30 min, before being centrifuged at 1,000 g for 10 min at 4°C. Serum was immediately frozen at −80°C for later use. For haemoglobin, isotonic buffer (0.9% NaCl with 5 mmol/l sodium phosphate, pH 8) was added to whole blood. After trituration, the suspension was centrifuged at 600 g for 10 min. Supernatant was discarded and the pellet resuspended in isotonic buffer by trituration, then re-centrifuged. The erythrocyte pellet was resuspended in hypotonic buffer (5 mmol/l sodium phosphate, pH 8) to lyse the cells and the samples were centrifuged at 12,000 g for 10 min. Protein concentrations of the haemoglobin samples were measured with a BCA Protein Assay Kit (Pierce, Rockford, IL, USA). Haemoglobin samples were immediately frozen at −80°C for later use.

FDP-lysine assay

Serum and haemoglobin levels of FDP-lysine were determined with a competitive ELISA assay using a well-characterised monoclonal antibody specific for acrolein-modified FDP-lysine residue (mAb5F6) prepared as previously described [27]. FDP-lysine-human serum albumin (HSA) was prepared by incubating 10 mg/ml HSA with 10 mmol/l acrolein (stabilised; Sigma, Poole, UK) in 50 mmol/l PBS (pH 7.2) at 37°C for 24 h in the dark. After dialysis against PBS, FDP-lysine-HSA was stored at −20°C. Microtitre plates (96 wells) were coated with 0.1 μg/ml FDP-lysine-HSA in coating buffer (0.05 mol/l carbonate buffer, pH 9.6, 200 μl per well) overnight at 4°C. ELISA plates were washed extensively three times with washing buffer (0.2 mmol/l KH2PO4, 1.4 mmol/l K2HPO4, 3 mmol/l NaCl, 0.005 mmol/l sorbic acid potassium, 0.01% Tween-20) and wells were blocked for 2 h with 3% skim milk. After rinsing with washing buffer, 50 μl competing antigen (FDP-lysine-HSA standard [0.1–3 μg/ml]), serum (1:4 vol./vol.) or haemoglobin (30 mg/ml) in dilution buffer (0.05% Tween-20 and 0.2% BSA in 75 mmol/l PBS pH 7.4) and 50 μl of mAb5F6 (50 ng/ml) were added for 2 h at room temperature. After washing, peroxidase conjugated anti-mouse IgG (diluted 1:5,000 vol./vol.) was added and incubated for 1 h. Following three subsequent washing steps, the colour reaction was induced with 150 μl per well 3,3′,5,5′-tetramethyl-benzidine (TMB) diluted in substrate buffer (340 μl TMB solution [52 mmol/l 3,3′,5,5′-TMB (Sigma), 1 ml DMSO, 5 ml methanol] plus 10 μl 30% H2O2 [wt/wt] per 20 ml substrate buffer [0.2 mol/l potassium dihydrogen citrate, 0.7 mmol/l sorbic acid potassium]) and stopped after 15 min by the addition of 50 μl per well of 1 mol/l H2SO4. Absorbance was read using a microtitre plate reader (Multifunction Microplate Reader, Safire; TECAN, Reading, UK) at 450 nm. Results were first determined as % B/B 0, where B = the absorbance in the presence of standard or sample and B 0 = absorbance in the absence of standard or sample. The FDP-lysine content of the samples was then quantified by using a standard calibration curve. Final results are given as: serum FDP-lysine concentrations (μg/ml) referring to the 1 μg/ml of synthesised FDP-lysine-HSA standard; haemoglobin FDP-lysine concentrations (arbitrary units [AU]/mg protein) is equal to 1 μg/ml of synthesised FDP-lysine-HSA standard per mg haemoglobin. All samples were run in triplicate. Intra-assay CV values of our ELISA system were 2.7% and inter-assay values were 6.6%.

Statistical analysis

Data are presented as means ± SEM. Statistical analyses were performed using Prism V4.02 (Graphpad Software, San Diego, CA, USA). Normal distribution was assessed by the D’Agostino and Pearson omnibus normality test. The Mann–Whitney U test was used for statistical analysis of serum and haemoglobin FDP-lysine between diabetic patients and control individuals. Multiple comparisons were performed using one-way ANOVA with Newmans–Keuls test for normally distributed data and Kruskal–Wallis analysis followed by Dunn’s test for non-normally distributed data. The χ 2 test was used to compare differences in proportions between groups. Bivariate logistic regression and receiver operating characteristics curve analyses were performed using SPSS statistical software V15 (SPSS, Chicago, IL, USA).


The serum level of FDP-lysine in diabetic patients was significantly greater than in non-diabetic volunteers (22.4 ± 0.51 and 19.61 ± 0.7 μg/ml, respectively; p = 0.04; Fig. 1a). Likewise, the haemoglobin level of FDP-lysine was elevated in diabetic patients compared with non-diabetic individuals (0.38 ± 0.02 AU/mg protein and 0.22 ± 0.03 AU/mg protein, respectively; p = 0.002; Fig. 1b).

Fig. 1
figure 1

Comparisons of FDP-lysine in serum of control individuals and diabetic patients (a), and from haemoglobin (Hb) from controls vs diabetic patients (b). The horizontal bars indicate mean values. *p < 0.05, **p < 0.01 vs control individuals

Patients were subdivided into three groups according to retinopathy status, and FDP-lysine levels were investigated. Eighty-one patients had no retinopathy and 30 patients were classified as having non-proliferative retinopathy, whilst 24 patients were categorised as having proliferative retinopathy (Table 1). Because of the criteria used for recruitment, no significant differences in age, sex or renal function were evident among the groups. Fasting blood glucose and HbA1c were significantly higher in diabetic patients compared with non-diabetic controls, and the severity of retinopathy was associated with a longer duration of diabetes. There was no significant difference in serum blood lipid profiles or diastolic BP among the groups; however, compared with controls, systolic BP and BMI were higher in diabetic patients with non-proliferative retinopathy and no retinopathy, respectively (Table 1). No association was found between levels of serum FDP-lysine and the severity of diabetic retinopathy (p = 0.97; Fig. 2a). In contrast, haemoglobin FDP-lysine levels were significantly different between the retinopathy groups (p = 0.04; Fig. 2b). Increased haemoglobin levels of FDP-lysine were seen when patients with proliferative retinopathy (0.47 ± 0.05 AU/mg protein) were compared with diabetic patients without retinopathy (0.36 ± 0.02 AU/mg protein; p < 0.05) and with non-proliferative retinopathy (0.32 ± 0.04 AU/mg protein; p < 0.05). A significant elevation in haemoglobin FDP-lysine levels was also evident in patients with no retinopathy (p < 0.01), non-proliferative retinopathy (p < 0.05) and proliferative retinopathy (p < 0.001) when compared with control individuals.

Fig. 2
figure 2

Association of serum (a) and haemoglobin (Hb) (b) concentrations of FDP-lysine with the severity of diabetic retinopathy. The horizontal bars indicate mean values. Post hoc analysis: *p < 0.05, **p < 0.01, ***p < 0.001 vs control individuals; p < 0.001 vs NPDR and no retinopathy. NPDR, non-proliferative diabetic retinopathy; PDR, proliferative diabetic retinopathy

Table 1 Clinical characteristics of controls and diabetic patients grouped according to retinopathy status

To determine whether haemoglobin FDP-lysine was independently associated with retinopathy status, we divided diabetic patients into two groups according to whether or not they had proliferative disease. Bivariate logistic regression analysis was then performed to examine the relationship between haemogloblin FDP-lysine and retinopathy before and after adjustment for potential confounding variables. Confounding variables chosen for inclusion in the logistic regression analysis were those known from previous studies to be risk factors for the progression of diabetic retinopathy, including HbA1c, diabetes duration, sex, BMI, diastolic BP and hypertension [28, 30, 31]. Type of diabetes and treatment with insulin were also included as possible confounding factors since these parameters varied between the two groups (p < 0.001 in both cases; χ 2 test). In an unadjusted analysis, a significant association was found between the haemoglobin level of FDP-lysine and retinopathy (Table 2; OR 1.38; p = 0.024). After adjustment for confounders, ORs remained essentially unaltered (Table 2), suggesting that FDP-lysine residues on haemoglobin are independently associated with the development of proliferative retinopathy.

Table 2 Unadjusted and adjusted ORs and 95% CIs for haemoglobin FDP-lysine and retinopathy

We constructed a receiver operating characteristics curve plot to compare haemoglobin FDP-lysine levels of patients with non-proliferative retinopathy with those of diabetic patients with proliferative retinopathy (Fig. 3). This facilitated selection of the optimal cut-point (threshold) associated with proliferative diabetic retinopathy. The optimal cut-point for balancing sensitivity and specificity of the haemoglobin FDP-lysine test was defined as the point on the curve closest to the upper left hand corner (dotted line, Fig. 3). This point corresponded to an FDP-lysine level of 0.3753 AU/mg protein, giving a sensitivity of 72% and a specificity of 71% for proliferative retinopathy.

Fig. 3
figure 3

Receiver operating characteristics curve plot comparing haemoglobin FDP-lysine levels of patients with non-proliferative retinopathy with those of diabetic patients with proliferative retinopathy. The diagonal broken line is the line of no discrimination. For the presence of proliferative retinopathy a ‘decision threshold’ haemoglobin FDP-lysine level of 0.3753 AU/mg protein was defined as the far left point in the curve, giving 72% sensitivity and 71% specificity. AUC = 0.743


To the best of our knowledge, this is the first clinical study examining serum and haemoglobin levels of FDP-lysine in diabetes. We found that both serum and haemoglobin levels of FDP-lysine were significantly elevated in patients with diabetes compared with control individuals. Our results are consistent with previous reports showing that urinary levels of acrolein-lysine adduct are elevated in type 1 and type 2 diabetic patients [25, 26]. In addition, they further confirm the widely accepted view that increased oxidative stress and lipid peroxidation are characteristic of the diabetic milieu.

A primary goal of the present study was to evaluate whether protein-bound FDP-lysine adduct was associated with diabetic retinopathy progression and to determine if it could provide a robust biomarker for oxidative stress over prolonged time-frames. We have shown that serum FDP-lysine levels were not significantly associated with retinopathy status. In contrast, haemoglobin FDP-lysine adducts were associated with the severity of diabetic retinopathy, with elevated levels being detected in patients with proliferative disease. The disparity between these findings may lie in the different half-lives of serum proteins and haemoglobin. Albumin, the most abundant serum protein, has a relatively short half-life of ∼20 days [32] meaning that serum FDP-lysine levels may only reflect integrated oxidative stress over ∼2–3 weeks. In contrast, erythrocytes have a lifespan of ∼120 days [32] and thus haemoglobin FDP-lysine adducts may represent the accumulation of ROS-induced damage over several months.

In the ELISA assay, haemoglobin FDP-lysine was normalised to haemoglobin content using the BCA protein assay kit (haemoglobin constitutes ∼95% of intracellular protein in erythrocytes [33]). In future studies, the precision of the assay may be improved by normalising to the exact haemoglobin concentration of the samples. This may be particularly important in studies where patients and controls are not matched for age, sex or ethnicity, since all of these factors may affect the haemoglobin concentration of erythrocytes [3436].

Mounting evidence from animal studies suggests that oxidative stress is a major pathogenic factor for diabetic retinopathy. In particular, exogenous antioxidants [12, 13] or overproduction of endogenous defence enzymes such as mitochondrial SOD protect against the development of retinal microvascular lesions in animal models of diabetes [14, 15]. Despite this, results from clinical studies have failed to unambiguously establish if oxidative stress plays any role in the development of retinopathy in diabetic patients. The outcomes of trials of antioxidants on the progression of retinopathy in diabetic patients have been contradictory, although in some cases the concentrations of antioxidants in the retina may not have been sufficiently elevated to produce beneficial effects [5]. Other studies investigating the association between serum markers of oxidative stress and the progression of diabetic retinopathy have also been inconclusive. For example, Hartnett et al. (2000) demonstrated increased levels of thiobarbituric acid-reacting substances and decreased levels of SOD and glutathione peroxidase in serum of diabetic patients when compared with non-diabetic controls, but no significant associations were found with the severity of retinopathy [21]. Most commonly used markers only reflect short-term levels of oxidative stress and this has greatly contributed to the ambiguity in this area. Despite the obvious limitations on the cross-sectional design of the current investigation, we believe that the haemoglobin FDP-lysine data presented herein provide some of the strongest evidence to date that oxidative stress is associated with the development of human diabetic retinopathy.

Although hyperglycaemia-induced ROS production is widely considered to be a causal factor in the initiation and progression of diabetic retinopathy, the exact mechanism by which oxidative stress contributes to the development of this condition remains to be elucidated. Mitochondria are believed to be the main source of superoxide in the diabetic retina [10] and this can lead to further biochemical imbalance such as activation of the polyol, hexosamine, diacylglycerol–protein kinase C pathways and accelerated accumulation of AGEs [6]. Furthermore, ROS may directly contribute to progressive vasodegeneration in the diabetic retina by causing mitochondrial cytochrome c release and activation of caspase-3, thereby initiating capillary cell apoptosis [5, 11, 37]. Findings from the present study also support a possible role for ROS-mediated lipid peroxidation in the development of diabetic retinopathy. Lipid aldehydes (e.g. acrolein, 4-hydroxynonenal, malondialdehyde), produced as a consequence of lipid peroxidation reactions, are believed to be important mediators of oxidative stress and can induce cell death by activating both the mitochondrial and death receptor pathways of apoptosis [38, 39]. ALEs and their precursors have been implicated in macrovascular endothelium dysfunction and atheromatous lesion formation [27], but their precise role in the development of diabetic microvascular complications remains unresolved. Currently, no drugs have been designed that specifically inhibit ALE formation [40], but future development of such compounds should allow for a clearer understanding of the contribution of lipid-derived aldehydes and ALEs in the aetiology of retinopathy and other diabetic complications.

There is no question that time-averaged mean HbA1c levels are closely linked to the development and progression of diabetic eye disease [3, 4]. However, as demonstrated from the Diabetes Control and Complications Trial, patients with identical HbA1c levels may in fact have significantly different risk profiles for diabetic retinopathy [41]. Recent work has suggested that monitoring oxidative stress in parallel with hyperglycaemia may help to identify subsets of patients at high risk of advancement of diabetic retinopathy [1921]. Carboxymethyl-lysine (CML), a product of glycooxidation and lipoxidation reactions, is often regarded as a general marker of oxidative stress [42] and a number of studies have shown cross-sectional associations between serum CML levels and the severity of diabetic retinopathy [4345]. In most cases, however, serum CML levels strongly correlate with HbA1c [46], suggesting that serum CML may not provide any additional risk information beyond that provided by monitoring mean levels of glycaemia. In contrast, other studies have reported that CML concentrations in protein from memory T cells [16] and in skin collagen [47] are of predictive value for the risk of future progression of retinopathy and appear to be independent of HbA1c. However, the complexity of sample preparation may preclude the widespread use of these biomarkers. In the present study we found that FDP-lysine residues on haemoglobin, which reflect cumulative oxidative stress, are also independent predictors of retinopathy in type 1 and type 2 diabetic patients. Thus haemoglobin FDP-lysine, which is relatively straightforward to measure, may prove to be a very useful tool for identifying patients at high risk of diabetic retinopathy not evident from their HbA1c profiles. To specifically address this issue, prospective studies are now warranted.



advanced lipoxidation end-product


arbitrary units


\(N^{\text{ $ \varepsilon $ }} \)-(3-formyl-3,4-dehydropiperidino)lysine


human serum albumin


reactive oxygen species


superoxide dismutase


  1. Aiello LP, Gardner TW, King GL et al (1998) Diabetic retinopathy. Diabetes Care 21:143–156

    PubMed  CAS  Google Scholar 

  2. Klein R, Klein BE, Moss SE (1989) The Wisconsin epidemiological study of diabetic retinopathy: a review. Diabetes Metab Rev 5:559–570

    PubMed  CAS  Article  Google Scholar 

  3. Anonymous (1993) The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. The Diabetes Control and Complications Trial Research Group. N Engl J Med 329:977–986

    Article  Google Scholar 

  4. Anonymous (1998) Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). UK Prospective Diabetes Study (UKPDS) Group. Lancet 352:837–853

    Article  Google Scholar 

  5. Kowluru RA, Chan PS (2007) Oxidative stress and diabetic retinopathy. Exp Diabetes Res 2007:43603

    PubMed  Google Scholar 

  6. Nishikawa T, Edelstein D, Du XL et al (2000) Normalizing mitochondrial superoxide production blocks three pathways of hyperglycaemic damage. Nature 404:787–790

    PubMed  Article  CAS  Google Scholar 

  7. Kowluru RA, Kern TS, Engerman RL (1997) Abnormalities of retinal metabolism in diabetes or experimental galactosemia. IV. Antioxidant defense system. Free Radic Biol Med 22:587–592

    PubMed  Article  CAS  Google Scholar 

  8. Li W, Yanoff M, Jian B, He Z (1999) Altered mRNA levels of antioxidant enzymes in pre-apoptotic pericytes from human diabetic retinas. Cell Mol Biol (Noisy-le-grand) 45:59–66

    CAS  Google Scholar 

  9. Taniyama Y, Griendling KK (2003) Reactive oxygen species in the vasculature: molecular and cellular mechanisms. Hypertension 42:1075–1081

    PubMed  Article  CAS  Google Scholar 

  10. Du Y, Miller CM, Kern TS (2003) Hyperglycemia increases mitochondrial superoxide in retina and retinal cells. Free Radic Biol Med 35:1491–1499

    PubMed  Article  CAS  Google Scholar 

  11. Kowluru RA, Abbas SN (2003) Diabetes-induced mitochondrial dysfunction in the retina. Invest Ophthalmol Vis Sci 44:5327–5334

    PubMed  Article  Google Scholar 

  12. Kowluru RA, Odenbach S (2004) Effect of long-term administration of alpha-lipoic acid on retinal capillary cell death and the development of retinopathy in diabetic rats. Diabetes 53:3233–3238

    PubMed  Article  CAS  Google Scholar 

  13. Kowluru RA, Tang J, Kern TS (2001) Abnormalities of retinal metabolism in diabetes and experimental galactosemia. VII. Effect of long-term administration of antioxidants on the development of retinopathy. Diabetes 50:1938–1942

    PubMed  Article  CAS  Google Scholar 

  14. Kanwar M, Chan PS, Kern TS, Kowluru RA (2007) Oxidative damage in the retinal mitochondria of diabetic mice: possible protection by superoxide dismutase. Invest Ophthalmol Vis Sci 48:3805–3811

    PubMed  Article  Google Scholar 

  15. Kowluru RA, Kowluru V, Xiong Y, Ho YS (2006) Overexpression of mitochondrial superoxide dismutase in mice protects the retina from diabetes-induced oxidative stress. Free Radic Biol Med 41:1191–1196

    PubMed  Article  CAS  Google Scholar 

  16. Hammes HP, Brownlee M, Lin J, Schleicher E, Bretzel RG (1999) Diabetic retinopathy risk correlates with intracellular concentrations of the glycoxidation product Nepsilon-(carboxymethyl) lysine independently of glycohaemoglobin concentrations. Diabetologia 42:603–607

    PubMed  Article  CAS  Google Scholar 

  17. Monnier L, Mas E, Ginet C et al (2006) Activation of oxidative stress by acute glucose fluctuations compared with sustained chronic hyperglycemia in patients with type 2 diabetes. JAMA 295:1681–1687

    PubMed  Article  CAS  Google Scholar 

  18. Uhlmann K, Kovacs P, Boettcher Y, Hammes HP, Paschke R (2006) Genetics of diabetic retinopathy. Exp Clin Endocrinol Diabetes 114:275–294

    PubMed  Article  CAS  Google Scholar 

  19. Hofer SE, Bennetts B, Chan AK et al (2006) Association between PON 1 polymorphisms, PON activity and diabetes complications. J Diabetes Complications 20:322–328

    PubMed  Article  Google Scholar 

  20. Petrovic MG, Kunej T, Peterlin B, Dovc P, Petrovic D (2005) Gly482Ser polymorphism of the peroxisome proliferator-activated receptor-gamma coactivator-1 gene might be a risk factor for diabetic retinopathy in Slovene population (Caucasians) with type 2 diabetes and the Pro12Ala polymorphism of the PPARgamma gene is not. Diabetes Metab Res Rev 21:470–474

    PubMed  Article  CAS  Google Scholar 

  21. Hartnett ME, Stratton RD, Browne RW, Rosner BA, Lanham RJ, Armstrong D (2000) Serum markers of oxidative stress and severity of diabetic retinopathy. Diabetes Care 23:234–240

    PubMed  Article  CAS  Google Scholar 

  22. Uchida K (2000) Role of reactive aldehyde in cardiovascular diseases. Free Radic Biol Med 28:1685–1696

    PubMed  Article  CAS  Google Scholar 

  23. Uchida K, Kanematsu M, Morimitsu Y, Osawa T, Noguchi N, Niki E (1998) Acrolein is a product of lipid peroxidation reaction. Formation of free acrolein and its conjugate with lysine residues in oxidized low density lipoproteins. J Biol Chem 273:16058–16066

    PubMed  Article  CAS  Google Scholar 

  24. Noiri E, Yamada S, Nakao A et al (2002) Serum protein acrolein adducts: utility in detecting oxidant stress in hemodialysis patients and reversal using a vitamin E-bonded hemodialyzer. Free Radic Biol Med 33:1651–1656

    PubMed  Article  CAS  Google Scholar 

  25. Daimon M, Sugiyama K, Kameda W et al (2003) Increased urinary levels of pentosidine, pyrraline and acrolein adduct in type 2 diabetes. Endocr J 50:61–67

    PubMed  Article  CAS  Google Scholar 

  26. Tsukahara H, Sekine K, Uchiyama M et al (2003) Formation of advanced glycosylation end products and oxidative stress in young patients with type 1 diabetes. Pediatr Res 54:419–424

    PubMed  Article  CAS  Google Scholar 

  27. Uchida K, Kanematsu M, Sakai K et al (1998) Protein-bound acrolein: potential markers for oxidative stress. Proc Natl Acad Sci USA 95:4882–4887

    PubMed  Article  CAS  Google Scholar 

  28. Klein R, Klein BE, Moss SE, Cruickshanks KJ (1998) The Wisconsin Epidemiologic Study of Diabetic Retinopathy: XVII. The 14-year incidence and progression of diabetic retinopathy and associated risk factors in type 1 diabetes. Ophthalmology 105:1801–1815

    PubMed  Article  CAS  Google Scholar 

  29. Shotliff K, Duncan G (2006) Diabetic retinopathy: summary of grading and management criteria. Pract Diab Int 23:418–420

    Article  Google Scholar 

  30. Porta M, Sjoelie AK, Chaturvedi N et al (2001) Risk factors for progression to proliferative diabetic retinopathy in the EURODIAB Prospective Complications Study. Diabetologia 44:2203–2209

    PubMed  Article  CAS  Google Scholar 

  31. Zhang L, Krzentowski G, Albert A, Lefebvre PJ (2001) Risk of developing retinopathy in Diabetes Control and Complications Trial type 1 diabetic patients with good or poor metabolic control. Diabetes Care 24:1275–1279

    PubMed  Article  CAS  Google Scholar 

  32. Skipper PL, Tannenbaum SR (1990) Protein adducts in the molecular dosimetry of chemical carcinogens. Carcinogenesis 11:507–518

    PubMed  Article  CAS  Google Scholar 

  33. Maugh TH (1975) Hemoglobin: model systems shed light on oxygen binding. Science 187:154–156

    PubMed  Article  Google Scholar 

  34. Garn SM, Ryan AS, Owen GM, Abraham S (1981) Income matched black-white hemoglobin differences after correction for low transferrin saturations. Am J Clin Nutr 34:1645–1647

    PubMed  CAS  Google Scholar 

  35. Skjelbakken T, Langbakk B, Dahl IM, Lochen ML (2005) Haemoglobin and anaemia in a gender perspective: the Tromso Study. Eur J Haematol 74:381–388

    PubMed  Article  Google Scholar 

  36. Yip R, Johnson C, Dallman PR (1984) Age-related changes in laboratory values used in the diagnosis of anemia and iron deficiency. Am J Clin Nutr 39:427–436

    PubMed  CAS  Google Scholar 

  37. Kowluru RA, Koppolu P (2002) Diabetes-induced activation of caspase-3 in retina: effect of antioxidant therapy. Free Radic Res 36:993–999

    PubMed  Article  CAS  Google Scholar 

  38. Tanel A, verill-Bates DA (2007) Activation of the death receptor pathway of apoptosis by the aldehyde acrolein. Free Radic Biol Med 42:798–810

    PubMed  Article  CAS  Google Scholar 

  39. Tanel A, verill-Bates DA (2005) The aldehyde acrolein induces apoptosis via activation of the mitochondrial pathway. Biochim Biophys Acta 1743:255–267

    PubMed  Article  CAS  Google Scholar 

  40. Aldini G, le-Donne I, Facino RM, Milzani A, Carini M (2007) Intervention strategies to inhibit protein carbonylation by lipoxidation-derived reactive carbonyls. Med Res Rev 27:817–868

    PubMed  Article  CAS  Google Scholar 

  41. Anonymous (1995) The relationship of glycemic exposure (HbA1c) to the risk of development and progression of retinopathy in the diabetes control and complications trial. Diabetes 44:968–983

    Article  Google Scholar 

  42. Fu MX, Requena JR, Jenkins AJ, Lyons TJ, Baynes JW, Thorpe SR (1996) The advanced glycation end product, Nepsilon-(carboxymethyl)lysine, is a product of both lipid peroxidation and glycoxidation reactions. J Biol Chem 271:9982–9986

    PubMed  Article  CAS  Google Scholar 

  43. Boehm BO, Schilling S, Rosinger S et al (2004) Elevated serum levels of N(epsilon)-carboxymethyl-lysine, an advanced glycation end product, are associated with proliferative diabetic retinopathy and macular oedema. Diabetologia 47:1376–1379

    PubMed  Article  CAS  Google Scholar 

  44. Miura J, Yamagishi S, Uchigata Y et al (2003) Serum levels of non-carboxymethyllysine advanced glycation endproducts are correlated to severity of microvascular complications in patients with type 1 diabetes. J Diabetes Complications 17:16–21

    PubMed  Article  Google Scholar 

  45. Wautier MP, Massin P, Guillausseau PJ et al (2003) N(carboxymethyl)lysine as a biomarker for microvascular complications in type 2 diabetic patients. Diabetes Metab 29:44–52

    PubMed  Article  CAS  Google Scholar 

  46. Monnier VM, Sell DR, Genuth S (2005) Glycation products as markers and predictors of the progression of diabetic complications. Ann N Y Acad Sci 1043:567–581

    PubMed  Article  CAS  Google Scholar 

  47. Genuth S, Sun W, Cleary P et al (2005) Glycation and carboxymethyllysine levels in skin collagen predict the risk of future 10-year progression of diabetic retinopathy and nephropathy in the diabetes control and complications trial and epidemiology of diabetes interventions and complications participants with type 1 diabetes. Diabetes 54:3103–3111

    PubMed  Article  CAS  Google Scholar 

Download references


We thank the Research and Development Office, Northern Ireland, UK, the Juvenile Diabetes Research Foundation, USA, and Fight for Sight, UK, for financial support. We thank C. Patterson for statistical advice.

Duality of interest

The authors declare that there is no duality of interest associated with this manuscript.

Author information

Authors and Affiliations


Corresponding author

Correspondence to T. M. Curtis.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Zhang, X., Lai, Y., McCance, D.R. et al. Evaluation of N ε-(3-formyl-3,4-dehydropiperidino)lysine as a novel biomarker for the severity of diabetic retinopathy. Diabetologia 51, 1723–1730 (2008).

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI:


  • Acrolein
  • Advanced lipoxidation end-products
  • Diabetic retinopathy
  • FDP-lysine
  • Haemoglobin
  • Oxidative stress
  • Serum