Abstract
More than 30% of patients with type 1 diabetes develop diabetic kidney disease (DKD), which significantly increases mortality risk. The Diabetes Control and Complications Trial (DCCT) and follow-up study, Epidemiology of Diabetes Interventions and Complications (EDIC), established that glycemic control measured by HbA1c predicts DKD risk. However, the continued high incidence of DKD reinforces the urgent need for additional biomarkers to supplement HbA1c. Here, we assessed biomarkers induced by methylglyoxal (MG), a metabolic by-product that forms covalent adducts on DNA, RNA, and proteins, called MG adducts. Urinary MG adducts were measured in samples from patients with type 1 diabetes enrolled in DCCT/EDIC who did (case patients; n = 90) or did not (control patients; n = 117) develop DKD. Univariate and multivariable analyses revealed that measurements of MG adducts independently predict DKD before established DKD biomarkers such as glomerular filtration rate and albumin excretion rate. Elevated levels of MG adducts bestowed the greatest risk of developing DKD in a multivariable model that included HbA1c and other clinical covariates. Our work establishes a novel class of biomarkers to predict DKD risk and suggests that inclusion of MG adducts may be a valuable tool to improve existing predictors of complications like DKD prior to overt disease, and to aid in identifying at-risk individuals and personalized risk management.
Article Highlights
Diabetic kidney disease (DKD) is a common cause of death in patients with type 1 diabetes, and there is an unmet need to identify novel biomarkers to predict DKD.
We sought to elucidate if methylglyoxal adducts, which are a marker of altered metabolism, have predictive utility for DKD in patients with type 1 diabetes.
We found that methylglyoxal adducts predict the risk of DKD at least 16 years before diagnosis.
Our data present a novel class of biomarkers with potential utility for predicting DKD risk to supplement existing tools such as HbA1c, albumin excretion rate, and glomerular filtration rate.
Graphical Abstract
Introduction
Diabetic kidney disease (DKD) is a microvascular complication affecting ∼30% of patients with type 1 diabetes. DKD increases mortality risk because the disease is often not diagnosed until significant kidney damage defined by macroalbuminuria (>300 mg/day urinary albumin) and decreased glomerular filtration rate (GFR; <60 mL/min/1.73 m2) occur (1–5). This late diagnosis increases the risk of additional complications, thus enhancing morbidity and mortality (6). Although intensive glycemic control is the standard of care for patients with type 1 diabetes, it is often insufficient in preventing DKD. This may be because, at least partially, metabolic flux is not captured by HbA1c, the difficulties in maintaining HbA1c of 5–6% (31–42 mmol/mol), and inaccurate HbA1c measurements in patients with hemoglobinopathies (7). Furthermore, HbA1c only provides a 3-month snapshot of blood glucose levels and does not reflect longer-term changes. To improve personalized disease management and patient outcomes, there is an unmet need to identify novel biomarkers to predict DKD risk before irreversible kidney damage occurs.
A potential source of DKD biomarkers is the by-products of metabolic pathways affected by dysglycemia, including glucose, protein, and lipid metabolism. These pathways, which are dysregulated in patients with diabetes, form methylglyoxal (MG), a reactive by-product that induces stable adducts on DNA, RNA, and proteins (8). MG is proposed to be the most important in vivo glycation agent and may offer a more comprehensive view of metabolic changes compared with HbA1c (9). MG modifies DNA to form N2-carboxyethyl-2′-deoxyguanosine (CEdG), the most abundant MG-DNA adduct in vivo (Fig. 1A) (10). We recently demonstrated that MG also modifies RNA to form N2-carboxyethyl-guanosine (CEG) (Fig. 1A) (11). MG reacts with lysine on proteins to form several adducts, including carboxyethyllysine (CEL) (Fig. 1A) (12). The formation and clinical relevance of MG and its adducts were recently reviewed (8).
Figure 1.
Study schematic. A) MG reacts with deoxyguanosine, guanosine, and lysine, forming the covalent adducts CEdG, CEG, and CEL, respectively. dR, deoxyribose, R, ribose. B) Graphical abstract of the DCCT/EDIC trial. Patients were assigned a treatment regimen: conventional (CON, CONV) or intensive (INT) therapy. Cases are patients who progressed to DKD during EDIC, and controls are patients who did not progress to DKD either during the DCCT or EDIC. C) An aliquot of frozen urine was thawed and spiked with a fixed amount of 15N5-CEdG, 15N5-CEG, d4-CEL, and d3-CRE. Analytes were enriched using cation exchange solid-phase extraction followed by analysis using multiplex liquid chromatography–tandem mass spectrometry (LC-MS/MS). D) Binning strategy for grouping samples for univariate and multivariable analysis.
Prior literature supports investigating MG adducts as predictors of DKD; MG adducts such as CEdG, CEG, CEL, carboxymethyllysine, and hydroimidazolones are correlated with diabetes and complications (11,13–16). However, a limitation of these studies is that adducts were measured after DKD diagnosis, preventing the study of their predictive ability. Here, we sought to determine the association of MG adducts with DKD before diagnosis to identify a novel class of biomarkers for predicting DKD risk.
Research Design and Methods
Study Design
Urine samples from patients in the DCCT/EDIC trial were obtained from the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) Biorepository (Fig. 1B). We designed a nested case-control study, designating controls as patients who did not develop DKD during DCCT or EDIC, and cases as patients who developed DKD during EDIC (Fig. 1B). Patients who developed DKD during the DCCT were excluded from our cohort. DKD was diagnosed using standard clinical practices (of at least two measures of macroalbuminuria, with albumin excretion rate [AER] >300 mg/day). Samples from 207 patients (n = 117 control patients and 90 case patients) were used based on biorepository availability and DCCT/EDIC committee approval. Clinical variables are referenced in Table 2 but were not used to match case patients with control patients.
Table 2.
Multivariable logistic regression of CEG and clinical variables with DKD progression
| Characteristic | Years 1–3 (case, n = 68; control, n = 70) | ||
|---|---|---|---|
| OR | 95% CI | P value | |
| Age (years) | 1.0 | 0.9, 1.1 | 0.48 |
| Sex | |||
| Female | — | — | |
| Male | 3.7 | 1.2, 12.1 | 0.025 |
| Group | |||
| Conventional (CONV) | — | — | |
| Intensive (INT) | 0.6 | 0.2, 2.1 | 0.39 |
| BMI (kg/m2) | 1.4 | 1.1, 1.7 | 0.002 |
| HbA1c (% or mmol/mol) | 1.7 | 1.2, 2.6 | 0.007 |
| Triglycerides (mg/dL) | 1.0 | 1.0, 1.0 | 0.28 |
| GFR (mL/min) | 1.0 | 1.0, 1.1 | 0.054 |
| DUR (years) | 1.0 | 1.0, 1.0 | 0.51 |
| AER (mg/day) | 1.0 | 1.0, 1.1 | 0.009 |
| MBP (mmHg) | 1.0 | 0.9, 1.0 | 0.56 |
| CEG (pmol/µmol CRE) | 6.4 | 3.0, 16.5 | <0.001 |
CEG is independently and significantly associated with DKD progression in a multivariable logistic regression model with other clinical variables in consideration.
Variables at DCCT baseline between DCCT and NCCT cohorts were compared to assess if NCCT was representative of the DCCT cohort. Significant differences in age (NCCT: 25.4 ± 7.7 years; DCCT: 27.0 ± 7.1 years; P = 0.007) and HbA1c (NCCT: 78.56 ± 17.36 mmol/mol; DCCT: 76 ± 17.5 mmol/mol; P = 0.048) were found, but not in diabetes duration (DUR) or BMI (Supplementary Table 1). Further breakdown of NCCT among the four subgroups, Case-Conventional (Case-CONV), Case-Intensive (Case-INT), Control-Conventional (Control-CONV), and Control-Intensive (Control-INT), is given in Supplementary Table 2.
Multiplex Mass Spectrometry Method to Measure MG Adducts
CEdG, CEG, creatinine (CRE), and CEL were quantified via multiplex liquid chromatography–mass spectrometry using isotopically labeled standards, as previously described (11,17,18) (Fig. 1C). Urine aliquots were spiked with fixed amounts of 15N5-R,S-CEdG, 15N5-R,S-CEG, CEL-d4, and CRE-d3. Technical triplicates were processed with replicate variation <5%. Each urine aliquot was analyzed twice independently. Representative chromatograms for CEdG and CEG are shown in Supplementary Fig. 1. CRE was used to normalize levels of MG adducts to account for discrepancies in urine concentration because of its excretion at a relatively constant rate. CEL and CRE were quantified similarly with the following modifications: the gradient used was 0–2 min 0% B, 2–2.5 min, 98% B, and 2.5–3 min 0% B with a constant flow rate of 0.2 mL/min. Mass transitions were monitored in positive-ion mode: CEL m/z 219 to 130; CEL-d4 m/z 223 to 134; CRE m/z 114 to 86; and CRE-d3 m/z 117 to 89. The retention time of CEL and CEL-d4 was 0.72 min, and 0.76 min for CRE and CRE-d3 (Supplementary Fig. 2).
Statistical Analysis
To account for different urine collection time points during DCCT/EDIC, we used a binning system with a 3-year sliding window for DCCT years 1–3, 4–6, and 7–9 (Fig. 1D). Where MG adduct measurements were available, each patient was sampled once per bin and at the closest to the end of the window, but this did not exclude patients from being sampled in subsequent bins. Therefore, each patient is only represented once in each bin. This approach allowed us to achieve a higher sampling population within bins to maintain statistical power in our analyses. The final case and control patient numbers for years 1–3, 4–6, and 7–9 were 68 and 70, 63 and 83, and 10 and 83, respectively (Fig. 1D). For each patient, clinical covariates such as BMI, HbA1c, levels of triglycerides (TRIG), GFR, DUR, AER, and mean blood pressure (MBP), measured at the same time as MG adducts, were used. Data are presented as mean ± SD.
Univariate logistic regression was performed using case and control status as the response variable with one independent variable in each model (Table 2). MG adducts were log2-transformed to account for right-sided skew. Univariate analyses were reported with 95% CIs, P values, and odds ratios (ORs); P < 0.05 was considered statistically significant. ORs were based on one-unit change in the independent variable. The pairwise correlation among covariates was assessed using Pearson correlation. For each MG adduct, we performed one multivariable logistic analysis with the following covariates: age, sex, treatment group, BMI, HbA1c, TRIG, GFR, DUR, AER, and MBP. Dashes in tables (—) represent analysis with a binary variable.
Data and Resource Availability
Data will be deposited with the NIDDK Central Repository per the agreement in NIDDK award X01DK118575 (to S.C.S.). Data reported here from the DCCT and EDIC studies are available on request at the NIDDK Central Repository website, Resources for Research (https://repository.niddk.nih.gov/).
Results
MG Adducts Are Significantly Associated With DKD Progression in a Univariate Model
To determine the association of clinical variables and MG adducts with DKD risk, we quantified MG adducts in 3-year bins spanning the DCCT (Fig. 1D). We focused on years 1–3 to study the time point closest to type 1 diabetes diagnosis and furthest from DKD diagnosis. At years 1–3, case patients had significantly higher concentrations of MG adducts than control patients (Fig. 2A–C). Univariate analysis of years 1–3 revealed significant association of CEdG, CEG, and CEL with DKD progression (CEdG: OR 1.8, P < 0.0001; CEG: OR 5.3, P < 0.0001; CEL: OR 1.5, P < 0.0001) (Table 1). Comparison of time of DKD diagnosis with patient adduct measurement in years 1–3 revealed that MG adducts were associated with DKD at least 16.3 ± 4.6 years before DKD diagnosis, supporting the utility of the MG adducts in early DKD risk prediction. In addition to MG adducts, HbA1c, male sex, DUR, and AER were also significantly associated with DKD progression (Table 1). MG adducts and HbA1c had the strongest association with DKD progression (P < 0.001). However, MG adducts had a higher OR (CEdG OR 1.8; CEG OR 5.3; CEL OR 1.5) than did AER or GFR (for both, OR 1.0) (Table 1). HbA1c had an OR of 1.8 (Table 1). In years 4–6 and 7–9, CEG was consistently correlated with a higher OR than HbA1c (Supplementary Tables 3 and 4). Additional longitudinal analysis revealed elevated levels of MG adducts in case patients compared with control patients over 10 years of the DCCT, with 95% CIs (Fig. 2D–F).
Figure 2.
Levels of MG adducts are elevated during the DCCT in patients who will develop DKD. Urinary CEG (A), CEdG (B), and CEL (C) levels are significantly elevated (P < 0.0001) at years 1–3 in patients who later developed DKD (case) during the EDIC compared with those who did not (control). Outliers were removed using robust regression and outlier removal (ROUT), and groups were compared using unpaired t tests with Welch’s correction. Urinary CEG (D), CEdG (E), and CEL (F) levels are elevated throughout the DCCT in case patients compared with control patients. The 95% CIs were plotted and obtained using ggplot2 software in R (R Foundation for Statistical Computing).
Table 1.
Univariate association analysis of clinical variables and MG adducts with DKD progression
| Characteristic | Years 1–3 (case, n = 68; control, n = 70) | ||
|---|---|---|---|
| OR | 95% CI | P value | |
| Sex | 0.002 | ||
| Female | — | — | |
| Male | 3.1 | 1.5, 6.5 | |
| Race | 0.62 | ||
| White | — | — | |
| Non-White | 1.6 | 0.2, 12.2 | |
| Group | 0.003 | ||
| Conventional | — | — | |
| Intensive | 0.4 | 0.2, 0.7 | |
| Age (years) | 1.0 | 1.0, 1.1 | 0.55 |
| BMI (kg/m2) | 1.1 | 1.0, 1.2 | 0.19 |
| HbA1c (% or mmol/mol) | 1.8 | 1.4, 2.2 | <0.001 |
| Cholesterol (mg/dL) | 1.0 | 1.0, 1.0 | 0.32 |
| TRIG (mg/dL) | 1.0 | 1.0, 1.0 | 0.6 |
| GFR (mL/min) | 1.0 | 1.0, 1.0 | 0.24 |
| DUR (years) | 1.0 | 1.0, 1.0 | 0.01 |
| Insulin (units/kg) | 0.5 | 0.1, 2.0 | 0.32 |
| AER (mg/day) | 1.0 | 1.0, 1.1 | 0.021 |
| CLR (mL/min/1.73 m2) | 1.0 | 1.0, 1.0 | 0.56 |
| MBP (mmHg) | 1.0 | 0.9, 1.0 | 0.24 |
| ANYCCN | 0.94 | ||
| 0 | — | — | |
| 1 | 1.0 | 0.4, 2.6 | |
| CEG (pmol/µmol CRE) | 5.3 | 3.0, 10.0 | <0.001 |
| CEDG (pmol/µmol CRE) | 1.8 | 1.3, 2.4 | <0.001 |
| CEL (pmol/µmol CRE) | 1.5 | 1.2, 1.9 | <0.001 |
| ETDRSPAT.cat | 0.15 | ||
| 0 | — | — | |
| 1 | 0.5 | 0.1, 1.9 | |
| 2 | 0 | ||
CEdG, CEG, and CEL are significantly associated with DKD progression. ANYCCN, neuropathy (0 = no; 1 = yes); CLR, creatinine clearance; ETDRSPAT.cat = retinopathy (0 = none; 1–2 = mild).
MG Adducts Are Independently, Significantly Associated With DKD Progression in a Multivariable Model
To determine the impact of clinical covariates on the association of MG adducts with DKD, we performed multivariable logistic regression with age, sex, treatment group, BMI, HbA1c, TRIG, GFR, AER, MBP, and each MG adduct independently. CEdG, CEG, and CEL remained significantly associated with DKD progression independent of covariates, with OR values of 1.9 (P = 0.001), 6.4 (P < 0.001), and 1.4 (P = 0.036), respectively, at years 1–3. CEG, CEdG, and CEL values at years 1–3 are presented in Table 2 and Supplementary Tables 5 and 6, respectively. Data for the remaining years are presented in Supplementary Table 7. Across years 1–9, all three MG adducts had higher ORs than AER or GFR, which remained at an OR of 1.0 during the DCCT. CEG also had a higher OR than HbA1c throughout the DCCT (Table 2, Supplementary Table 7). A multivariable analysis containing all three MG adducts was not performed, because of multicollinearity between MG adducts.
Discussion
Diabetes-related complications are significant health concerns to patients with type 1 diabetes, despite intensive insulin therapy. Although maintaining glycemic control is the standard of care for these patients, many struggle to achieve this, and, by itself, maintaining glycemic control is not sufficient to prevent DKD. Existing DKD biomarkers include HbA1c, a marker of glycemic control, and AER and GFR, markers of kidney function. However, HbA1c may be unable to capture the full metabolic landscape, and AER and GFR offer limited utility until they surpass clinical thresholds for DKD diagnosis, at which point significant kidney damage has occurred.
In univariate and multivariable analyses, elevated levels of MG adducts consistently associated more significantly with DKD progression and corresponded to higher ORs than AER and GFR, independently of HbA1c, suggesting MG adducts may enhance the predictive power of either alone. We report that MG adducts demonstrate significant association with DKD at least 16 years before diagnosis, lending support to measuring MG adducts for early detection of DKD. Notably, CEG maintained a significant association with DKD throughout the DCCT and had a higher OR than GFR, AER, and HbA1c, supporting its potential to be a stronger biomarker candidate for DKD risk. Furthermore, ORs for AER and GFR remained 1.0 throughout the DCCT, suggesting they have limited predictive utility and minimal association with DKD progression at these time points, a direct contrast to the ORs for MG adducts at the same time points. HbA1c also was significantly associated with DKD in univariate analysis. We recapitulated previous associations between CEL and DKD progression (16). However, these prior findings only included sex, HbA1c, and DUR in multivariable analysis, whereas our model also accounted for age, BMI, treatment, TRIG, GFR, AER, and MBP. Furthermore, our results indicate CEL does not remain significant in multivariable analysis over time, limiting its predictive capability.
Our results reveal a promising modality of biomarkers for diabetic complications and other metabolic diseases. An advantage of MG adducts is their noninvasive measurement (urine), a direct contrast to HbA1c (blood). Furthermore, HbA1c is conventionally measured via liquid chromatography, whereas MG adducts are measured via liquid chromatography–mass spectrometry, potentially offering higher sensitivity and specificity.
There are limitations in our study to be considered, however. Although DCCT/EDIC had excellent patient recruitment and retention with regular sampling, NCCT was limited to samples that would not deplete archived samples. An additional limitation was an inequality within NCCT regarding sex, age, and ethnic diversity. Similarly, it is of interest to measure MG adducts in people without diabetes and with type 2 diabetes to assess association with their DKD risk. Furthermore, cross-cohort validations across larger diverse populations of patients of different ethnic, socioeconomic, geographic, and racial backgrounds are warranted. Another limitation is that our study defined “cases” and “controls” by AER, a canonical marker for DKD diagnosis. DKD was diagnosed using standard clinical testing based on AER and/or albumin-to-CRE ratio with repeated elevation. However, patients with AER 299 mg/day were not diagnosed with DKD and therefore were excluded, although they may have had significant kidney damage. This extends to patients who experience AER flux, which may qualify them for DKD diagnosis at one time point, but subsequently regress below the 300 mg/day threshold later. We are also interested in identifying ways to normalize MG adducts to total excreted DNA or RNA to control for changes in DNA or RNA excretion independent of MG adducts.
Our findings indicate that MG adducts, which reflect systemic MG burden, may provide a more comprehensive indication of metabolic changes than captured via HbA1c, which only indicates the glucose environment of erythrocytes. A direction of interest is uncovering the dynamics of MG adduct levels in response to changes in blood glucose levels and HbA1c. Additionally, we are working toward validating these findings using cross-cohort analyses. The discovery of the RNA MG adduct and the use of MG adducts, particularly CEG, as a predictive tool for DKD represents a step forward in early risk detection and intervention for diabetic complications such as DKD.
This article contains supplementary material online at https://doi.org/10.2337/figshare.24550600.
Article Information
Funding. This work was supported by National Institutes of Health (NIH; grants R21DK127285 to S.C.S.; R01DK065073 and R01DK081705 to R.N.; and R01CA176611 to J.T.), with additional support from the Diabetic Complications Consortium (to S.C.S.) and the Wanek Family Project for the Cure of Type 1 Diabetes at City of Hope (to S.C.S., R.N., and Z.C.). Research reported in this publication included work performed in the Integrated Mass Spectrometry Core supported by the National Cancer Institute of the NIH under award number P30CA033572. Archived samples were obtained from the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) Central Repository through NIDDK award X01DK118575 (to S.C.S.).
The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.
Duality of Interest. R.N. and Z.C. are members of the DCCT/EDIC Study Group. No other potential conflicts of interest relevant to this article were reported.
Author Contributions. S.W.T.L. performed data analysis and data interpretation, prepared figures, and wrote the manuscript. C.H.-C. processed samples and performed data analysis. E.D.J.L.G. prepared synthetic standards. T.Z. processed samples. M.T. performed statistical analysis. N.P. organized data. Z.C. and B.O.R. provided scientific input. J.S.K. provided bioinformatics support, data, statistical analysis, and scientific input. R.N. and J.T. provided experimental planning, data interpretation, and scientific input. S.C.S. performed data analysis, experimental planning, and data interpretation, prepared figures, supervised the laboratory, and wrote the manuscript. All authors edited, reviewed, and approved the final version of the manuscript. S.C.S. and J.T. are the guarantors of this work and, as such, had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Prior Presentation. Parts of this study were presented by S.W.T.L. in poster form at the 83rd Scientific Sessions of the American Diabetes Association, San Diego, CA, 23–26 June 2023.
Funding Statement
This work was supported by National Institutes of Health (NIH; grants R21DK127285 to S.C.S.; R01DK065073 and R01DK081705 to R.N.; and R01CA176611 to J.T.), with additional support from the Diabetic Complications Consortium (to S.C.S.) and the Wanek Family Project for the Cure of Type 1 Diabetes at City of Hope (to S.C.S., R.N., and Z.C.). Research reported in this publication included work performed in the Integrated Mass Spectrometry Core supported by the National Cancer Institute of the NIH under award number P30CA033572. Archived samples were obtained from the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) Central Repository through NIDDK award X01DK118575 (to S.C.S.).
References
- 1. Afkarian M, Sachs MC, Kestenbaum B, et al. Kidney disease and increased mortality risk in type 2 diabetes. J Am Soc Nephrol 2013;24:302–308 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Groop PH, Thomas MC, Moran JL, et al. ; FinnDiane Study Group . The presence and severity of chronic kidney disease predicts all-cause mortality in type 1 diabetes. Diabetes 2009;58:1651–1658 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Livingstone SJ, Looker HC, Hothersall EJ, et al. Risk of cardiovascular disease and total mortality in adults with type 1 diabetes: Scottish registry linkage study. PLoS Med 2012;9:e1001321. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Orchard TJ, Secrest AM, Miller RG, Costacou T.. In the absence of renal disease, 20 year mortality risk in type 1 diabetes is comparable to that of the general population: a report from the Pittsburgh Epidemiology of Diabetes Complications Study. Diabetologia 2010;53:2312–2319 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Orchard TJ, Costacou T, Kretowski A, Nesto RW.. Type 1 diabetes and coronary artery disease. Diabetes Care 2006;29:2528–2538 [DOI] [PubMed] [Google Scholar]
- 6. Kovesdy CP. Epidemiology of chronic kidney disease: an update 2022. Kidney Int Suppl (2011) 2022;12:7–11 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Gunton JE, McElduff A.. Hemoglobinopathies and HbA(1c) measurement. Diabetes Care 2000;23:1197–1198 [DOI] [PubMed] [Google Scholar]
- 8. Lai SWT, Lopez Gonzalez EJ, Zoukari T, Ki P, Shuck SC.. Methylglyoxal and its adducts: induction, repair, and association with disease. Chem Res Toxicol 2022;35:1720–1746 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Rabbani N, Thornalley PJ.. Measurement of methylglyoxal by stable isotopic dilution analysis LC-MS/MS with corroborative prediction in physiological samples. Nat Protoc 2014;9:1969–1979 [DOI] [PubMed] [Google Scholar]
- 10. Shuck SC, Wuenschell GE, Termini JS.. Product studies and mechanistic analysis of the reaction of methylglyoxal with deoxyguanosine. Chem Res Toxicol 2018;31:105–115 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Shuck SC, Achenbach P, Roep BO, et al. Methylglyoxal products in pre-symptomatic type 1 diabetes. Front Endocrinol (Lausanne) 2023;14:1108910. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Ahmed MU, Brinkmann Frye E, Degenhardt TP, Thorpe SR, Baynes JW.. N-epsilon-(carboxyethyl)lysine, a product of the chemical modification of proteins by methylglyoxal, increases with age in human lens proteins. Biochem J 1997;324:565–570. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Waris S, Winklhofer-Roob BM, Roob JM, et al. Increased DNA dicarbonyl glycation and oxidation markers in patients with type 2 diabetes and link to diabetic nephropathy. J Diabetes Res 2015;2015:915486. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Beisswenger PJ, Makita Z, Curphey TJ, et al. Formation of immunochemical advanced glycosylation end products precedes and correlates with early manifestations of renal and retinal disease in diabetes. Diabetes 1995;44:824–829 [DOI] [PubMed] [Google Scholar]
- 15. Sun JK, Keenan HA, Cavallerano JD, et al. Protection from retinopathy and other complications in patients with type 1 diabetes of extreme duration: the Joslin 50-Year Medalist Study. Diabetes Care 2011;34:968–974 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Beisswenger PJ, Howell SK, Russell GB, Miller ME, Rich SS, Mauer M.. Early progression of diabetic nephropathy correlates with methylglyoxal-derived advanced glycation end products. Diabetes Care 2013;36:3234–3239 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Takahashi N, Boysen G, Li F, Li Y, Swenberg JA.. Tandem mass spectrometry measurements of creatinine in mouse plasma and urine for determining glomerular filtration rate. Kidney Int 2007;71:266–271 [DOI] [PubMed] [Google Scholar]
- 18. Mazza MC, Shuck SC, Lin J, et al. DJ-1 is not a deglycase and makes a modest contribution to cellular defense against methylglyoxal damage in neurons. J Neurochem 2022;162:245–261 [DOI] [PMC free article] [PubMed] [Google Scholar]