Circulating Concentrations of Soluble Receptor for AGE Are Associated With Age and AGER Gene Polymorphisms in Children With Newly Diagnosed Type 1 Diabetes

A variety of environmental agents have been studied for their potential role in the pathogenesis of type 1 diabetes (1). To date, there is limited evidence of the significance of specific nutritional factors in the etiology of type 1 diabetes (2). Advanced glycation end products (AGEs) are novel dietary candidates, which could contribute to the autoimmune process leading to type 1 diabetes. AGEs are formed by nonenzymatic glycation reactions, and they are present in excess amounts in western style diets (3). AGEs are also produced endogenously within the human body, although the physiological role of the endogenously produced AGEs remains unknown (3). Pathologically, excess accumulation of AGEs contributes to a variety of microvascular and macrovascular complications of diabetes (4). There is also evidence that AGEs and the glycation reaction promote β-cell dysfunction and apoptosis (5,6). Furthermore, dietary restriction of AGEs reduced the incidence of autoimmune diabetes in an animal model, the NOD mice (7). A previous study (8) has also identified an association between AGEs and the appearance of islet autoantibodies.

AGEs can exert their biological effects by engaging the receptor for AGEs (RAGEs) (9). The RAGE is a member of the Ig superfamily of cell surface molecules and a multiligand receptor involved in immune and inflammatory responses (10). The AGE-specific receptor gene (AGER) encoding RAGEs is located on the short arm of chromosome 6 within the HLA class III region near the junction with the class II loci, which are the main loci associated with risk for type 1 diabetes (11). Earlier studies have implicated that specific AGER polymorphisms are associated with increased risk of type 1 diabetes (12,13) and complications of diabetes (14). In addition, AGER polymorphisms correlate with the circulating soluble RAGE (sRAGE) concentrations in adult populations (15,16). While signaling of the full-length membrane-bound form of RAGEs is involved in various pathological conditions and inflammation, the circulating forms of the molecule, the endogenous secretory RAGEs (esRAGEs) generated by alternative splicing of the AGER gene transcript (17) and the sRAGEs generated from the proteolytic cleavage of the full-length membrane-bound RAGEs, are considered to be cytoprotective (18). The blockade of RAGE activation and RAGE/ligand interaction via introducing external sRAGEs prevents the adoptive transfer of autoimmune diabetes in an animal model (19).

The incidence of type 1 diabetes has been increasing among Finnish children over the last few decades (20), although an encouraging plateau in the rate over the last few years was reported recently (21). Based on nationwide register data, the annual incidence of type 1 diabetes increased from 31.4 per 100,000 children under 15 years of age in 1980 to the peak incidence of 64.9 per 100,000 children in 2006 (21). Polymorphisms of AGER and the biological effects of AGEs offer an interesting link among genetic disease susceptibility, aspects of specific environmental factors, and the increasing incidence rate of type 1 diabetes seen in most developed countries. We defined three single nucleotide polymorphisms (SNPs) of AGER and analyzed the circulating sRAGE concentrations in a large population of children and adolescents under 15 years of age with newly diagnosed type 1 diabetes to assess the possible associations among sRAGE levels, AGER genotype, and the disease phenotype of the patients.

The study subjects were derived from the Finnish Pediatric Diabetes Register and Biobank. Samples from children and adolescents with newly diagnosed type 1 diabetes and from their first-degree relatives are collected for the Biobank. Data on family history of type 1 diabetes and other forms of diabetes, and on the degree of metabolic decompensation at the time of diagnosis are collected for the register with a structured questionnaire (22). The questionnaire does not include data on ethnicity, but <3% of the families reported roots outside Finland, and 75% of those families came from other European countries. Islet cell antibodies, insulin autoantibodies (IAAs), autoantibodies to GAD, islet antigen 2, and zinc transporter 8 (ZnT8A) are measured from the serum samples of the affected children and their family members. The autoantibody measurement protocol has been described previously (23). HLA-DR/DQ genotypes are analyzed as well. The parents of the children gave their written informed consent and children ≥10 years of age gave their written assent. The protocol was approved by the Ethical Committee of the Hospital District of Helsinki and Uusimaa. The population in this study includes 2,115 children in whom type 1 diabetes had been diagnosed between 2 February 2002 and 7 January 2009. The blood samples were drawn within 1 month after the diagnosis (median time 5 days). A majority of these subjects were males (1,194; 56%), and the mean age at diagnosis was 7.96 years (median age 8.16 years, age range 0.28–14.98 years). The control population comprised 316 healthy children and adolescents who had negative test results for diabetes-associated autoantibodies. They were siblings of nondiabetic children who carried HLA-conferred susceptibility to type 1 diabetes and took part in the Finnish Type 1 Diabetes Prediction and Prevention study (24). At the time of the recruitment of the index child to the Finnish Type 1 Diabetes Prediction and Prevention study, there were five cases (<2%) of parental type 1 diabetes among the families of the control children. The age, sex distribution, and distribution of HLA genotypes were similar in the study population and in the control subjects (Table 1).

Serum samples were assayed undiluted according to the manufacturer’s instructions (Human RAGE ELISA; R&D Systems, Minneapolis, MN). The interassay coefficient of variation was 7.6%, whereas the intra-assay coefficient of variation was 3.5%. The analysis covers the whole pool of circulating sRAGEs, both the esRAGE and sRAGE components.

AGER genotyping was performed in 1,444 patients with the Sequenom multiplex platform (Sequenom, Hamburg, Germany) according to the manufacturer instructions. To ensure the quality of Sequenom genotyping, all three SNP markers were tested for Mendelian inconsistencies (where the genotypes of parents were available), Hardy-Weinberg equilibrium, and missingness of genotyping results. Mendelian and missingness tests were based on a larger set of 72 SNPs (from AGER and other genomic regions). All markers were in Hardy-Weinberg equilibrium, and per-marker missingness of genotyping results was on an acceptable level of <2%. Families with >2% Mendelian errors or >50% missingness in 72 SNPs were removed from the analysis (<4%), leaving 1,390 case patients who passed the Sequenom quality control.

HLA typing of the major predisposing and protective DR-DQ haplotypes was performed with a PCR-based lanthanide-labeled hybridization method using time-resolved fluorometry for detection. The presence of the (DR3)-DQA1*05-DQB1*02 haplotype is shortened to DR3 and that of HLA-DRB1*04-DQB1*0302 to DR4, according to convention.

Blood pH, plasma glucose, and β-hydroxybutyrate concentrations were available for 2,077, 2,092, and 1,739 subjects, respectively. Height and weight data were available from 2,073 of the subjects with type 1 diabetes. Age-adjusted BMI was calculated according to World Health Organization growth charts (25). Information on the type 1 and type 2 diabetes status of the father and mother of the index case patient at the time of diagnosis was available in 2,065 and 2,105 of the 2,115 subjects, respectively. According to the register data, 80% of the index case patients (1,698 case patients) had siblings at diagnosis, and information on type 1 and type 2 diabetes status of the latter was available in 1,695 families. Data on AGER and HLA genotypes of both parents were available for 1,761 diabetic subjects (1,390 of them overlapping with the 2,115 subjects analyzed for sRAGE concentrations), which were used to infer (pseudo) control allele frequencies of AGER and HLA.

All statistical analyses were performed with the SPSS version 21.0 (IBM, Chicago, IL). Comparisons between groups were made using the Student t test for independent samples, or ANOVA followed by post hoc analysis with Tukey test or Tamhane test when appropriate, if there were more than two groups. Differences in proportions between groups were tested using the χ² test. Correlations were analyzed using the Pearson product moment correlation coefficient (r) or the Spearman rank correlation test (r_s). Logistic or linear regression was used when appropriate to assess the unique contribution of a specific variable to the model. For all analyses, P values <0.05 were considered significant.

There was no significant difference in the circulating sRAGE concentrations between boys and girls with type 1 diabetes (1,173 vs. 1,168 pg/mL, P = 0.80). Neither could any difference be seen in the circulating sRAGE concentrations between the patients with newly diagnosed type 1 diabetes and the control subjects (1,171 vs. 1,153 pg/mL, P = 0.48). However, there was a modest correlation between age at diagnosis and the sRAGE serum concentrations (r = 0.10, P < 0.001) among the affected children. When the subjects with type 1 diabetes were divided into tertiles according to age, the youngest age group (i.e., children <5 years of age) had the lowest sRAGE concentrations (1,113 pg/mL), and the difference was significant when compared with children 5–9.99 years of age (1,178 pg/mL, P = 0.01) or children and adolescents in the oldest age group (10–14.99 years of age, 1,208 pg/mL, P < 0.001). Children in whom type 1 diabetes had been diagnosed before 2 years of age had even lower concentrations, when compared with the older children (1,027 vs. 1,181 pg/mL, P < 0.001). The sRAGE concentrations of the children divided into quintiles according to age are presented in Fig. 1. There was no significant correlation between age and the sRAGE concentrations in the control population (r = −0.006, P = 0.95). The mean sRAGE concentrations in the age tertile groups were similar (1,150, 1,155, and 1,152 pg/mL, respectively, P = 0.996). The nondiabetic children <2 years of age had slightly higher concentrations than the rest of the control population (1,329 vs. 1,140 pg/mL, P = 0.04). In fact, when only the children <2 years of age were included in the analysis, there was a significant difference in the circulating sRAGE concentrations between the case patients (n = 133) and the control subjects (n = 21) (1,027 vs. 1,329 pg/mL, P = 0.005).

Approximately 18% of the patients with the blood pH information available (n = 365) had diabetic ketoacidosis (DKA), which was defined as a blood pH of <7.30, and 4% of the patients (n = 86) had severe DKA, which was defined as a blood pH of <7.10 at diagnosis. Both DKA and severe DKA were associated with lower sRAGE concentrations (1,081 vs. 1,190 pg/mL, P < 0.001; and 1,009 vs. 1,178 pg/mL, P < 0.001, respectively). The associations among metabolic state at diagnosis, demographic data, and sRAGE concentrations are presented in Table 2.

There was no association between the sRAGE serum concentrations and the number of autoantibodies detectable at the diagnosis of type 1 diabetes (P = 0.31). Children testing positive for ZnT8A had higher concentrations than the other index case patients (1,196 vs. 1,128 pg/mL, P < 0.001). IAA-positive subjects had lower concentrations than the IAA-negative subjects (1,137 vs. 1,198 pg/mL, P = 0.001), but the association was no longer significant when the effect of age at diagnosis was considered (P = 0.09). There were no significant associations between the sRAGE concentrations and the positivity or titers of the other diabetes-associated autoantibodies. The associations between islet autoantibodies and sRAGE at diagnosis are presented in Table 2.

When both the study population and the control subjects were analyzed together, subjects with the highest genetic risk of type 1 diabetes, the HLA-DR3/DR4 heterozygotes, had somewhat lower sRAGE concentrations than the rest of the population (1,127 vs. 1,180 pg/mL, P = 0.008). When logistic regression was performed with the presence of the HLA-DR3/DR4 genotype as the dependent variable, the relationship with sRAGE concentration remained significant (P = 0.02), with age at diagnosis and sex as other covariates. Carriers of the HLA-DR3 haplotype had lower circulating sRAGE levels as well (1,131 vs. 1,190 pg/mL, P = 0.001), and the difference remained significant (P = 0.002) when the effects of age and sex were taken into account in a logistic regression analysis. The inverse relationship between the circulating sRAGE concentrations and the HLA-DR3/DR4 genotype was seen both in the population with type 1 diabetes (1,133 vs. 1,181 pg/mL, P = 0.02) and the control population, although it was not significant in the latter (1,084 vs. 1,171 pg/mL, P = 0.13). The HLA-DR3 haplotype was associated with lower sRAGE concentrations in both populations (diabetic population 1,144 vs. 1,186 pg/mL [P = 0.03]; control subjects 1,033 vs. 1,214 pg/mL [P < 0.001]). Carriers of the HLA-DR4 haplotype had higher serum concentrations in the control population (1,186 vs. 1,055 pg/mL, P = 0.01). In the study population, the difference remained nonsignificant (1,175 vs. 1,152 pg/mL, P = 0.64). However, the carriers of the HLA-DR4/non-DR3 genotype had higher concentrations both among the population of patients with type 1 diabetes (1,192 vs. 1,152 pg/mL, P = 0.03) and in the control population (1,224 vs. 1,070 pg/mL, P = 0.001).

When three SNPs of the AGER gene previously associated with an increased risk of type 1 diabetes (rs2070600, rs9469089, and rs17493811) were analyzed in 1,390 subjects with type 1 diabetes, we observed that rs2070600 and rs9469089 were strongly associated with the sRAGE concentrations (P < 0.001 and P = 0.001, respectively, ANOVA), whereas the rs17493811 SNP was not (Table 3). In a linear regression analysis with sRAGE concentration as the dependent variable, carrying the minor allele of rs2070600 or rs9469089 had an independent effect with sex, age, DR3 positivity, and DR4 positivity as other covariates (β −0.32, P < 0.001; β 0.09, P = 0.002, respectively). Subjects carrying the predisposing minor allele of rs2070600 were younger at the diagnosis of type 1 diabetes than the rest of the population (7.4 vs. 8.0 years of age, P = 0.03). Carrying the protective minor allele of rs9469089 was not associated with age at diagnosis (7.8 vs. 7.9 years of age, P = 0.61). In both rs9469089 and rs2070600, the major G alleles, which have opposite effects on sRAGE concentrations and type 1 diabetes risk, are present in most DR3 and DR4 haplotypes in the control population, while the minor C and A alleles are rare in both DR3 and DR4 (Supplementary Table 1). The frequency of IAAs was higher among carriers of the minor allele of the rs2070600 than among the remaining patients (55% vs. 43%, respectively, P < 0.001), whereas the frequency of ZnT8A was higher among carriers of the minor allele of the rs9469089 SNP than among the other patients (66% vs. 58%, P = 0.01), while the frequency of IAAs was lower when comparing the two groups (38% vs. 48%, respectively, P = 0.001). The associations between the SNPs and autoantibody frequencies remained significant after controlling for the effect of age, sex, and HLA genotype in logistic regression analyses (data not shown).

To our knowledge, this is the first study to report an association between circulating sRAGE concentrations and age at diagnosis of type 1 diabetes. This is particularly interesting, because in the control population there was no correlation between age and sRAGE concentration, and, as a matter of fact, the very young healthy children had higher concentrations than the rest of the control population. This observation may indicate that a low sRAGE concentration is associated with accelerated progression of the disease process leading to type 1 diabetes, or that higher concentrations are protective and associated with a milder and less rapidly progressing process. This notion is supported by our observation that DKA at diagnosis is associated with lower sRAGE concentrations as well. However, causality is not evident since the sRAGE was analyzed after the compensation of the DKA.

Decreased sRAGE concentrations have previously been described as being associated with the acute phase of various immune-mediated diseases, such as Kawasaki disease (26) and juvenile idiopathic arthritis (26,27). A relatively rapid increase in sRAGE concentration was reported in Kawasaki disease after successful treatment with intravenous Ig (26). However, reports on circulating sRAGE levels in individuals with chronic conditions with inflammation and autoimmunity are somewhat controversial. Stable patients with chronic juvenile idiopathic arthritis or rheumatoid arthritis have similar or even increased sRAGE concentrations when compared with healthy control subjects, although there was an inverse association between metabolic markers of inflammation and sRAGE concentrations (26–28). In previous studies on adult and pediatric patients with type 1 diabetes, the sRAGE concentrations have been increased when compared with those in healthy control subjects (29,30). Type 1 diabetes in the participants in those studies was, however, not newly diagnosed, but had afflicted those participants already for some years. One could speculate that in patient with chronic, stable conditions with autoimmunity and inflammation, compensatory mechanisms are activated to increase the circulating protective sRAGE concentrations. In contrast, some investigators have interpreted the higher sRAGE concentrations in patients with pathological conditions as reflecting a possible damaging mechanism (31).

Whether our results reflect an increased risk of rapid progression to clinical disease in subjects with initially low sRAGE concentrations or a secondary decrease in the sRAGE concentrations in patients with a more aggressive and rapidly developing disease, still remains to be determined in future studies. A recent study (32) on obese and normal weight children reported that being born small or large for gestational age correlated with decreased sRAGE and esRAGE concentrations. In the current study, the children in whom type 1 diabetes was diagnosed at a very young age (<2 years of age) had the lowest sRAGE concentrations, whereas in the control population the concentrations were the highest in this age group. This finding suggests that the reduced sRAGE concentrations are induced at a very early age, at least in some cases. It has been shown previously that children <2 years of age in whom type 1 diabetes had been diagnosed have a very aggressive disease process that is characterized by severe metabolic decompensation, poorly preserved residual β-cell function, and strong humoral autoimmunity against islet cells and insulin at diagnosis (33). Accordingly, the current observation of reduced circulating sRAGE concentrations in very young children may reflect a more aggressive disease process.

In contrast to our initial hypothesis, the circulating sRAGE concentrations did not correlate with the number of autoantibodies detectable at the diagnosis of type 1 diabetes. In fact, the only significant relationship between autoantibody specificities and sRAGE concentration was the positive association between sRAGE concentration and the frequency of ZnT8A. The strong inverse association between circulating sRAGE concentration and the predisposing minor allele of rs2070600 has been described previously (15,16). The rs2070600 is suggested to play a functional role in the RAGE/ligand interaction (34), and the minor allele was associated with increased risk of type 1 diabetes in a recent study (13). Interestingly, the minor allele of rs9469089, which was associated with decreased risk of type 1 diabetes (13), was associated with higher sRAGE concentration in the current study.

To conclude, circulating sRAGE concentrations are associated with age at diagnosis of type 1 diabetes, whereas there is no correlation between age and sRAGE concentration in healthy control subjects. Higher sRAGE concentrations seem to be related to a lower frequency of ketoacidosis at diagnosis. The AGER and/or HLA class II genotype regulate the circulating sRAGE concentrations. There is an association between the AGER genotype and the humoral autoimmunity against ZnT8 and insulin, suggesting that the RAGE genotype might influence the autoimmune process leading to type 1 diabetes. Together, the inverse association between circulating sRAGE concentrations, on one hand, and age at diagnosis, DKA at diagnosis, and HLA DR3/DR4 heterozygosity, on the other hand, indicate that reduced sRAGE concentrations at the time of the diagnosis of type 1 diabetes reflect a more aggressive disease process. Studies in prediabetic children will be needed to assess the role of sRAGEs in the development of type 1 diabetes.

The Finnish Pediatric Diabetes Register and Biobank comprise the following investigators: Principal Investigator: Mikael Knip (Children’s Hospital, Helsinki University Central Hospital). Steering Committee: Per-Henrik Groop (Folkhälsan Research Center), Jorma Ilonen (Immunogenetics Laboratory, University of Turku), Anneli Lappi (Children’s Hospital, Helsinki University Central Hospital), Timo Otonkoski (Children’s Hospital, Helsinki University Central Hospital), Marja-Terrtu Saha (Department of Pediatrics, Tampere University Hospital), Olli Simell (Chair, Department of Pediatrics, Turku University Central Hospital), Timo Talvitie (Department of Pediatrics, South Ostrobothnia Central Hospital), Outi Vaarala (Department of Vaccination and Immune Protection, National Institute for Health and Welfare), and Riitta Veijola (Department of Pediatrics, Oulu University Hospital). Locally responsible investigators: Henrikka Aito (Department of Pediatrics, Porvoo Hospital), Jonas Bondestam (Department of Pediatrics, Lohja Hospital), Thomas Dahllund (Department of Paediatrics, Västra Nyland Hospital), Johanna Granvik (Department of Pediatrics, Jakobstad Hospital), Maarit Haapalehto-Ikonen (Department of Pediatrics, Rauma Hospital), Anu-Maaria Hämäläinen (Department of Pediatrics, Jorvi Hospital), Hanna Huopio (Department of Pediatrics, Kuopio University Hospital), Christian Johansson (Department of Pediatrics, Åland Central Hospital), Anne Kinnala (Department of Pediatrics, Salo Hospital), Jussi Korhonen (Department of Pediatrics, North Karelia Central Hospital), Paavo Korpela (Department of Pediatrics, Kanta-Häme Central Hospital), Maarit Korteniemi (Department of Pediatrics, Central Hospital of Lapland), Pentti Lautala (Department of Pediatrics, Päijät-Häme Central Hospital), Kaija Lindström (Department of Pediatrics, Hyvinkää Hospital), Päivi Miettinen (Children’s Hospital, Helsinki University Central Hospital), Taina Mustila (Department of Pediatrics, Vaasa Central Hospital), Anja Nuuja (Department of Pediatrics, Central Hospital of Central Finland), Päivi Nykänen (Department of Pediatrics, Mikkeli Central Hospital), Jussi Ojanperä (Department of Pediatrics, Central Ostrobothnia Central Hospital), Anne- Putto-Laurila (Department of Pediatrics, Turku University Central Hospital), Marja-Terttu Saha (Department of Pediatrics, Tampere University Hospital), Juhani Sankila (Department of Pediatrics, Savonlinna Central Hospital), Anne-Maarit Suomi (Department of Pediatrics, South Ostrobothnia Central Hospital), Sirpa Tenhola (Department of Pediatrics, Kymenlaakso Central Hospital), Pentti Varimo (Department of Pediatrics, Kainuu Central Hospital), Riitta Veijola (Department of Pediatrics, Oulu University Hospital), Ritva Virransalo (Department of Pediatrics, South Karelia Central Hospital), Pentti Vuolukka (Department of Pediatrics, Länsi-Pohja Central Hospital), and Samuli Ylitalo (Department of Pediatrics, Satakunta Central Hospital).

Acknowledgments. The authors thank Markku Lehto, Maikki Parkkonen, Anna-Reetta Salonen, Tuula Soppela, Matti Koski, and Sirpa Nolvi for their assistance.

Funding. This study was supported by the Academy of Finland (Centre of Excellence in Molecular Systems Immunology and Physiology Research 2012–2017, Decision No. 250114), the Sigrid Jusélius Foundation, the Novo Nordisk Foundation, the Liv and Hälsa Fund, and the National Graduate School of Clinical Investigation.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. K.M.S. analyzed the data, wrote the first version of the manuscript, and edited the manuscript. S.J.R. and J.M.F. reviewed the manuscript and contributed to the discussion. T.H. was in charge of the autoantibody laboratory, reviewed the manuscript, and contributed to the discussion. J.I. was responsible for the HLA and AGER genotyping, reviewed the manuscript, and contributed to the discussion. A.-P.L. was responsible for the AGER genotyping, reviewed the manuscript, and contributed to the discussion. P.-H.G. and M.K. designed the study, contributed to the discussion, and reviewed the manuscript. The participants in the Finnish Pediatric Diabetes Register were involved in the planning of the study design and the collection of data. M.K. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.