Strong Association of Autoantibodies Targeting Deamidated Extracellular Epitopes of Insulinoma Antigen-2 With Clinical Onset of Type 1 Diabetes

Type 1 diabetes (T1D) is an autoimmune disease characterized by the presence of circulating islet autoantibodies (IAbs) in peripheral blood. The established biochemical IAbs targeting glutamate decarboxylase (GADA), insulinoma antigen-2 (IA-2A), zinc transporter 8 (ZnT8A), and insulin (IAA) typically appear years before overt clinical disease. IAbs are the most reliable biomarkers for predicting disease progression in T1D (1): two or more IAbs confer significantly higher risk of progression to clinical diabetes (2).

A major islet autoantigen, IA-2, is a transmembrane glycoprotein that belongs to the tyrosine phosphatase-like protein family and is localized to the insulin-secretory granules of the pancreatic β-cells. The IA-2 molecule contains three domains: the N-terminal extracellular domain (IA-2ec; amino acids [aa] 1–576), the transmembrane domain (aa 577–600), and the C-terminal intracellular domain (IA-2ic; aa 601–979). The major humoral immunoreactivities targeting the IA-2 protein have been mapped to the C-terminal intracellular domain, IA-2ic. Although extensive research has focused on studying autoantibodies to the intracellular domain of IA-2 (IA-2icA), there has been limited investigation into autoantibodies targeting the IA-2 extracellular domain (IA-2ecA).

Over the past decade, accumulating evidence has indicated that some pathogenic T cells in T1D recognize endoplasmic reticulum stress-induced epitopes formed by post-translational modification of self-antigens that could increase antigenic diversity. T cells recognizing post-translationally modified (PTM) epitopes may avert thymic selection, which could play an important role in the initiation and amplification of autoimmunity (3,4). Post-translational modification of existing proteins, such as oxidation, deamidation, citrullination, and phosphorylation, could also give rise to modified epitopes with altered immunogenicity (5). In particular, deamidation generates negatively charged aa preferred by T1D high-risk HLA-DQ molecules in several known islet antigens, including preproinsulin, ZnT8, IA-2, phogrin, GAD65, IGRP, and ICA69 (6). Recent studies found that CD4⁺ T cells that recognized deamidated peptides in the IA-2ec were present in peripheral blood and in the pancreatic draining lymph nodes from individuals with T1D (7). Although the analysis of immune responses to PTM neoepitopes has focused predominantly on T cells, evidence regarding autoantibodies specific for neoepitopes in T1D is limited (8,9). Because CD4⁺ T cells undergo interactions with B cells that can lead to the production of antibodies, we hypothesized that deamidated IA-2ec may serve as an antigen to promote humoral autoimmunity in T1D.

In this study, we compared the profiles of autoantibodies targeting the IA-2ec Q > E epitopes (PTM IA-2ecA) with autoantibodies targeting the IA-2ec unmodified epitopes (IA-2ecA), assessing their relationship to T1D across different stages of disease development, as well as in other types of diabetes and other autoimmune diseases.

Archived serum samples at the Barbara Davis Center for Diabetes clinics from patients diagnosed with diabetes within 26 weeks of onset were randomly selected for the study. This included 1,049 patients diagnosed with T1D who were positive for at least one IAb (GADA, IAA, IA-2A, or ZnT8A), and 115 patients diagnosed with type 2 diabetes (T2D) who were negative for all IAbs. Of the 1,049 patients with T1D, 790 were children (aged <18 years) and 259 were adults. Control samples were from 509 children (aged 1–17 years) who participated in the Autoimmunity Screen for Kids (ASK) study (10) and who tested negative for all IAbs (GADA, IAA, IA-2A, and ZnT8A) and antitissue transglutaminase antibodies (tTGAs). Serum samples from 271 patients with latent autoimmune diabetes in adults (LADA) who were all positive for GADA, had duration of diabetes < 5 years, and were aged 30–70 years, and participated in the Action LADA study (11) were included in the present study. In addition, 130 participants from the ASK study who were tTGA positive and IAb negative, representing a distinct autoimmune cohort, were analyzed. Furthermore, a total of 360 serum samples from a birth cohort of 222 participants in the Diabetes Autoimmunity Study in the Young (DAISY) (12) were analyzed. These participants were followed longitudinally from birth to stage 1 or stage 3 of T1D. The selected participants were tested for both IA-2ecAs at different time points, including prior to IAb seroconversion (n = 33), at single-IAb seroconversion (n = 161), at multiple-IAb seroconversion (n = 106), and at diagnosis of stage 3 T1D (n = 60).

All participants (or the guardian, when the participant was younger than 18 years) provided written or electronic informed consent. The studies were approved by the University of Colorado Multiple Institutional Review Board.

The T-cell epitopes in IA-2ec containing deamidated Q > E residues lie in four short segments: aa 198–216, aa 467–482, aa 523–536 and aa 545–562 (7) (Fig. 1A). Initially, a cDNA encoding IA-2ec (aa 26–576, devoid of the signal sequence) was cloned with the following PCR primers: forward: GTAGGTACCACCATGAGCAGCCGCCCTGGAGGCTGCAGCGCCGTTAGTGCC; reverse: CATGCGGCCGCCTATCACACTGAGCGCATGGGTGAGGTGCTGTG, using a full-length IA-2 cDNA template. The amplification product was cut with restriction enzymes KpnI and NotI and cloned into a vector equipped for coupled in vitro transcription (SP6)/translation (pCMVTnT; Promega). Site-specific mutations were sequentially substituted in the IA-2ec construct to generate the PTM IA-2ec construct that encodes E residues for the Q residues at the positions shown in Fig. 1B. Forward primers used for mutagenesis were as follows:

Epitope 1: AAGTTATGAACCTGCTTTACTAGAGCCCTATCTGTTCCAT, GAGCCCTATCTGTTCCATGAGTTTGGTTCACGTGATG;

Epitope 2: CTACATCGTCACTGATGAGAAGCCCCTGAGCCTGGCT;

Epitope 3: CTGTCTTTGGCTGATGTGACCGAACAAGCAGGGCTGGTGAAGTC, GTCTTTGGCTGATGTGACCGAAGAAGCAGGGCTGGTGAAGTCTG;

Epitope 4: CTGGAAGCACAGACAGGGCTCGAAATCTTGCAGACAGGAGTG, CAGACAGGGCTCGAAATCTTGGAGACAGGAGTGGGACAGAGGG, GAAATCTTGGAGACAGGAGTGGGAGAGAGGGAGGAGGCAGCTGCAG.

Reverse complement primers were used for each corresponding forward primer. Mutagenesis was conducted with in-house reagents similar to the QuikChange Kit (Agilent Technologies) using 18 PCR cycles with PfuTurbo (Agilent Technologies). Amplification products were digested with DpnI (New England Biolabs) to remove methylated template DNA and transformed into competent bacteria. Several clones from each mutagenesis were analyzed by automated sequencing of plasmid DNA to verify the mutation(s). These constructs were used to generate ³⁵S-labeled antigen probes for radio-binding assays (RBAs) to establish the presence of antibodies targeting the unmodified and PTM IA-2ec antigens.

The standard RBA methods for GADA, IAA, IA-2A, and ZnT8A have been published (13). In the 2023 Islet Autoantibody Standardization Program workshop, the sensitivities and specificities of RBA IAb assays in our laboratory were: 62% and 99% for IAA, 78% and 99% for GADA, 72% and 100% for IA-2A, and 74% and 100% for ZnT8A.

The RBA method for IA-2ecA and PTM IA-2ecA is similar to the standard RBA for other IAbs. Purified plasmid DNA of IA-2ec and PTM IA-2ec constructs were transcribed and translated in vitro using the TnT SP6-coupled rabbit reticulocyte lysate system (Promega, Madison, WI) in the presence of ³⁵S-methionine (Revvity, Hopkinton, MA), according to the manufacturer's instructions. Unincorporated ³⁵S-methionine was removed by size exclusion chromatography on a NAP5 column (Cytiva, Marlborough, MA). SDS-PAGE and autoradiography were used to verify that the molecular weight of the translated IA-2ec protein matched that predicted from the aa sequence. To verify the antigen quality, ³⁵S-labeled IA-2ic protein was produced in parallel as a control. The ³⁵S-labeled IA-2ec antigens were tested using positive and negative controls before testing serum samples. For the RBA method, ³⁵S-labeled IA-2ec (20,000 counts per minute [cpm]) or PTM IA-2ec (20,000 cpm) was incubated with 2.5 μL of patient serum at a final 1:25 dilution overnight at 4°C. Autoantibody-bound antigen was precipitated with protein A Sepharose (Cytiva, Marlborough, MA). After washing eight times with PBST, scintillation fluid was added, and radioactivity was counted on a TopCount β-counter (Revvity, Hopkinton, MA). Positive and negative control sera for IA-2ecA and PTM IA-2ecA were included in every assay plate. The levels of both antibodies were expressed as an index (index = [(cpm_sample – cpm_{negative control})/(cpm_{positive control} – cpm_{negative control})]. The assay cutoff was set at the 99th percentile of 509 negative control samples, with the index of 0.226 for IA-2ecA and 0.180 for PTM IA-2ecA. Each serum sample was tested in triplicate. The intra-assay coefficient of variation (CV) was 7.1% (n = 10) for IA-2ecA and 6.5% (n = 10) for PTM IA-2ecA. The interassay CV was 9.2% (n = 40) for IA-2ecA and 8.1% (n = 40) for PTM IA-2ecA.

Antibody absorption assays were conducted to assess the specificities of IA-2ecAs, using excess unlabeled antigen proteins to compete with labeled antigens. The assay included two serum samples positive for IA-2ecA but negative for PTM IA-2ecA, and two samples positive for PTM IA-2ecA but negative for IA-2ecA. Unlabeled IA-2ec and PTM IA-2ec proteins were incubated separately with each sample to evaluate potential cross-binding activity between these two IA-2ecAs. The assay was performed using the RBA method. Unlabeled IA-2ec and PTM IA-2ec were expressed in the reticulocyte lysate system with an aa mixture containing unlabeled methionine to serve as a competitive inhibitor. Serum samples were incubated overnight at 4°C with either ³⁵S-labeled or excessive unlabeled IA-2ec proteins. The antibody–antigen complexes were then precipitated with protein A Sepharose, and radioactive signals (cpm) were counted. Signals of unabsorbed serum (without incubation with unlabeled antigen) were used as the positive control for the assay. The unabsorbed rate (%) was calculated using the formula: [(cpm_{absorbed sample} – cpm_{negative control})/(cpm_{unabsorbed sample} – cpm_{negative control})] ×100, to assess antibody specificity. A lower unabsorbed rate indicates higher antibody specificity for the unlabeled antigen.

The data that support the findings of this study are available from the corresponding author upon reasonable request.

All statistical analyses were performed using GraphPad Prism software (version 8.0; GraphPad Software Institute) and SAS, version 9.4 (SAS Institute Inc). Continuous variables are presented as the mean ± SD. Frequency and its percentage were generated for categorical variables. Between-group comparisons were performed with the t test for continuous variables and with the χ² test or Fisher exact test for categorical variables, as appropriate. The McNemar test assessed the proportions of IA-2ecA positivity and PTM IA-2ecA positivity in patients with new-onset T1D. Multivariable logistic regression was used to identify characteristics associated with IA-2ecA positivity or PTM IA-2ecA positivity, with predictor variables including age, sex, diabetes duration, GADA positivity (yes or no [Y/N]), IA-2A positivity (Y/N), IAA positivity (Y/N), and ZnT8A positivity (Y/N). Pearson correlation analysis was performed to assess the strength of correlation between levels of IA-2ecA and PTM IA-2ecA. In figures, the indexes of IA-2ecA and PTM IA-2ecA were presented as an index score (index score = sample index divided by corresponding cutoff index) for easy comparison. Two-tailed P < 0.05 was considered significant.

The specificities of IA-2ecA and PTM IA-2ecA were set at 99% of 509 negative controls, and the cutoff indexes were 0.226 and 0.180, respectively (Fig. 2). The positivity of PTM IA-2ecA in patients with new-onset T1D (n = 274; 26.1%) was significantly higher than the positivity of IA-2ecA (n = 205 [19.5%]; P < 0.0001). Comparing patients with new-onset T1D with and without IA-2ecA, with and without PTM IA-2ecA, with either IA-2ecA, and without both IA-2ecA types revealed several trends: patients with IA-2ecAs, compared with those without, were generally younger, had a longer T1D duration, and were more likely to be negative for IA-2icA (Table 1). In multivariable logistic regression analysis, the prevalence of PTM IA-2ecA was associated with younger age and longer T1D duration, whereas no associations were found for IA-2ecA (Supplementary Table 1). The positivity of both IA-2ecA and PTM IA-2ecA were significantly higher in children with new-onset T1D (22.2% and 29.0%, respectively) than in adults (11.6%, P = 0.0001; and 17.4%, P = 0.0002, respectively).

The prevalences of both IA-2ecA and PTM IA-2ecA were comparable between patients with single or multiple IAb positivity; 23.6% (n = 17 of 72) and 34.7% (n = 25 of 72), respectively, in participants with single IAb positivity; 19.2% (n = 188 of 977) and 25.5% (n = 249 of 977), respectively, in participants with multiple IAbs. Notably, among 72 patients with T1D with a single IAb, 24–35% were reclassified as having multiple IAbs with the inclusion of IA-2ecA or PTM IA-2ecA measurements.

The positivity of both IA-2ecA and PTM IA-2ecA was independent of IA-2icA, and both IA-2ecAs were detected more frequently in patients negative for IA-2icA. In 171 patients with T1D who were negative for IA-2icA, 42 patients (24.6%) tested positive for IA-2ecA, and 64 patients (37.4%) tested positive for PTM IA-2ecA (Fig. 3A). Among 876 patients with T1D who were positive for IA-2icA, 163 (18.6%) were positive for IA-2ecA and 209 (23.9%) were positive for PTM IA-2ecA (Fig. 3B). By incorporating IA-2ecA and PTM IA-2ecA, the overall detection of autoantibodies to IA-2 in these patients increased from 83.5 to 89.8%.

In addition, the positivity of IA-2ecA and PTM IA-2ecA was mostly superimposable. Among 310 patients with T1D who were positive for IA-2ecA or PTM IA-2ecA, 168 were positive for both and 142 were positive only for one, including 37 patients positive only for IA-2ecA and 105 positive only for PTM IA-2ecA (Fig. 3C). In the participants who were positive for both IA-2ecA and PTM IA-2ecA, the levels of IA-2ecA and PTM IA-2ecA were well correlated (r = 0.410; P < 0.0001). The relationships of IA-2ecA and PTM IA-2ecA with the four established IAbs are shown in Fig. 3C.

The antibody absorption assays revealed that the specificity of IA-2ecA and PTM IA-2ecA partially overlapped. In PTM IA-2ecA–positive and IA-2ecA–negative serum samples, unlabeled PTM IA-2ec antigen protein was able to completely absorb the signals of PTM IA-2ecA, whereas the same amount of unlabeled IA-2ec antigen exhibited only partial absorption (74–83%), with unabsorbed rates ranging from 17% to 26% (Fig. 4A). A similar pattern was observed with IA-2ecA–positive and PTM IA-2ecA–negative serum samples, in which unlabeled IA-2ec protein was able to completely absorb the signals of IA-2ecA, whereas unlabeled PTM IA-2ec showed partial absorption (61–81%), with unabsorbed rates ranging from 19% to 39% (Fig. 4B).

To investigate the timing of IA-2ecA and PTM IA-2ecA appearance during T1D development, 222 participants from DAISY who were prospectively followed from birth to stages 1 to 3 of T1D were studied. Neither IA-2ecA nor PTM IA-2ecA was detected in serum samples prior to IAb seroconversion or at single IAb seroconversion. At stage 1 T1D (multiple IAb positive), IA-2ecA or PTM IA-2ecA was detected only in three participants at low levels (2.8%) with overlap in one participant. At the diagnosis of stage 3 T1D, five of 60 individuals were positive for IA-2ecA (8.3%) and eight were positive for PTM IA-2ecA (13.3%), with overlap in two participants (Fig. 5). Compared with the prevalence in patients with overt clinical T1D, both IA-2ecA and PTM IA-2ecA were much less frequent in preclinical stages of T1D and stage 3 T1D in closely monitored children from the DAISY study.

Further analysis of the demographic information of DAISY participants who progressed to stage 3 T1D revealed that they had notably lower HbA_1c levels, fewer instances of diabetic ketoacidosis (DKA), and shorter durations of diabetes than the new-onset T1D patients diagnosed in routine clinical settings with overt symptoms (Table 2). Taken together, these results indicate that IA-2ecA and PTM IA-2ecA appear at later stages of T1D development and are more prevalent in those individuals presenting with marked metabolic abnormalities such as elevated HbA_1c levels and DKA.

Next, we investigated IA-2ecA and PTM IA-2ecA in patients with other types of diabetes and autoimmunity, including those with LADA, T2D, and celiac disease autoimmunity (positive for tTGA). In 271 patients with LADA, only four (1.5%) and five (1.8%) were positive for IA-2ecA and PTM IA-2ecA, respectively: much less than in the cohort with new-onset T1D (P < 0.0001 for both). Of 115 patients with T2D, none had either IA-2ecA or PTM IA-2ecA. Among 130 individuals with celiac disease autoimmunity (tTGA positive, negative for all IAbs), only two (1.5%) and one (0.8%) were positive for IA-2ecA and PTM IA-2ecA, respectively (Fig. 6). These results indicate that IA-2ecA and PTM IA-2ecA are predominantly present in patients with new-onset T1D, especially those with overt clinical T1D at a younger age.

Post-translational modification of self-antigens is one explanation of how autoreactive T cells may escape thymic selection. In T1D, pathogenic T cells recognize epitopes formed by post-translational modification, including deamidation, citrullination, oxidization, peptide fusion, and use of alternate reading frames (4,7,14–18). Recent observations demonstrate that PTM IA-2 peptides within the extracellular domain are the legitimate targets of autoreactive CD4⁺ T cells in patients with T1D (19–21). It is well established that T and B cells interact, and the interaction of antigen-specific CD4⁺ T cells with B cells may aid in the production of antibodies. Autoantibodies to oxidized PTM insulin (oxPTM-INS) have been found in the sera of patients with T1D (8). Concordant autoimmune responses to oxidized PTM insulin peptides by CD4⁺ and CD8⁺ T cells (simultaneously), along with autoantibodies to oxPTM-INS have been observed in the same patients with T1D. This suggests that the CD4⁺ T-cell response to these peptides is required to generate CD8⁺ and/or antibody responses to oxPTM-INS (22). Similarly, elevated T-cell reactivity to hybrid insulin peptides (HIPs) (23) and autoantibodies to HIPs (9) have been found in T1D patients. Here, we report that antibodies targeting deamidated epitopes within the extracellular domain of IA-2, based on antigen-specific T-cell responses in patients with new-onset T1D, are detectable in the circulation.

Antibodies to both IA-2ec and PTM IA-2ec were identified in patients with new-onset T1D, especially in those with younger-onset cases, but rarely in LADA, T2D, or in individuals with only celiac autoimmunity. Both IA-2ecA and PTM IA-2ecA were independent autoantibodies directed against the intracellular domain of IA-2 protein. Our results for IA-2ecA are in line with previously published studies indicating 8–27.8% prevalence in patients with T1D (21, 24) but rarely found in patients with T2D (24). The prevalence of PTM IA-2ecA compared with IA-2ecA in patients with newly diagnosed T1D was significantly higher, suggesting that the extracellular domain of IA-2 undergoes post-translational protein modifications during the development of T1D.

We used samples from the prospective birth cohort DAISY study to determine the timing of IA-2ecA and PTM IA-2ecA appearance in the natural history of T1D development. Antibodies to the IA-2ec, whether unmodified or PTM, were not present before, at, or after seroconversion to any of the four major biochemical IAbs during the very early stage of T1D development. PTM IA-2ecA was predominantly detected at the clinical onset of T1D. Unexpectedly, we noted a lower prevalence of both IA-2ecA and PTM IA-2ecA in DAISY participants with stage 3 T1D, who were closely monitored for metabolic decompensation, compared with patients with new-onset T1D diagnosed in diabetes clinics, where these patients presented with higher HbA_1c levels and more frequent DKA (25). Autoantibodies to IA-2ec or PTM IA-2ec tend to develop at later stages of T1D, potentially explaining this differences between the cohort with new-onset T1D diagnosed in the clinic and those patients diagnosed through the closely monitored DAISY study. Taken together, these data indicate that deamidation of IA-2ec is unlikely to reflect the early change contributing to the loss of immune tolerance to islet antigens. Instead, this late PTM process on islet antigen proteins might reflect the metabolic chaos in the β-cells at the late disease stage resulting from autoimmunity and endoplasmic reticulum stress.

All the established biochemical IAbs target either secreted (e.g., insulin) or intracellular domains of islet proteins (e.g., GAD65, IA-2, ZnT8). There are few studies of autoantibodies targeting extracellular domains or epitopes within the transmembrane domains of proteins in autoimmune diseases (26). We recently identified autoantibodies against the extracellular epitopes of the ZnT8 protein, which were positive in 23.6% of patients with new-onset T1D and were present earlier than any other IAbs in a subset of DAISY participants (27). In autoimmune hyperthyroidism (Graves’ disease), autoantibodies to the extracellular thyroid-stimulating hormone receptor were infrequent (2%) 5–7 years before diagnosis, but far more frequent (55%) at diagnosis (28). In contrast, autoantibodies to intracellular thyroid peroxidase gradually increased from 31% (5–7 years prior to diagnosis) to 57% at diagnosis (28), similar to the pattern we observed in autoantibodies against the IA-2ec in T1D.

Our work is not without limitations. First, the post-translational modification for IA-2ec used in the antibody assay was designed to reflect CD4⁺ T-cell epitopes, and there may be other PTM epitopes for antibodies independent of CD4⁺ T cell epitopes. Second, T and B cells can interact even if they recognize distinct specific epitopes. Although the deamidated epitopes can simulate T-cell reactivity, B cells may or may not recognize the same epitopes. More studies are needed to verify whether the other deamidated domains of IA-2ec serve as autoantibody epitopes and whether these same deamidated epitopes can also simulate T-cell reactivity in the same individuals. Third, our PTM IA-2ec included four T-cell epitopes with eight separate deamidated aa residues, and more detailed studies are warranted to investigate which individual deamidated site(s) are necessary for antibody binding. Similarly, it is not clear whether post-translational modification affects other known islet antigens in T1D.

In conclusion, we identified autoantibodies that bind to the extracellular domain of IA-2, a known self-antigen in T1D, as well as distinct autoantibodies that recognize deamidated IA-2ec. These deamidated IA-2ec autoantibodies are present late in the course of the disease, which does not support the hypothesis of loss of immune tolerance via post-translational modification of islet antigens. However, these new autoantibodies are, indeed, biomarkers of islet autoimmunity, potentially assisting in the differential diagnosis of diabetes, particularly in individuals positive for a single islet autoantibody.

Acknowledgments. The authors thank Bing Wang and Guy Alonso at the DRC Clinical Core for collection of the clinical demographic data on the patients with new-onset T1D. The authors thank the patients and their families as well as the investigators and staff involved in the DAISY, ASK, and Action LADA studies.

Funding. This study was supported by the National Institutes of Health (NIH) grants AI153665, DK032083, DK108868, and DK099317; and Diabetes Research Center grant P30 DK116073. The DAISY study was supported by NIH grant DK 032493. The ASK study was supported by Juvenile Diabetes Research Foundation grant 1-SRA-2016-208-S-B. The Action LADA study was supported by the 5th Framework Programme of the European Union.

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

Author Contributions. X.J., J.M.W., and L.Y. designed the study, researched data, and wrote and edited the manuscript. C.Z., F.D., K.W., R.D.L., M.J.R., and A.W.M. researched data and reviewed and edited the manuscript. L.Y. is the guarantor of this work and, as such, has 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.

Prior Presentation. Parts of this work were presented in abstract form at the 83rd Scientific Sessions of the American Diabetes Association, 23–26 June 2023, San Diego, CA (https://doi.org/10.2337/db23-174-LB), and the 82nd Scientific Sessions of the American Diabetes Association, 3–7 June 2022, New Orleans, LA (https://doi.org/10.2337/db22-186-OR).

Diabetes Autoimmunity Study in the Young (DAISY) and Autoimmunity Screening for Kids (ASK) Study Group. Members include, from the Barbara Davis Center for Diabetes University of Colorado Denver, Aurora, CO, Marian Rewers, Jill Norris, Andrea Steck, Brigitte Frohnert, Cristy Geno, Liping Yu, Kimber Simmons, and Holly O’Donnell and, from the Children’s Hospital University of Colorado, Aurora, CO, Ed Liu.

Action LADA Consortium. Members include Richard David Leslie and Mohammed I. Hawa (deceased), Blizard Institute, Queen Mary University of London, London, U.K.; Paolo Pozzilli, University Campus Bio-Medico, Rome, Italy; Henning Beck-Nielsen and Knud Yderstraede, University Hospital of Odense, Odense, Denmark; Steven Hunter and David Hadden (deceased), Royal Victoria Hospital, Belfast, U.K.; Raffaella Buzzetti, University La Sapienza, University of Rome, Rome, Italy; Werner Scherbaum and Hubert Kolb, University of Düsseldorf, Düsseldorf, Germany; Nanette C. Schloot, German Diabetes Centre and Clinic for Metabolic Diseases at University Hospital Düsseldorf, Düsseldorf, Germany (currently Lilly Deutschland, Bad Homburg, Germany); Jochen Seissler, Ludwig-Maximilians-University, Munich, Germany; Guntram Schernthaner, Rudolfstiftung Hospital, Vienna, Austria; Jaakko Tuomilehto and Cinzia Sarti, Finnish Institute for Health and Welfare, Helsinki, Finland; Alberto De Leiva and Eulalia Brugues, Universitat Autonoma de Barcelona, Barcelona, Spain; Didac Mauricio, Hospital de la Santa Creu i Sant Pau, Barcelona, Spain; and Charles Thivolet, Hopital Edouard Herriot, Lyon, France.