Pathophysiology/Complications ORIGINAL ARTICLE

The Tyrosine Phosphatase Nonreceptor 22 (PTPN22) Is Associated With High GAD Antibody Titer in Latent Autoimmune Diabetes in Adults Non Insulin Requiring Autoimmune Diabetes (NIRAD) Study 3

1 5 ANTONIO PETRONE, PHD EFISIO COSSU, MD CONCLUSIONS — In adult-onset auto- 2 6 CONCETTA SURACI, MD PAOLO POZZILLI, MD immune diabetes, the PTPN22 1858T variant 1 7 MARCO CAPIZZI, MD ALBERTO FALORNI, MD is associated only with a high GADA titer, pro- 3 1 ANDREA GIACCARI, MD, PHD RAFFAELLA BUZZETTI, MD viding evidence of a genetic background to 4 EMANUELE BOSI, MD FOR THE TUDY ROUP clinical heterogeneity identified by GADA 1 NIRAD S G * CLAUDIO TIBERTI, ATA titer.

Diabetes Care 31:534–538, 2008 OBJECTIVE — We previously demonstrated the presence of two different populations among individuals with adult-onset autoimmune diabetes: those having either a high titer or a consistent fraction of subjects (4– low titer of antibodies to GAD (GADAs). Protein tyrosine phosphatase nonreceptor type 22 10%) with adult-onset non–insulin- (PTPN22) has been identified as a new susceptibility for type 1 diabetes and other autoim- requiring diabetes at diagnosis, also mune diseases. The aim of the present study was to evaluate whether the phenotypic heteroge- A referred to as latent autoimmune diabetes neity of adult-onset autoimmune diabetes based on the GADA titer is associated with the PTPN22 in adults or non–insulin-requiring auto- C1858T polymorphism. immune diabetes, has autoimmune fea- RESEARCH DESIGN AND METHODS — Analysis for the C1858T polymorphism tures, specifically the presence of GAD using the TaqMan assay was performed in 250 subjects with adult-onset autoimmune diabetes, autoantibodies (GADAs). These patients divided into two subgroups with low (Յ32 arbitrary units) or high (Ͼ32 arbitrary units) GADA do not initially require insulin treatment titers and 450 subjects with classic type 2 diabetes (from the Non Insulin Requiring Autoimmune and are extremely heterogeneous in terms Diabetes [NIRAD] Study cohort of 5,330 subjects with adult-onset diabetes) and in 558 subjects of clinical presentation, ranging across the with juvenile-onset type 1 diabetes and 545 normoglycemic subjects. whole spectrum between classic pheno- types of type 1 and type 2 diabetes (1–4). RESULTS — Genotype, allele, and phenotype distributions of the PTPN22 C1858T variant We have recently demonstrated the revealed similar frequencies in autoimmune diabetes with high GADA titer and juvenile-onset presence of two different populations type 1 diabetes. An increase in TT and CT genotypes was observed in individuals with a high among individuals with adult-onset auto- GADA titer compared with a low GADA titer, those with type 2 diabetes, and control subjects Ͻ immune diabetes (5); analysis of GADA (P 0.002 for all comparisons). The PTPN22 1858T allele and phenotype frequencies were titers showed a bimodal distribution that increased in high GADA titer compared with a low GADA titer, type 2 diabetic, and control identified two subgroups of patients with subjects (P Ͻ 0.001 for all comparisons, odds ratio 2.6). either a low or a high GADA titer. Com- pared with patients with a low GADA titer, patients with a high GADA titer had more prominent traits of insulin defi- ●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●● ciency and a profile of more severe , resulting in a higher From the 1Department of Clinical Sciences, Sapienza University, Rome, Italy; the 2Sandro Pertini Hospital, 3 4 prevalence of protein tyrosine phospha- Rome, Italy; Endocrinology, Catholic University, Rome, Italy; General Medicine, Diabetes and Endocri- nology, San Raffaele Vita-Salute University, Milan, Italy; the 5Department of Endocrinology and Metabolism, tase IA-2, thyroid peroxidase antibodies, University of Cagliari, Cagliari, Italy; 6Endocrinology, University Campus Bio-Medico, Rome, Italy; and the and DRB1*03-DQB1*0201 and a de- 7Department of Internal Medicine, University of Perugia, Perugia, Italy. creasing frequency of DQB1*0602 and Address correspondence and reprint requests to Professor Raffaella Buzzetti, Department of Clinical DRB1*0403. Science, Sapienza University of Rome, Viale del Policlinico 155, 00161 Rome, Rome, Italy. E-mail: raffaella. [email protected]. A new susceptibility gene to type 1 Received for publication 31 July 2007 and accepted in revised form 17 November 2007. diabetes has been recently identified out- Published ahead of print at http://care.diabetesjournals.org on 4 December 2007. DOI: 10.2337/dc07- side the HLA region, protein tyrosine 1457. phosphatase nonreceptor type 22 *The list of centers and physicians participating in the NIRAD Study has been published in ref. 5. This article is dedicated to the memory of Professor Umberto Di Mario who greatly contributed to the (PTPN22) (6), which encodes a lym- design and implementation of the study. phoid-specific phosphatase known as Abbreviations: GADA, GAD autoantibody; NIRAD, Non Insulin Requiring Autoimmune Diabetes; LYP, a powerful inhibitor of T-cell activa- PTPN22, protein tyrosine phosphatase nonreceptor type 22. tion (7). Several studies showed that a © 2008 by the American Diabetes Association. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby missense single nucleotide polymor- marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. phism, C1858T, in the PTPN22 gene is

534 DIABETES CARE, VOLUME 31, NUMBER 3, MARCH 2008 Petrone and Associates

Table 1—Clinical characteristics of the groups of subjects investigated

High GADA titer Low GADA titer autoimmune Type 1 diabetes autoimmune diabetes diabetes Type 2 diabetes Control n (male/female) 296/262 65/58 66/61 234/216 278/267 Age at diagnosis (years) 14.9 Ϯ 7.8 49.3 Ϯ 12.6 51.8 Ϯ 13.3 51.6 Ϯ 10.8 30 Ϯ 5 BMI (kg/m2) 18.33 Ϯ 3.6 26.1 Ϯ 5.03 28.2 Ϯ 5.12 29.4 Ϯ 5.01 21.8 Ϯ 2.2 Fasting glucose (mg/dl) 256 Ϯ 137 170.6 Ϯ 62.8 165 Ϯ 55 144 Ϯ 48.9 78.15 Ϯ 10 A1C (%) 10.62 Ϯ 2.49 7.6 Ϯ 1.7 7.1 Ϯ 1.6 6.5 Ϯ 1.4 — Data are means Ϯ SD. associated with type 1 diabetes (6) and going project aimed to identify autoim- control samples designated at 10 GADA other autoimmune diseases (8–10); so far mune diabetic subjects for a series of arbitrary units were 9 and 17%, respec- it is unclear how the 1858T allele can in- clinical studies). Adult-onset autoim- tively. The distribution of GADA titer in fluence the activity of the LYP phospha- mune diabetic subjects were selected us- patients with autoimmune diabetes was tase. In a recent study Vang et al. (11) ing the following inclusion criteria: 1)an independent of diabetes duration and demonstrated that the Arg620Trp variant initial diagnosis of type 2 diabetes accord- showed a bimodal distribution. Consis- (which corresponds to the C1858T poly- ing to the American Diabetes Association tent with this observation, patients with morphism: the 1858T variant changes (14), 2) documented antibody positivity autoimmune diabetes (GADA titer Ͼ3 codon 620 from arginine [Arg] to trypto- for GADAs (15), 3) no insulin require- arbitrary units) were divided into sub- phan [Trp]) is a gain-of-function form of ment and no evidence of ketosis from di- groups representing the two distribu- the protein, but the mechanism by which agnosis to screening time, and 4) disease tions, namely low (taken to be Յ32 the PTPN22 Trp20 variant exerts the dis- duration between 6 months and 5 years. arbitrary units) and high (Ͼ32 arbitrary ease-promoting effect has yet to be estab- Type 1 diabetic subjects (n ϭ 558) (age at units) GADA titer. Samples with low lished. Evidence has also been provided onset 14.9 Ϯ 7.8 years) were recruited by GADA titer were validated for GAD- regarding a permissive role played by the participating centers of the Immunother- specific binding by a competition assay PTPN22 C1858T variant on disease pro- apy Diabetes (IMDIAB) group in the Lazio with an excess of cold insulin (5). Based gression from pre-diabetes to clinical region of central Italy (16). The control on the Diabetes Antibody Standardization disease (12). Finally, it has been hypo- group comprised normoglycemic sub- Program as a reference (17), the threshold thesized that the PTPN22 polymorphism jects (n ϭ 545) with no family history of of 32 arbitrary units was equivalent to 300 is associated primarily with autoanti- autoimmune disorders (aged 30 Ϯ 5 World Health Organization units (5). body-positive autoimmune diseases (13). years), collected from the Blood Transfu- The aim of the present study was to study sion Service of Sapienza University of HLA and PTPN22 genotyping whether the phenotypic heterogeneity of Rome. The clinical characteristics of the Genomic DNA was extracted using the adult-onset autoimmune diabetes based five groups investigated are reported in salting out method (18). HLA-DRB1 and on the GADA titer is supported by the Table 1. The study was approved by all DQB1 typing was performed using allele genetic analysis, evaluating whether the local ethics committees of participating group–specific amplifications. A reverse PTPN22 C1858T polymorphism is asso- centers, and written informed consent line blot method, kindly provided by H.A. ciated with a high GADA titer instead of was obtained from all patients. Erlich and T. Bugawan (Roche Molecular antibody positivity per se. Systems, Alameda, CA), was used as the Autoantibody measurement detection system (19). The C1858T was RESEARCH DESIGN AND GADAs were measured by a radiobinding genotyped using the fluorogenic 5Ј nucle- METHODS — Four groups of patients assay with in vitro translated [35S]methi- ase assay application of the ABI PRISM and one of control subjects were investi- onine-labeled GAD65 (15) and IA-2IC 7900 HT Sequence Detection System gated. All subjects were unrelated and of (amino acids 605–979) (15). Results for (Applied Biosystems, Foster City, CA). exclusively Italian origin (with parents GADAs were converted into arbitrary The genotyping was performed using the and grandparents of Italian origin). units by extrapolation from a standard following primers: forward, 5Ј-CAA Adult-onset autoimmune diabetic curve with a local standard designated CTGCTCCAAGGATAGATGATGA-3Ј; subjects (n ϭ 250) (mean Ϯ SD age at 100 arbitrary units. The thresholds for reverse, 5ЈCCAGCTTCCTCCTCAAC onset 50.3 Ϯ 12.8 years) and age- and positivity were determined from the 99th CAATAAATG-3Ј; and TaqMan MGB sex-matched GADA-negative type 2 dia- centile of control subjects and corre- probes Fam TCAGGTGTCCGTACAGG betic subjects (n ϭ 450) (aged 51.6 Ϯ sponded to 3 arbitrary units for GADAs. and Vic TCAGGTGTGTCCATACAGG. 10.8 years) were selected from the Non The following performances were ob- Of the 10 ng/␮l stock of DNA, 4 ␮l Insulin Requiring Autoimmune Diabetes tained at the first, second, and third assay was dispensed into 384-well PCR plates (NIRAD) Study cohort of 5,330 type 2 proficiency evaluations of the Diabetes using a Biomek FX robot (Beckman diabetic subjects recruited between Feb- Antibody Standardization Program (17) Coulter, Fullerton, CA) to which 2 ␮lofa ruary 2001 and January 2006 (4,250 performed between 2002 and 2005: sen- mix containing primers, MGB probes, from February 2001 to June 2004 [5] and sitivity 84, 86, and 88%; and specificity and TaqMan Universal PCR Master Mix an additional 1,080 from July 2004 to 97, 97, and 92%. The intra- and interas- (Applied Biosystems) was added in accor- January 2006; the NIRAD study is an on- say coefficients of variation of GADA for dance with the manufacturer’s instruc-

DIABETES CARE, VOLUME 31, NUMBER 3, MARCH 2008 535 PTPN22 gene in adult-onset autoimmune diabetes

Table 2—Distribution of genotype, allele, and phenotype frequencies of the PTPN22 C1858T polymorphism in type 1 diabetic, autoimmune diabetic according to GADA titer, type 2 diabetic, and control subjects

Type 1 High GADA titer Low GADA titer Type 2 diabetes autoimmune diabetes autoimmune diabetes diabetes Control n 558 123 127 450 545 Genotypes* C1858C 448 (80.3) 98 (79.7) 120 (94.5) 424 (94.2) 496 (91.1) C1858T 105 (18.8) 24 (19.5) 7 (5.5) 25 (5.6) 47 (8.6) T1858T 5 (0.9) 1 (0.8) 0 (0) 1 (0.2) 2 (0.4) Alleles† C 1001 (89.6) 220 (89.5) 247 (97.3) 873 (97) 1039 (95.4) T 115 (10.4) 26 (10.5) 7 (2.7) 27 (3) 51 (4.6) Phenotypes‡ CT/TT genotypes 110 (19.7) 25 (20.3) 7 (5.5) 26 (5.8) 49 (9) CC genotype 448 (80.3) 98 (79.7) 120 (94.5) 424 (94.2) 496 (91) Data are n (%). *␹2 3 ϫ 2 high GADA titer vs. low GADA titer, type 2 diabetic, and control subjects, P Յ 0.008 after Bonferroni correction. Type 1 diabetic vs. low GADA titer, type 2 diabetic, and control subjects, P Յ 0.002 after Bonferroni correction. †␹2 2 ϫ 2 high GADA titer vs. low GADA titer, type 2 diabetic, and control subjects, P Յ 0.003 after Bonferroni correction. Type 1 diabetic vs. low GADA titer, type 2 diabetic, and control subjects, P Ͻ 0.0001 after Bonferroni correction. ‡␹2 2 ϫ 2 high GADA titer vs. low GADA titer and type 2 diabetic and control subjects (OR 2.6 [95% CI 1.5–4.4]), P Յ 0.002 after Bonferroni correction. Type 1 diabetic vs. low GADA titer, type 2 diabetic, and control subjects, P Ͻ 0.0001 after Bonferroni correction. tions. These were sealed with optical seals [DRB1*04 different from DRB1*0403]; high GADA titer (20.3%) compared with (Applied Biosystems) and incubated at moderate risk genotypes: DRB1*04- low GADA titer (5.5%), type 2 diabetic 95°C for 10 min followed by 40 cycles at DQB1*0302/DRB1*04-DQB1*0302, (5.8%), and control groups (9%) (P Ͻ 95°C for 15 s and 60°C for 1 min before DRB1*03-DQB1*0201/DRB1*03- 0.001 comparisons) conferring an OR of analysis on a 7900HT plate reader (Ap- DQB1*0201, DRB1*04-DQB1*0302/X, 2.6 (95% CI 1.5–4.4). No differences plied Biosystems). and DRB1*03/X [X different from were observed in the C1858T genotype DRB1*03, DRB1*0403-DQB1*0302, distribution between men and women Statistical analysis and DQB1*0602/03]; and low-risk geno- (data not shown). Moreover, no signifi- Statistical analysis was performed using types: all the other genotypes) (18). cant differences were observed in the fre- SPSS (version 13; SPSS, Chicago, IL). Ge- quency of the 1858T allele between notype, allele, and phenotype frequency RESULTS — Table 2 shows the geno- subjects carrying high and moderate HLA distributions were compared using the ␹2 type, allele, and phenotype frequencies of risk genotypes and those carrying low test or Fisher’s exact test when the criteria the PTPN22 C1858T variant in type 1 di- HLA risk genotypes either in high or low of the ␹2 test were not fulfilled. Allele and abetic, high GADA titer, low GADA titer, GADA titer subjects (Table 3). genotype frequencies of the PTPN22 type 2 diabetic, and control subjects. The C1858T were in Hardy-Weinberg equi- analysis of the PTPN22 C1858T variant CONCLUSIONS — In the present librium in all groups analyzed. We con- revealed similar frequencies in subjects study we have shown that the PTPN22 sidered all P Ͻ 0.05 values as statistically with type 1 diabetes and high GADA titers 1858T variant is associated only with high significant. A Bonferroni correction of a regarding genotype, allele, and pheno- GADA titer adult-onset autoimmune dia- factor 4 was applied. Based on an odds type distributions; the risk conferred by betes, giving important support to our ratio (OR) of 2.31 conferred by the 1858T the 1858T variant in type 1 diabetes (OR previous findings in which the GADA titer variant in a previous study (6) and on pre- 2.48, 95% CI 1.7–3.5) resulted indepen- identified two subgroups of subjects liminary evaluations, the present study dently from age at disease onset (data not (with high or low GADA titers) with dif- should be able to identify statistical differ- shown). No significant differences in the ferences in clinical phenotype supported ences between type 1 diabetic, high frequencies were observed between low by a different HLA class II susceptibility GADA titer, low GADA titer, type 2 dia- GADA titer, type 2 diabetic, and control (5). Our results further extend previous betic, and control subjects with a power subjects. Conversely, a significant in- studies indicating an association between of 75% and P value of 5%. Conditioning crease in T1858T and C1858T genotypes the PTPN22 gene variant and classic type analysis based on HLA risk genotypes was was observed in high GADA titer com- 1 diabetes in young subjects (6). The re- also performed. HLA genotypes were pared with low GADA titer, type 2 dia- ported association of this variant with classified in three risk categories (high, betic, and control subjects (P Յ 0.002 for other autoimmune diseases such as rheu- moderate, and low) based on the absolute all comparisons). matoid arthritis (8), systemic ery- risk values for type 1 diabetes previously The PTPN22 1858T allele frequency thematosus (9), Graves’ disease (10), and estimated in Italian population (18). HLA was significantly increased in high GADA Hashimoto’s thyroiditis (20) underscore genotypes were introduced as a dichoto- titer (10.5%) compared with low GADA the importance of the product of the mous variable in the analysis as follows: titer (2.7%), type 2 diabetic (3%), and PTPN22 gene in the immune regulation. high and moderate risk ϭ 1, low risk ϭ 0 control subjects (4.6%) (P Ͻ 0.0001 for Kyogoku et al. (9) proposed that the (high-risk genotype: DRB1*03- all comparisons). The frequency of 1858T C1858T polymorphism of the PTPN22 DQB1*0201/DRB1*04-DQB1*0302 carriers was significantly increased in gene may predispose individuals to auto-

536 DIABETES CARE, VOLUME 31, NUMBER 3, MARCH 2008 Petrone and Associates

Table 3—Distribution of genotype, allele, and phenotype frequencies of the PTPN22 C1858T polymorphism in subjects carrying high and moderate HLA risk genotypes and in low HLA risk genotypes with either high or low GADA titer

High GADA titer autoimmune diabetes (n ϭ 123) Low GADA titer autoimmune diabetes (n ϭ 127) High and moderate HLA risk* Low HLA risk† High and moderate HLA risk* Low HLA risk† n 42 81 35 92 Genotypes C1858C 33 (78.6) 65 (80.2) 33 (94.3) 87 (94.6) C1858T 8 (19) 16 (19.8) 2 (5.7) 5 (5.4) T1858T 1 (2.4) 0 (0) 0 (0) 0 (0) Alleles C 74 (88.1) 146 (90.1) 68 (97.1) 179 (97.3) T 10 (11.9) 16 (9.9) 2 (2.9) 5 (2.7) Phenotypes CT/TT genotypes 9 (21.4) 16 (19.8) 2 (5.7) 5 (5.4) CC genotype 33 (78.6) 65 (80.2) 33 (94.3) 87 (94.6) Data are n (%). *High-risk genotypes: DRB1*03-DQB1*0201/DRB1*04-DQB1*0302 genotype (DRB1*04 different from 0403); moderate-risk genotypes: DRB1*04-DQB1*0302/DRB1*04-DQB1*0302, DRB1*03-DQB1*0201/DRB1*03-DQB1*0201, DRB1*04-DQB1*0302/X, and DRB1*03/X (X different from DRB1*03, DRB1*0403-DQB1*0302, or DQB1*0602/03) genotypes. †Low risk genotypes: all the other genotypes. immune disease by facilitating the gener- children possessing the PTPN22 T1858T References ation of certain disease-associated genotype compared with children with 1. Turner R, Stratton I, Horton V, Manley S, autoantibodies, thereby contributing to the CC genotype. Altered LYP function, Zimmet P, Mackay IR, Shattock M, Bot- disease progression. Thus, the existence codified by the 1858T variant in tazzo GF, Holman R: UKPDS 25: autoan- ϩ ϩ of humoral abnormalities in the PTPN8 CD4 CD25 T regulatory cells to make tibodies to islet-cell cytoplasm and knockout mouse, the ortholog mouse them less potent in suppressing immune glutamic acid decarboxylase for predic- gene of PTPN22, is consistent with the fact response could explain more aggressive tion of insulin requirement in type 2 dia- that autoantibody production is a promi- ␤-cell destruction and consequently a betes: UK Prospective Diabetes Study Group. Lancet 350:1288–1293, 1997 nent feature of all human autoimmune major loss in ␤-cell function in patients diseases that are significantly associated 2. Bottazzo GF, Bosi E, Cull CA, Bonifacio E, carrying this gene variant (24). Locatelli M, Zimmet P, Mackay IR, Hol- with the PTPN22 C1858T gene variant In summary, the association of this (20). Overall these data suggest that the man RR: IA-2 antibody prevalence and genetic variant with high GADA titer au- risk assessment of early insulin require- PTPN22 gene plays a role in autoanti- toimmune diabetes supports the hypoth- ment in subjects presenting with type 2 body-related autoimmune diseases as re- diabetes (UKPDS 71). Diabetologia cently supported by data in systemic esis that in these subjects the autoimmune process induces diabetes with no major 48:703–708, 2005 lupus erythematosus (9). 3. Tuomi T, Carlsson A, Li H, Isomaa B, Mi- In type 1 diabetes, early histological contribution from other concomitant ettinen A, Nilsson A, Nissen M, Ehrn- and functional studies indicate that the pathogenic processes. On the other hand, strom BO, Forsen B, Snickars B, Lahti K, disease is caused by T lymphocytes infil- the lack of association with low GADA Forsblom C, Saloranta C, Taskinen MR, trating the ␤-cells (21). The disease- titer autoimmune diabetes, possibly re- Groop LC: Clinical and genetic character- specific immune events are, however, flecting a less intense autoimmune pro- istics of type 2 diabetes with and without reflected in the appearance of ␤-cell– cess, implies that the appearance of a low GAD antibodies. Diabetes 48:150–157, specific autoantibodies, which are useful GADA titer alone seems not to be related 1999 4. Zinman B, Kahn SE, Haffner SM, O’Neill tools for prediction of progression to clin- to the diabetes-specific autoimmune pro- MC, Heise MA, Freed MI, ADOPT Study ical disease (22,23). In this view, the cess (5). Finally, our data suggest that the Group: Phenotypic characteristics of GAD present findings support and extend our PTPN22 1858T variant controlling T cells antibody-positive recently diagnosed pa- previous study, demonstrating an associ- is involved in humoral autoimmunity. tients with type 2 diabetes in North Amer- ation of PTPN22 exclusively with high ica and Europe. Diabetes 53:3193–3200, GADA titer, thus suggesting that the 2004 5. Buzzetti R, Di Pietro S, Giaccari A, Petrone C1858T gene variant may be associated to Acknowledgments— This study was spon- A, Locatelli M, Suraci C, Capizzi M, Arpi autoimmune diseases only when autoan- sored by the Foundation for the Research of tibodies are a marker of pathogenic cell ML, Bazzigaluppi E, Dotta F, Bosi E, Non Societa` Italiana di Diabetologia (Fo.ri.SID) Insulin Requiring Autoimmune Diabetes destruction. based on an unconditioned research grant The PTPN22 C1858T variant has Study Group: High titer of autoantibodies from Novo Nordisk, Italy. to GAD identifies a specific phenotype of been found to be a marker of disease pro- The authors are indebted to Professors Ric- adult-onset autoimmune diabetes. Diabe- gression and to regulate diabetes-specific cardo Giorgino and Riccardo Vigneri of tes Care 30:932–938, 2007 autoimmunity (12). This progression was Fo.ri.SID for their valuable and continuous 6. Bottini N, Musumeci L, Alonso A, Rah- demonstrated by a fourfold higher risk of support. We also acknowledge the IMDIAB mouni S, Nika K, Rostamkhani M, Mac- developing an additional autoantibody group, Rome, for providing genetic data on Murray J, Meloni GF, Lucarelli P, carried out by islet cell antibody–positive their type 1 diabetic patients. Pellecchia M, Eisenbarth GS, Comings D,

DIABETES CARE, VOLUME 31, NUMBER 3, MARCH 2008 537 PTPN22 gene in adult-onset autoimmune diabetes

Mustelin T: A functional variant of lym- Autoimmune-associated lymphoid ty- Locatelli M, Bottazzo G, Di Mario U: Ge- phoid tyrosine phosphatase is associated rosine phosphatase is a gain-of-function netic prediction of type 1 diabetes in a with type I diabetes. Nat Genet 36:337– variant. Nat Genet 37:1317–1319, 2005 population with low frequency of HLA 338, 2004 12. Hermann R, Lipponen K, Kiviniemi M, risk genotypes and low incidence of the 7. Cloutier JF, Veillette A: Cooperative inhi- Kakko T, Veijola R, Simell O, Knip M, disease (the DIABFIN study). Diabetes bition of T-cell antigen receptor signaling Ilonen J: Lymphoid tyrosine phosphatase Metab Res Rev 20:137–143, 2004 by a complex between a kinase and a (LYP/PTPN22) Arg620Trp variant regu- 19. Erlich H, Bugawan T, Begovich AB, Scharf phosphatase. J Exp Med 189:111–121, lates insulin autoimmunity and progres- S, Griffith R, Saiki R, Higuchi R, Walsh PS: 1999 sion to type 1 diabetes. Diabetologia 49: HLA-DR, DQ and DP typing using PCR 8. Begovich AB, Carlton VE, Honigberg LA, 1198–1208, 2006 amplification and immobilized probes. Schrodi SJ, Chokkalingam AP, Alexander 13. Lee YH, Rho YH, Choi SJ, Ji JD, Song GG, Eur J Immunogenet 18:33–55, 1991 HC, Ardlie KG, Huang Q, Smith AM, Nath SK, Harley JB: The PTPN22 C1858T 20. Criswell LA, Pfeiffer KA, Lum RF, Gonza- Spoerke JM, Conn MT, Chang M, Chang functional polymorphism and autoim- les B, Novitzke J, Kern M, Moser KL, Be- SY, Saiki RK, Catanese JJ, Leong DU, Gar- mune diseases—a meta-analysis. Rheu- govich AB, Carlton VE, Li W, Lee AT, cia VE, McAllister LB, Jeffery DA, Lee AT, matology (Oxford) 46:49–56, 2007 Ortmann W, Behrens TW, Gregersen PK: Batliwalla F, Remmers E, Criswell LA, Sel- 14. Palmer JP, Fleming GA, Greenbaum CJ, Analysis of families in the multiple auto- din MF, Kastner DL, Amos CI, Sninsky JJ, Herold KC, Jansa LD, Kolb H, Lachin JM, immune disease genetics consortium Gregersen PK: A missense single-nucle- Polonsky KS, Pozzilli P, Skyler JS, Steffes (MADGC) collection: the PTPN22 620W otide polymorphism in a gene encoding a MW: C-peptide is the appropriate out- allele associates with multiple autoim- protein tyrosine phosphatase (PTPN22) is come measure for type 1 diabetes clinical mune phenotypes. Am J Hum Genet 76: associated with . Am J trials to preserve ␤-cell function: report of 561–571, 2005 Hum Genet 75:330–337, 2004 an ADA workshop, 21–22 October 2001. 21. Roep BO, Arden SD, de Vries RR, Hutton 9. Kyogoku C, Langefeld CD, Ortmann WA, Diabetes 53:250–264, 2004 JC: T-cell clones from a type-1 diabetes Lee A, Selby S, Carlton VE, Chang M, 15. Bonifacio E, Genovese S, Braghi S, Bazzi- patient respond to insulin secretory gran- Ramos P, Baechler EC, Batliwalla FM, No- galuppi E, Lampasona V, Bingley PJ, ule . Nature 345:632–634, 1990 vitzke J, Williams AH, Gillett C, Rodine P, Rogge L, Pastore MR, Bognetti E, Bottazzo 22. Ziegler AG, Hummel M, Schenker M, Graham RR, Ardlie KG, Gaffney PM, GF, Gale EAM, Bosi E: Islet autoantibody Bonifacio E: Autoantibody appearance Moser KL, Petri M, Begovich AB, Gre- markers in insulin dependent diabetes: and risk for development of childhood di- gersen PK, Behrens TW: Genetic associa- risk assessment strategies yielding high abetes in offspring of parents with type 1 tion of the R620W polymorphism of sensitivity. Diabetologia 38:816–822, diabetes: the 2-year analysis of the Ger- protein tyrosine phosphatase PTPN22 1995 man BABYDIAB Study. Diabetes 48:460– with human SLE. J Hum Genet 75:504– 16. Visalli N, Sebastiani L, Adorisio E, Conte 468, 1999 507, 2004 A, De Cicco AL, D’Elia R, Manfrini S, 23. Kukko M, Kimpimaki T, Korhonen S, 10. Velaga MR, Wilson V, Jennings CE, Owen Pozzilli P, IMDIAB Group: Environmental Kupila A, Simell S, Veijola R, Simell T, CJ, Herington S, Donaldson PT, Ball SG, risk factors for type 1 diabetes in Rome Ilonen J, Simell O, Knip M: Dynamics of James RA, Quinton R, Perros P, Pearce and province. Arch Dis Child 88:695–698, diabetes-associated autoantibodies in SH: The codon 620 tryptophan allele of 2003 young children with human leukocyte an- the lymphoid tyrosine phosphatase (LYP) 17. Bingley PJ, Bonifacio E, Mueller PW: Dia- tigen-conferred risk of type 1 diabetes re- gene is a major determinant of Graves’ betes Antibody Standardization Program: cruited from the general population. J Clin disease. J Clin Endocrinol Metab 89:5862– first assay proficiency evaluation. Diabetes Endocrinol Metab 90:2712–2717, 2005 5865, 2004 52:1128–1136, 2003 24. Bottini N, Vang T, Cucca F, Mustelin T: 11. Vang T, Congia M, Macis MD, Musumeci 18. Buzzetti R, Galgani A, Petrone A, Del Role of PTPN22 in type 1 diabetes and L, Orru V, Zavattari P, Nika K, Tautz L, Buono ML, Erlich HA, Bugawan TL, other autoimmune diseases. Semin Immu- Tasken K, Cucca F, Mustelin T, Bottini N: Lorini R, Meschi F, Multari G, Pozzilli P, nol 18:207–213, 2006

538 DIABETES CARE, VOLUME 31, NUMBER 3, MARCH 2008