It has been proposed that type 1 and 2 diabetes might share common pathophysiological pathways and, to some extent, genetic background. However, to date there has been no convincing data to establish a molecular genetic link between them. We have genotyped three single nucleotide polymorphisms associated with type 2 diabetes in a large type 1 diabetic family collection of European descent: Gly972Arg in the insulin receptor substrate 1 (IRS1) gene, Glu23Lys in the potassium inwardly-rectifying channel gene (KCNJ11), and Pro12Ala in the peroxisome proliferative-activated receptor γ2 gene (PPARG2). We were unable to confirm a recently published association of the IRS1 Gly972Arg variant with type 1 diabetes. Moreover, KCNJ11 Glu23Lys showed no association with type 1 diabetes (P > 0.05). However, the PPARG2 Pro12Ala variant showed evidence of association (RR 1.15, 95% CI 1.04–1.28, P = 0.008). Additional studies need to be conducted to confirm this result.

Type 1 and 2 diabetes have been considered genetically and pathophysiologically distinct diseases, although clinically, it can sometimes be difficult to distinguish between them in individuals with deficient or absent endogenous insulin secretion, a feature common to both categories. Evidence does exist suggesting that there may be shared etiological features (13). Occasionally, individuals who develop diabetes late in life initially present in a similar manner to type 2 diabetes, but develop progressive β-cell failure together with evidence of autoimmunity to islet cells, which is characteristic of type 1 diabetes. This so-called latent autoimmune diabetes in adults could suggest an overlap in the pathogenesis of type 1 and 2 diabetes in some cases. Furthermore, a recent study (4) has reported that early in the course of type 1 diabetes, individuals with evidence of islet autoimmunity may have normal fasting glucose but abnormal postprandial glucose tolerance. This pattern of defective postprandial secretion of insulin in response to nutrient intake is often observed in type 2 diabetes as well (4).

There has been some evidence of familial clustering of type 1 and 2 diabetes (5,6). However, the HLA region, the major locus in type 1 diabetes, has only infrequently been associated with type 2 diabetes (7). Moreover, both diseases have been associated with variation in the insulin gene but with predisposition determined by opposite alleles of the 5′ variable number tandem repeat locus (8). Recently, genetic studies in type 2 diabetes have begun to reveal disease-associated polymorphisms that have established some level of general acceptance because they have been replicated by independent studies. Two polymorphisms that are well established as associated with type 2 diabetes are the single nucleotide polymorphisms (SNPs) Pro12Ala in the peroxisome proliferator-activated receptor γ gene isoform 2 (PPARG2), resulting from a C>G transversion, and Glu23Lys in the potassium inwardly rectifying channel gene (KCNJ11), resulting from an A>G transition (9). The reported effect sizes are modest (odds ratio [OR] 1.2–1.3) in meta-analyses (9,10). A large collaborative collection of type 1 diabetic families is now available for genetic analysis (11) with sufficient power to detect effects in the OR range of 1.2–1.3. Hence, in the present report we investigated the possibility that the two type 2 diabetes-associated SNPs were also associated with type 1 diabetes in an effort to begin to establish if overlaps exist in the two disease pathways. We have also tested the insulin receptor substrate 1 (IRS1) Gly972Arg polymorphism because it has been reported to be associated with type 1 diabetes (12), although it is not as well established in type 2 diabetes as the PPARG2 and KCNJ11 SNPs.

The PPARG gene encodes peroxisome proliferator-activated receptor γ (PPARγ), a member of the nuclear receptor superfamily that has regulatory functions in adipocyte differentiation and glucose homeostasis (13). PPARγ is a target for a large family of antidiabetic drugs, the thiazolidinediones, that act as PPARγ agonists and increase hepatic and peripheral insulin action in type 2 diabetes (14). In a meta-analysis of published studies through February 2003 (9), an OR of 1.27 was estimated for the Pro allele of isoform 2 (P value <2 × 10−8) in type 2 diabetes. This association is supported by functional data, as in vitro studies have shown that the Ala allele decreases the DNA-binding affinity of the PPARγ2, thus reducing its transcriptional activity (15).

The KCNJ11 Glu23Lys variant has also been consistently reported (10,16,17) to be associated with type 2 diabetes. KCNJ11 encodes for KIR6.2, which comprises one of the two subunits of the β-cell ATP-sensitive potassium channel that regulates insulin secretion. A meta-analysis combining published case-control studies up to September 2002 (10) gave an OR of 1.23 (95% CI 1.12–1.36, P = 1.5 × 10−5) for the Lys allele, suggesting that this allele is associated with type 2 diabetes. In addition, functional data shows that the Lys variant may alter type 2 diabetes susceptibility by increasing the threshold ATP concentration necessary for insulin release, thus inducing spontaneous overactivity of pancreatic β-cells (18). Increasing the metabolic activity of β-cells might confer susceptibility to type 1 diabetes as well (19).

The insulin receptor substrate proteins (IRS1 and IRS2) are expressed in a variety of insulin responsive cells and tissues, mediating metabolic and growth-promoting actions of insulin and IGF-1. Observations from knockout mice suggest a role for these proteins in regulation of β-cell function (20), and the human IRS1 Gly972Arg variant has been reported to be associated with type 2 diabetes (9). Meta-analysis, based on 3,408 case subjects and 5,419 control subjects from studies before January 2002 (21), resulted in an OR of 1.25 (95% CI 1.05–1.48). Of direct relevance to our study is a recent study (12) reporting an association of this variant with type 1 diabetes, with an OR of 2.5 (P = 0.0008) in a case-control study and a transmission/disequilibrium test (TDT) in simplex families giving a P value <0.02.

Therefore, we genotyped these three polymorphisms in 2,434 type 1 diabetic families of European descent. Parental genotype frequencies were found to be consistent with Hardy-Weinberg equilibrium. TDT and conditional logistic regression analyses were performed, and the results are summarized in Table 1. In this large dataset of type 1 diabetic families, we were unable to confirm the reported association of IRS1 (12), although we had >95% power to detect the reported effect (OR 2.5, P = 0.0008, minor allele frequency = 8%). Similarly, the KCNJ11 Glu23Lys variant was not associated with type 1 diabetes, as a nonsignificant deviation from 50:50 transmission from heterozygous parents was observed.

The PPARG2 Pro12Ala SNP, however, showed some evidence of association with the more common Pro allele conferring risk to type 1 diabetes (relative risk [RR] 1.15, 95% CI 1.04–1.28, P = 0.008), as is the case for type 2 diabetes. TDT results indicated a transmission to affected individuals of 54% (769 of 1,437 informative transmissions, P = 0.008). To confirm that this result was not due to genotyping error, we regenotyped the PPARG2 Pro12Ala SNP on the reverse strand and obtained 99.77% concordance. In agreement with the reports in type 2 diabetes, the minor allele (G) is protective in type 1 diabetes. Genotype RRs were calculated by conditional logistic regression analysis for the Pro12Ala SNP, giving values of 0.85 (0.76–0.95, P = 0.005) and 0.87 (0.61–1.24, P = 0.44) for the C/G and G/G genotypes, respectively, relative to the C/C genotype. Conversely, relative to the G/G genotype, the C/G and C/C genotypes have RRs of 0.98 (0.69–1.39) and 1.15 (0.81–1.64), respectively. The nonsignificant result for the G/G genotype (and the risks relative to the G/G genotype) is likely a result of its low frequency.

Our failure to replicate the Italian study (12), which reported an association between IRS1 Gly972Arg and type 1 diabetes, is perhaps not surprising given the modest sample size previously studied. Nevertheless, we genotyped this SNP with two technologies, as family studies of rare variants may be compromised because apparent undertransmission of alleles can be observed due to genotyping errors. A concordance rate of 99.74% was observed between genotypes generated by the two methods. It is unlikely that this common SNP could be genetically associated with type 1 diabetes in central Italy and not in other European countries. In these data we found no evidence of population heterogeneity (χ2 [4 df] = 1.93, P = 0.75, in five populations). Hence, our study highlights the importance of large datasets to investigate the multigenic basis of type 1 diabetes, type 2 diabetes, and other common multifactorial diseases.

If the PPARG2 Pro12Ala association is confirmed in other studies, then a molecular link would be established between type 1 and 2 diabetes involving a protective effect of the minor allele (Ala). Other studies have reported that thiazolidinediones possess anti-inflammatory properties and decrease diabetes incidence in nonobese diabetic (NOD) mice (22,23). Prevention of autoimmune diabetes in NOD mice by thiazolidinediones has been associated with suppression of intercellular adhesion molecule 1 (ICAM1) expression and alteration of Th1/Th2 cytokine balance (24). Interestingly, we have recently reported (25) an association of the Gly241Arg polymorphism of ICAM1 with type 1 diabetes. It is possible that PPARγ agonists may be worth testing in prevention of immune rejection of transplanted islets.

All families were Caucasian of European descent and were composed of two parents and at least one affected child. The families comprised up to 458 Diabetes U.K. Warren 1 multiplex from the U.K. (26), 328 U.S. multiplex from the Human Biological Data Interchange (27), 80 Yorkshire simplex from the U.K., 250 Belfast multiplex/simplex (28) from the U.K., 159 Norwegian simplex, 233 Romanian simplex, and 926 Finnish multiplex/simplex (29). All DNA samples were collected with appropriate ethical approval and informed consent.

Genotyping.

SNPs were genotyped by the TaqMan 5′ nuclease assay according to the manufacturer’s instructions (Applied Biosystems, Warrington, U.K.). TaqMan primers and probes were designed by Applied Biosystems. The PPARG2 Pro12Ala SNP was independently typed by the TaqMan 5′ nuclease assay on both strands. IRS1 Gly972 Arg was also genotyped by restriction enzyme digest with XmaI, incorrectly reported as MvaI in Federici et al. (12). All genotyping data were double scored to minimize error.

Statistical analysis.

All statistical analyses were performed in Stata (http://www.stata.com) using the Genassoc package (http://www-gene.cimr.cam.ac.uk/clayton/software/stata). A modified test was used to evaluate Hardy-Weinberg equilibrium that allows for allelic frequencies to differ between known population groups. Allelic association was tested using the TDT. Pseudo-control subjects, generated by conditioning on parental genotype, allowed us to test genotypic association with conditional logistic regression (30). This also allowed us to test for population heterogeneity by adding appropriate interaction terms into the regression equation. Robust variance estimates were used for the calculation of P values and 95% CIs in order to correct for nonindependence of transmissions within families with more than one affected offspring. As a quality control measure, the transmission of PPARG2 Pro12Ala alleles to unaffected offspring was evaluated by TDT and found to be nonsignificant (data not shown), ruling out genotyping error as a possible explanation for the association observed with type 1 diabetes.

TABLE 1

Type 1 diabetes SNP association analysis

SNPReported risk allele for type 2 diabetesFrequency (%)TDT
CLR
TUPPRR95% CI
*IRSI Gly972Arg Arg 5.9 312 327 0.54 0.77 0.95 0.82–1.11 
KCNJ11 Glu23Lys Lys 42 1268 1300 0.54 0.72 0.97 0.85–1.10 
PPARG2 Pro12Ala Pro 86 769 668 0.0084 0.018 1.15 1.04–1.28 
SNPReported risk allele for type 2 diabetesFrequency (%)TDT
CLR
TUPPRR95% CI
*IRSI Gly972Arg Arg 5.9 312 327 0.54 0.77 0.95 0.82–1.11 
KCNJ11 Glu23Lys Lys 42 1268 1300 0.54 0.72 0.97 0.85–1.10 
PPARG2 Pro12Ala Pro 86 769 668 0.0084 0.018 1.15 1.04–1.28 

P values were calculated based on the null hypothesis of no association of the polymorphism. Conditional logistic regression (CLR) P values are for χ2 on 2 degrees of freedom; RRs are for the reported risk allele. Frequency given is the frequency in unaffected parents of the reported risk allele for type 2 diabetes.

*

Successfully typed in 2,161 families (2,887 affected offspring).

Successfully typed in 2,090 families (2,734 affected offspring).

Successfully typed in 2,355 families (3,157 affected offspring). T, transmitted; U, untransmitted.

We thank the Juvenile Diabetes Research Foundation and the Wellcome Trust for financial support.

We also thank the Human Biological Data Interchange and Diabetes U.K. for U.S. and U.K. multiplex families, respectively, the Norwegian Study Group for Childhood Diabetes for sample collection of Norwegian samples, Sarah Field and Tasneem Hassanali for DNA preparation, and David Clayton, Jason Cooper, and Chris Lowe for discussions.

1.
Donath MY, Storling J, Maedler K, Mandrup-Poulsen T: Inflammatory mediators and islet beta-cell failure: a link between type 1 and type 2 diabetes.
J Mol Med
81
:
455
–470,
2003
2.
Mathis D, Vence L, Benoist C: Beta-cell death during progression to diabetes.
Nature
414
:
792
–798,
2001
3.
Wilkin TJ: The accelerator hypothesis: weight gain as the missing link between type I and type II diabetes.
Diabetologia
44
:
914
–922,
2001
4.
Greenbaum CJ, Prigeon RL, D’Alessio DA: Impaired β-cell function, incretin effect, and glucagon suppression in patients with type 1 diabetes who have normal fasting glucose.
Diabetes
51
:
951
–957,
2002
5.
Dahlquist G, Blom L, Tuvemo T, Nystrom L, Sandstrom A, Wall S: The Swedish childhood diabetes study: results from a nine year case register and a one year case-referent study indicating that type 1 (insulin-dependent) diabetes mellitus is associated with both type 2 (non-insulin-dependent) diabetes mellitus and autoimmune disorders.
Diabetologia
32
:
2
–6,
1989
6.
Li H, Lindholm E, Almgren P, Gustafsson A, Forsblom C, Groop L, Tuomi T: Possible human leukocyte antigen-mediated genetic interaction between type 1 and type 2 diabetes.
J Clin Endocrinol Metab
86
:
574
–582,
2001
7.
Tuomilehto-Wolf E, Tuomilehto J, Hitman GA, Nissinen A, Stengard J, Pekkanen J, Kivinen P, Kaarsalo E, Karvonen MJ: Genetic susceptibility to non-insulin dependent diabetes mellitus and glucose intolerance are located in HLA region.
BMJ
307
:
155
–159,
1993
8.
Bennett ST, Todd JA: Human type 1 diabetes and the insulin gene: principles of mapping polygenes.
Annu Rev Genet
30
:
343
–370,
1996
9.
Florez JC, Hirschhorn J, Altshuler D: The inherited basis of diabetes mellitus: implications for the genetic analysis of complex traits.
Ann Rev Genomics Hum Genet
4
:
257
–291,
2003
10.
Gloyn AL, Weedon MN, Owen KR, Turner MJ, Knight BA, Hitman G, Walker M, Levy JC, Sampson M, Halford S, McCarthy MI, Hattersley AT, Frayling TM: Large-scale association studies of variants in genes encoding the pancreatic β-cell KATP channel subunits Kir6.2 (KCNJ11) and SUR1 (ABCC8) confirm that the KCNJ11 E23K variant is associated with type 2 diabetes.
Diabetes
52
:
568
–572,
2003
11.
Dahlman I, Eaves IA, Kosoy R, Morrison VA, Heward J, Gough SC, Allahabadia A, Franklyn JA, Tuomilehto J, Tuomilehto-Wolf E, Cucca F, Guja C, Ionescu-Tirgoviste C, Stevens H, Carr P, Nutland S, McKinney P, Shield JP, Wang W, Cordell HJ, Walker N, Todd JA, Concannon P: Parameters for reliable results in genetic association studies in common disease.
Nat Genet
30
:
149
–150,
2002
12.
Federici M, Petrone A, Porzio O, Bizzarri C, Lauro D, D’Alfonso R, Patera I, Cappa M, Nistico L, Baroni M, Sesti G, di Mario U, Lauro R, Buzzetti R: The Gly972→Arg IRS-1 variant is associated with type 1 diabetes in continental Italy.
Diabetes
52
:
887
–890,
2003
13.
Lee CH, Olson P, Evans RM: Minireview: lipid metabolism, metabolic diseases, and peroxisome proliferator-activated receptors.
Endocrinology
144
:
2201
–2207,
2003
14.
Spiegelman BM: PPAR-γ: adipogenic regulator and thiazolidinedione receptor.
Diabetes
47
:
507
–514,
1998
15.
Deeb SS, Fajas L, Nemoto M, Pihlajamaki J, Mykkanen L, Kuusisto J, Laakso M, Fujimoto W, Auwerx J: A Pro12Ala substitution in PPARgamma2 associated with decreased receptor activity, lower body mass index and improved insulin sensitivity.
Nat Genet
20
:
284
–287,
1998
16.
Gloyn AL, Hashim Y, Ashcroft SJ, Ashfield R, Wiltshire S, Turner RC: Association studies of variants in promoter and coding regions of beta-cell ATP-sensitive K-channel genes SUR1 and Kir6.2 with type 2 diabetes mellitus (UKPDS 53).
Diabet Med
18
:
206
–212,
2001
17.
Hani EH, Boutin P, Durand E, Inoue H, Permutt MA, Velho G, Froguel P: Missense mutations in the pancreatic islet beta cell inwardly rectifying K+ channel gene (KIR6.2/BIR): a meta-analysis suggests a role in the polygenic basis of type II diabetes mellitus in Caucasians.
Diabetologia
41
:
1511
–1515,
1998
18.
Schwanstecher C, Meyer U, Schwanstecher M: K(IR)6.2 polymorphism predisposes to type 2 diabetes by inducing overactivity of pancreatic β-cell ATP-sensitive K+ channels.
Diabetes
51
:
875
–879,
2002
19.
Palmer JP: Beta cell rest and recovery: does it bring patients with latent autoimmune diabetes in adults to euglycemia?
Ann N Y Acad Sci
958
:
89
–98,
2002
20.
Burks DJ, White MF: IRS proteins and β-cell function.
Diabetes
50 (Suppl. 1)
:
S140
–S145,
2001
21.
Jellema A, Zeegers MP, Feskens EJ, Dagnelie PC, Mensink RP: Gly972Arg variant in the insulin receptor substrate-1 gene and association with type 2 diabetes: a meta-analysis of 27 studies.
Diabetologia
46
:
990
–995,
2003
22.
Beales PE, Liddi R, Giorgini AE, Signore A, Procaccini E, Batchelor K, Pozzilli P: Troglitazone prevents insulin dependent diabetes in the non-obese diabetic mouse.
Eur J Pharmacol
357
:
221
–225,
1998
23.
Beales PE, Pozzilli P: Thiazolidinediones for the prevention of diabetes in the non-obese diabetic (NOD) mouse: implications for human type 1 diabetes.
Diabete Metab Res Rev
18
:
114
–117,
2002
24.
Augstein P, Dunger A, Heinke P, Wachlin G, Berg S, Hehmke B, Salzsieder E: Prevention of autoimmune diabetes in NOD mice by troglitazone is associated with modulation of ICAM-1 expression on pancreatic islet cells and IFN-gamma expression in splenic T cells.
Biochem Biophys Res Commun
304
:
378
–384,
2003
25.
Nezhentsev S, Guja C, McCormack R, Cooper J, Howson JMM, Nutland S, Rance H, Walker N, Undlien D, Ronningen KS, Tuomilehto-Wolf E, Tuomilehto J, Ionescu-Tirgoviste C, Gale EAM, Bingley PJ, Gillespie KM, Savage DA, Carson DJ, Patterson CC, Maxwell AP, Todd JA: The ARG241 allele of the intercellular adhesion molecule-1 is associated with a reduced risk of type 1 diabetes.
Lancet
362
:
1723
–1724,
2003
26.
Bain SC, Todd JA, Barnett AH: The British Diabetic Association: Warren repository.
Autoimmunity
7
:
83
–85,
1990
27.
Lernmark A, Ducat L, Eisenbarth G, Ott J, Permutt MA, Rubenstein P, Spielman R: Family cell lines available for research.
Am J Hum Genet
47
:
1028
–1030,
1990
28.
Patterson CC, Carson DJ, Hadden DR, the Northern Ireland Diabetes Study Group: Epidemiology of childhood IDDM in Northern Ireland 1989–1994: low incidence in areas with highest population density and most household crowding.
Diabetologia
39
:
1063
–1069,
1996
29.
Tuomilehto J, Lounamaa R, Tuomilehto-Wolf E, Reunanen A, Virtala E, Kaprio EA, Akerblom HK, the Childhood Diabetes in Finland (DiMe) Study Group: Epidemiology of childhood diabetes mellitus in Finland: background of a nationwide study of type 1 (insulin-dependent) diabetes mellitus.
Diabetologia
35
:
70
–76,
1992
30.
Cordell HJ, Clayton DG: A unified stepwise regression procedure for evaluating the relative effects of polymorphisms within a gene using case/control or family data: application to HLA in type 1 diabetes.
Am J Hum Genet
70
:
124
–141,
2002