Polymorphic Variation in the CBLB Gene in Human Type 1 Diabetes

BRIEF COMMUNICATION Polymorphic variation in the CBLB gene in human type 1 diabetes

R Kosoy1,2, N Yokoi3, S Seino3,4 and P Concannon1,2 1Molecular Genetics Program, Benaroya Research Institute, Seattle, WA, USA; 2Department of Immunology, University of Washington School of Medicine, Seattle, WA, USA; 3Division of Cellular and Molecular Medicine, Kobe University Graduate School of Medicine, Kobe, Japan; 4Department of Cellular and Molecular Medicine, Graduate School of Medicine, Chiba University, Chiba, Japan

CBLB was evaluated as a candidate gene for type 1 diabetes (T1D) susceptibility based on its association with autoimmunity in animal models and its role in T-cell costimulatory signaling. Cblb is one of the two major diabetes predisposing loci in the Komeda diabetes-prone (KDP) rat. Cbl-b, a ubiquitin ligase, couples TCR-mediated stimulation with the requirement for CD28 costimulation, regulating T-cell activation. To identify variants with possible effects on gene function as well as haplotype tagging polymorphisms, the human CBLB coding region was sequenced in 16 individuals with T1D: no variants predicted to change the amino-acid sequence were identified. Seven single-nucleotide polymorphism (SNP) markers spanning the CBLB gene were genotyped in multiplex T1D families and assessed for disease association by transmission disequilibrium testing. No significant evidence of association was obtained for either individual markers or marker haplotypes. Genes and Immunity (2004) 5, 232–235. doi:10.1038/sj.gene.6364057 Published online 12 February 2004

Keywords: diabetes; T cell; SNP; haplotype

Introduction studies in human T1D families and studies in the NOD mouse model of T1D.9,10 Type I diabetes (T1D) arises from autoimmune destruc- In the current study, we have evaluated another tion of the insulin-secreting b-cells of the pancreas candidate T1D gene, CBLB, which is implicated based requiring insulin replacement therapy for survival. T1D on studies in another animal model of T1D, the Komeda is a multifactorial disorder with contributions from diabetes-prone (KDP) rat. The KDP rat model of multiple genetic and environmental components.1 There autoimmune diabetes is characterized by a rapid onset is intense interest in the identification of these risk factors of disease, but without the sex difference that char- both for their potential benefits in understanding specific acterizes the NOD mouse, or the lymphopenia that is pathology of T1D and the more general insights they characteristic of the BB rat.11 KDP rats also exhibit might provide into mechanisms of human autoimmu- additional autoimmune phenotypes that include lym- nity. In T1D, considerable effort has been expended on phocyte infiltration into the thyroid gland, adrenal, linkage studies in affected sib pairs as an approach to kidney and pituitary.10 Two loci, MHC and Cblb, have identify genetic risk factors.2–6 However, most successes been shown to account for the major part of the disease in the identification of susceptibility loci have come from risk in this model.11,12 A nonsense mutation in the Cblb the testing of candidate genes for allelic association with gene of the KDP rat results in the production of a the disease. Two regions, the HLA region on chromo- truncated Cbl-b protein lacking 484 amino acids. some 6p and the insulin gene region on 11p, are generally As a member of the ubiquitin ligase family of proteins, accepted as containing T1D susceptibility loci, desig- Cbl-b participates in the degradation and trafficking of nated IDDM1 and IDDM2, respectively. Both of these proteins. Studies in cell lines derived from CblbÀ/À mice regions were initially investigated based on the presence indicate that Cbl-b is a negative regulator of TCR of viable candidate genes.7,8 Recently, a third suscept- signaling. Activation of naive wild-type CD4 T cells ibility locus, the CTLA4 gene, which may play a more requires stimulation through both CD3 and CD28. Naı¨ve generalized role in autoimmunity, has been proposed CblbÀ/À T cells display comparable levels of proliferation based on its effects on T-cell costimulation, linkage and IL2 secretion when stimulated through CD3 alone. Thus, elimination of Cbl-b alleviates the requirement for CD28 costimulation in T-cell activation, and can rescue the T-cell proliferation defect in cd28À/À mice.13,14 Loss of Correspondence: Dr P Concannon, Molecular Genetics Program, Benaroya the requirement for CD28 costimulation allows activa- Research Institute, 1201 Ninth Avenue, Seattle, WA 98101, USA. E-mail: [email protected] tion of naı¨ve T cells in the periphery under conditions Received 22 October 2003; revised 09 December 2003; accepted 15 where they should be unresponsive, creating the December 2003 potential for autoimmunity. Effects on T-cell costimula- CBLB gene and type 1 diabetes R Kosoy et al 233 tion are also implicated in T1D pathogenesis by reports, within the LD block based on the stringent criteria used, suggesting that CTLA-4 is involved in T1D susceptibility four haplotypes accounted for the majority (90.9%) of all in both human and the NOD mouse.9,10,15 haplotypes observed in the CBLB region. Association analysis, using TDT,20 was performed using data for individual markers, as well as for Results and discussion two-, three-, four- and five-marker haplotypes. None of the individual markers displayed any significant asso- All exons and flanking intronic regions of CBLB were ciation with T1D (Table 2). Three haplotypes, most sequenced in 16 unrelated T1D affected individuals to notably that composed of the ‘C’ allele at CBLB4 and identify variants of possible functional significance and the ‘A’ allele at CBLB1 (35% of 61 transmissions, common SNPs that could serve as haplotype tags for P ¼ 0.017) were significantly undertransmitted to af- association testing. In total, 17 variants were detected by fected offspring. However, none of these effects resequencing, but none were judged likely to have an mained significant after correction for the number of impact on CBLB function or expression (Table 1). tests performed. Examination of the NCBI and Celera SNP databases The Cblb gene is mutated in the KDP rat model of T1D revealed that nine of these polymorphisms were novel. and transgenic rescue experiments confirm the patho- Two rare missense mutations listed in the databases, genic role of this mutation.12 Functional studies of the P522S and R964W, were not detected. product of the Cblb gene in knockout mice and cell lines Seven SNPs were selected from those identified by derived from such mice indicate that Cbl-b plays a key sequencing, and from the NCBI and Celera databases, role in T-cell costimulation.13,14 Defects in this signaling based on their having minor allele frequencies greater pathway that reduce the requirement for costimulation then 0.20 and relatively even spacing across the gene in T-cell activation are associated with autoimmunity. (Figure 1a,b). These SNPs were genotyped in T1D Knockouts of Cblb result in generalized autoimmunity, multiplex families. The number of families genotyped while the hypomorphic allele in the KDP rat, in the for different markers varied from 147 to 259. All seven context of a permissive MHC allele, results in organ- markers had similar heterozygosities, between 0.34 and specific autoimmunity directed primarily against the 0.41, and all were in Hardy–Weinberg equilibrium pancreatic islets, but also against the thyroid gland, (P40.05). Intermarker linkage disequilibrium (LD) ana- adrenal, kidney and pituitary. Recent studies implicate lyses revealed the presence of one large LD block of at polymorphic variation affecting the threshold level of T- least 101 kbp as defined by standardized disequilibrium cell costimulation as playing a role in T1D susceptibility coefficient D0 40.8 (Figure 1c). This LD block included in both human and NOD mice.9,10,15 These considerations most of CBLB, from exon 5–18 and was bordered by the led us to evaluate human CBLB as a possible candidate markers CBLB6 and CBLB2. Within this block, three gene for human T1D susceptibility. haplotypes composed of alleles at five markers, CBLB2/ Since human T1D appears to be much more genetically CBLB10/CBLB4/CBLB11/CBLB6, accounted for 85.8% of complex than the diabetes occurring in the KDP rat, it is all the observed haplotypes: TGCCC (40.4%), GATTT probably not surprising that resequencing in 16 indivi- (34.2%) and GACTT (11.2%). Indeed, even with the duals (32 chromosomes) revealed no obviously inactivat- inclusion of the marker CBLB13, which does not fall ing alleles. While deeper resequencing might reveal rare

Table 1 Variants detected by resequencinga of the CBLB gene

Exon Variantb Minor allele Allele frequencyc Identityd NCBI position (Build 34)

3 IVS3+30insAAA AAA 0.031 106893128 4 IVS4+61G/T G 0.031 106816080 IVS4+64delA delA 0.125 rs3836269 106816077 IVS4+94insA A 0.083 hCV11795476 106816046 IVS4+137insATTG ATTG 0.233 106816003 8 IVS8+60A/G A 0.312 rs2289746 106776856 10 1535C/T T 0.187 rs2305035 106759927 1604A/C C 0.187 rs2305036 106759858 11 1844A/G G 0.219 rs2305037 106743745 12 2126A/G A 0.031 rs3772534 106741935 17 IVS16-76G/A G 0.094 106718393 IVS16-31A/G A 0.062 106718347 18 IVS17-94T/A A 0.031 106710192 2876C/T C 0.187 hCV2821699 106710054 19 3511A/G A 0.179 rs1042852 106698416 3719delA delA 0.031 106698208 3570C/T C 0.031 106698091 aSequencing of all exons and flanking intronic sequences (50–100 nucleotides) was performed on amplified genomic DNA samples from 16 unrelated T1D patients using Big Dye V3.1 (Perkin Elmer Applied Biosystems) and an ABI3100 capillary sequencer. bNucleotide positions based on reference sequence NM170662. cAllele frequency based on the 32 chromosomes. dKnown variant ID numbers are derived from NCBI dbSNP where available (rs# designations), or from Celera RefSNP (hCV# designations).

Genes and Immunity CBLB gene and type 1 diabetes R Kosoy et al 234

Figure 1 CBLB gene and SNP markers used for TDT analyses. (a) Schematic structure of the CBLB gene and the positions of the SNPs selected for analysis. The upper numbers indicate the positions of the nucleotides on chromosome 3 according to NCBI Build 34. Exons are indicated by solid vertical lines with exon numbers indicated above. Arrows indicate the positions of individual SNP markers. (b) Characteristics of SNPs genotyped in T1D families. Genotyping was performed using either primer extension with fluorescence polarization detection, as described16 or single-strand conformation polymorphism (SSCP).17 (c) Linkage disequilibrium relationships between the genotyped SNP markers. Haplotype distributions were estimated using SIMWALK2.18 Standardized disequilibrium coefficients (D0) were calculated using GOLD.19

Table 2 TDT analysesa in T1D families using CBLB SNP markers Despite the failure to detect clear functional variants by resequencing, there could exist common variants that Marker Allele Tb NTc P-value have modest effects on transcription, splicing or stability, located outside of coding regions. Accordingly, we CBLB13 T 116 110 0.69 genotyped SNP markers spanning CBLB and evaluated CBLB6 C 124 125 0.95 them both for the extent of intermarker LD and for CBLB11 T 122 118 0.80 CBLB1 G 125 123 0.90 association to T1D. Much of the gene, including three CBLB4 T 105 103 0.89 of the four functional domains (RING finger, proline-rich CBLB10 A 123 117 0.70 domain and the leucine-zipper/ubiquitin-associated CBLB2 G 89 79 0.44 domain) were contained within a region of strong LD. However, none of the markers tagging common aTDT analyses were performed using GENEHUNTER21 for indivi- haplotypes in this region provided any significant dual markers, as well as for two-, three-, four-, and five-marker evidence of association with T1D. These data provide haplotypes. Only results from individual markers are shown. For little support for a functional role of variation in the each marker, genotyping was performed in 147–259 nuclear families CBLB gene in T1D susceptibility. The first three exons, each with two unaffected parents and two T1D-affected offspring. and possibly the fourth as well, were outside of the Disease onset was less than 30 years of age in all affected strong LD block, but one marker, CBLB13, was posi- individuals. Families were all of Caucasian European ancestry, tioned in the vicinity of the three first exons, while ascertained within the US, and all DNA samples were obtained CBLB6 was positioned close to the fourth exon. Lack of from the Human Biological Data Interchange (HBDI).22 association with alleles at CBLB13, either individually or b Number of alleles transmitted from heterozygous parents to as part of a haplotype, did not suggest a role for N- affected offspring. terminal coding sequences or the promoter of CBLB in c Number of alleles not transmitted from heterozygous parents to T1D predisposition. Knowledge of the haplotype struc- affected offspring. ture in the CBLB gene region and the identities of functional variants in CBLB, their apparent infrequency potential haplotype-tagging SNPs should facilitate the suggests that they are not likely to be major contributors evaluation of the role of this critical gene in other to T1D susceptibility in human populations. autoimmune disorders.

Genes and Immunity CBLB gene and type 1 diabetes R Kosoy et al 235 Acknowledgements 10 Kristiansen OP, Larsen ZM, Pociot F. CTLA-4 in autoimmune diseases—a general susceptibility gene to autoimmunity? We thank Suna Onengut-Gumuscu for assistance with Genes Immun 2000; 1: 170–184. TDT and LD analyses, V Anne Morrison and Lemuel 11 Yokoi N, Kanazawa M, Kitada K et al. A non-MHC Navarro for assistance with nucleotide sequencing, and locus essential for autoimmune type I diabetes in the Mary West for manuscript preparation. This work was Komeda diabetes-prone rat. J Clin Invest 1997; 100: supported by grants from the NIH (DK46635) and JDRF 2015–2021. (1-2001-433) to PC and from the Ministry of Education, 12 Yokoi N, Komeda K, Wang HY et al. Cblb is a major Culture, Sports, Science and Technology of Japan to SS susceptibility gene for rat type 1 diabetes mellitus. Nat Genet 2002; 31: 391–394. and NY. RK was supported by a Genome Training Grant 13 Chiang YJ, Kole HK, Brown K et al. Cbl-b regulates the (NHGRI, T32 HG00035). CD28 dependence of T-cell activation. Nature 2000; 403: 216–220. 14 Bachmaier K, Krawczyk C, Kozieradzki I et al. Negative References regulation of lymphocyte activation and autoimmunity by the molecular adaptor Cbl-b. Nature 2000; 403: 211–216. 1 Rich SS. Mapping genes in diabetes. Genetic epidemiological 15 Nistico L, Buzzetti R, Belgian Diabetes Registry et al. The perspective. Diabetes 1990; 39: 1315–1319. CTLA-4 gene region of chromosome 2q33 is linked to, and 2 Rich SS, Concannon P. Challenges and strategies for investi- associated with, type 1 diabetes. Hum Mol Genet 1996; 5: gating the genetic complexity of common human diseases. 1075–1080. Diabetes 2002; 51: S288–S294. 16 Chen X, Levine L, Kwok PY. Fluorescence polarization 3 Mein CA, Esposito L, Dunn MG et al. A search for type 1 in homogeneous nucleic acid analysis. Genome Res 1999; 9: diabetes susceptibility genes in families from the United 492–498. Kingdom. Nat Genet 1998; 19: 297–300. 17 Orita M, Iwahana H, Kanazawa H, Hayashi K, Sekiya T. 4 Cox NJ, Wapelhorst B, Morrison VA et al. Seven regions of the Detection of polymorphisms of human DNA by gel electro- genome show evidence of linkage to type 1 diabetes in a phoresis as single-strand conformation polymorphisms. Proc consensus analysis of 767 multiplex families. Am J Hum Genet Natl Acad Sci USA 1989; 86: 2766–2770. 2001; 69: 820–830. 18 Sobel E, Lange K. Descent graphs in pedigree analysis: 5 Nerup J, Pociot F. A genomewide scan for type 1-diabetes applications to haplotyping, location scores, and marker- susceptibility in Scandinavian families: identification of new sharing statistics. Am J Hum Genet 1996; 58: 1323–1337. loci with evidence of interactions. Am J Hum Genet 2001; 69: 19 Abecasis GR, Cookson WO. GOLD–graphical overview of 1301–1313. linkage disequilibrium. Bioinformatics 2000; 16: 182–183. 6 Pociot F, McDermott MF. Genetics of type 1 diabetes mellitus. 20 Spielman RS, McGinnis RE, Ewens WJ. Transmission test Genes Immun 2002; 3: 235–249. for linkage disequilibrium: the insulin gene region and 7 Bell GI, Horita S, Karam JH. A polymorphic locus near the insulin-dependent diabetes mellitus. Am J Hum Genet 1993; human insulin gene is associated with insulin-dependent 52: 506–516. diabetes mellitus. Diabetes 1984; 33: 176–183. 21 Kruglyak L, Daly MJ, Reeve-Daly MP, Lander ES. Parametric 8 Nerup J, Platz P, Andersen OO et al. HL-A antigens and and nonparametric linkage analysis: a unified multipoint diabetes mellitus. Lancet 1974; 2: 864–866. approach. Am J Hum Genet 1996; 58: 1347–1363. 9 Ueda H, Howson JM, Esposito L et al. Association of the T-cell 22 Lernmark A, Ducat L, Eisenbarth G et al. Family cell regulatory gene CTLA4 with susceptibility to autoimmune lines available for research. Am J Hum Genet 1990; 47: disease. Nature 2003; 423: 506–511. 1028–1030.

Genes and Immunity