Introduction

Animal models play an important role for the understanding of the pathogenesis as well as the genetics of type 1 diabetes since they provide the opportunity to combine genetic and functional characterisation of the syndrome [1–3]. The LEW.1AR1/Ztm-iddm rat is a model of human type 1 diabetes, which arose through a spontaneous mutation in the intra-MHC recombinant inbred strain LEW.1AR1 in 1997 [4]. This diabetes syndrome is of a proven autoimmune nature [5, 6] leading to apoptotic beta cell destruction, induced by proinflammatory cytokines released from islet infiltrating immune cells [7, 8].

The mode of inheritance is autosomal recessive with an incomplete penetrance of the mutant phenotype of about 60% [7, 8]. In a previous study in the LEW.1AR1-iddm model, three type 1 diabetes susceptibility loci have been discovered by linkage analysis using a Brown Norway (BN) backcross population (BN N2) [9].

Out of these identified three loci, one locus could be mapped to RNO20p12 within the MHC II region that also provides type 1 diabetes susceptibility in humans (IDDM1 [also known as HLA-DQB1]), the NOD mouse (Idd1), and the BioBreeding diabetes-prone (BB) and the Komeda rat models (Iddm1) [10]. Thus, the MHC II haplotype plays a pivotal role in permitting type 1 diabetes development [11].

The other two Iddm loci reside on rat chromosome 1 (RNO1) in the LEW.1AR1-iddm rat [9]. The Iddm8 locus was discovered within RNO1q51–55 at the telomeric end and Iddm9 could be localised in RNO1p11–1q11 near the centromere using the BN N2 backcross population [9]. In an additional backcross population with the Paris rat (PAR) strain (PAR N2) the Iddm1 and Iddm8 loci could be confirmed [12].

The Iddm8 locus could also be associated with a second phenotype in addition to type 1 diabetes development, described as a variable CD3+ T cell frequency in peripheral blood lymphocytes in the LEW.1AR1-iddm rat [13].

The aim of this study was to fine map the Iddm8 region and to identify the mutation that is accompanied by autoimmune diabetes development and a variable CD3+ T cell frequency in the LEW.1AR1-iddm rat.

Methods

Animals

To identify, in the LEW.1AR1/Ztm-iddm rat, a susceptibility locus of the mutation going along with diabetes, a ([LEW.1AR1-iddm × PAR] × LEW.1AR1-iddm) backcross population (PAR N2, diabetes incidence ∼6.5%) and a ([LEW.1AR1-iddm × BN] × LEW.1AR1-iddm) backcross population (BN N2, diabetes incidence ∼16%) were generated as described previously [9, 12]. Diabetic BN and PAR N2 animals were analysed for their genetic characteristics [9, 12]. To prove the impact of the identified mutation, an F2 colony was generated by mating F1 (LEW.1AR1 × LEW.1AR1-iddm) animals. The distribution of the genotypes in a (LEW.1AR1 × LEW.1AR1-iddm) F2 generation according to Mendelian rules is 25% homozygous for the wild-type, 25% homozygous for the mutation and 50% heterozygous. Because the mutation in the LEW.1AR1-iddm rat is autosomal recessive, only animals that are homozygous for the mutation become diabetic.

All animals were bred under specified pathogen free conditions and later housed together in the same hygienic unit in the Central Animal Facility of Hannover Medical School (Ztm) [13]. Blood glucose was measured by the glucose oxidase method (Glucometer Elite, Bayer, Leverkusen, Germany) in blood taken from the tail vein. Diabetic animals were killed within 48 h after onset of hyperglycaemia (≥10 mmol/l). Autoimmune diabetes was confirmed morphologically in each pancreas by diagnosis of an islet immune cell infiltration [7]. The same procedure was applied to non-diabetic animals at the age of 120 days. Experimental procedures were performed according to the German Animal Welfare Act and approved by the Local Institutional Animal Care and Research Advisory Committee of Hannover Medical School and the Lower Saxony State Office for Consumer Protection and Food Safety (Approval ID: 42500/1H).

Tissue collection

Tail and ear biopsies from BN N2 (n = 218), PAR N2 (n = 130) and (LEW.1AR1 × LEW.1AR1-iddm) F2 (n = 110) animals were collected for preparation of genomic DNA. DNA used for genotyping of other strains, as documented in Electronic Supplementary Material (ESM) Table 1, was taken from the existing DNA library of the Central Animal Facility of Hannover Medical School. DNA from BB rats (n = 8) and Komeda rats (n = 8) was isolated from tissue kindly provided by A. Bone (School of Pharmacy and Biomolecular Sciences, University of Brighton, Brighton, UK) and Y. Nakaya (Institute of Health Biosciences, Department of Nutrition and Metabolism, University of Tokushima Graduate School, Tokushima, Japan), respectively. For genotyping the LEW.1AR1-iddm rat, DNA samples were used, which were collected directly after manifestation of the mutation within the background strain LEW.1AR1 (1997) as well as 5, 10 and 15 years after establishment of the inbred strain LEW.1AR1-iddm (n = 20 at each time point). No data were excluded because of lack of quality.

LEW.1AR1 and LEW.1AR1-iddm rats (n = 5–10 of each strain) were killed for organ collection. Organs were frozen until RNA extraction and subsequent gene expression analyses (ESM Table 2). All experiments have been performed in triplicate.

DNA preparation

Genomic DNA was extracted from the tissues using the NucleoSpin Tissue kit (Macherey-Nagel, Düren, Germany) according to the manufacturer’s instructions.

Flow cytometry

From all (LEW.1AR1 × LEW.1AR1-iddm) F2 animals, blood was taken to analyse lymphocyte subpopulations in peripheral blood by flow cytometry as described [13]. Specifically, the natural killer T (NKT) cell population was identified by a double staining with antibodies for CD3 (G4.18) labelled with phycoerythrin (dilution 1:50; Becton Dickinson, Heidelberg, Germany) and for CD161 (NKR 10/78) labelled with fluorescein isothiocyanate (dilution 1:10, AbD Serotec, Munich, Germany). The specificity of the primary labelled antibodies was tested by isotype controls.

Single nucleotide polymorphism analyses

Single nucleotide polymorphisms (SNPs) within the Iddm8 region were selected based on new informative markers as described before [14] (CASCAD SNPview [http://cascad.niob.knaw.nl/snpview/; assessed on 10 August 2015]). All SNPs used were analysed with a specific SNP testing method from KBioscience (KASPar). SNPs were confirmed by sequencing.

Sequencing of candidate genes

For sequence analysis, candidate genes were chosen according to their annotated function. For sequencing, all exons of 49 candidate genes primers were designed using Primer3 software (http://bioinfo.ut.ee/primer3-0.4.0/; assessed on 10 August 2015). Parameters were set to design primers with an optimal melting temperature of 60°C. All sequenced genes are listed in ESM Table 3 (including accession number from GenBank).

Structure analyses

The structure of the dedicator of cytokinesis (DOCK) 8 was prepared with the PyMOL Molecular Graphics System, Version 1.5.0.4 Schrödinger LCC (Munich, Germany) using PDB-code 3VHL [15].

In situ RT-PCR

Sections of the pancreas-draining lymph nodes from the (LEW.1AR1 × LEW.1AR1-iddm) F2 animals and the background strain LEW.1AR1 were placed on 3-chamber slides. The in situ RT-PCR analysis was performed on a specific thermal cycler (MJ Research, Waltham, MN, USA) as described [16]. To identify the mRNA transcripts of the mutated or wild-type Dock8 gene, the primers showed on the 3′ end the specific base exchange as identified as the specific mutation in the Dock8 gene for this animal model. The sequences of the primers are provided in ESM Table 4.

Real-time RT-PCR

RNA was isolated using the RNeasy Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. Tissues were collected from LEW.1AR1 and LEW.1AR1-iddm rats and were homogenised in Isozol lysis reagent (5 Prime, Hamburg, Germany) using the Tissue Lyser LT (Qiagen) according to the manufacturer’s recommendations. RNA was quantified using the NanoDrop ND-100 spectrophotometer (Thermo Scientific, Schwerte, Germany) and the quality was tested in an agarose gel. From 2 μg of RNA, cDNA was synthesised using the Omniscript Reverse Transcription Kit (Qiagen) according to the manufacturer’s recommendations. Sequences of the primers used for real-time RT-PCR are provided in ESM Table 5.

Results

Genetic mapping of the mutation accompanying autoimmune diabetes development in the LEW.1AR1-iddm rat

Iddm8 was mapped on RNO1 (∼275 genes) and is harbouring the mutation that is accompanied by diabetes development and a variable CD3+ T cell frequency in peripheral blood in LEW.1AR1-iddm rats (Fig. 1) [9, 12, 13].

Fig. 1
figure 1

Long range physical map and non-recombination blocks of Iddm8. Iddm8 was identified by linkage analysis with microsatellite markers using the BN N2 and PAR N2 physical positions of microsatellite markers taken from previous studies [9, 12]. The non-recombinant blocks in Iddm8 (red boxes) were determined using SNP markers informative for the BN N2 and PAR N2

Full size image

Genotyping using SNP markers was performed in diabetic BN N2 and PAR N2 rats as well as in non-diabetic rats for both backcross strains. SNPs markers are located within the Iddm8 region in order to narrow down the susceptibility locus. For the BN N2 rats, 35 SNPs could be used as informative markers while for the PAR N2 rats, only 22 SNPs were usable. Two fragments in the diabetic BN N2 and one fragment in the diabetic PAR N2 could be identified within Iddm8 as homozygous for LEW.1AR1-iddm markers (Fig. 1, marked in red). A more detailed analysis of the SNPs is provided in ESM Figs 1–4. All SNPs are listed in ESM Table 6 and were confirmed by sequencing with specific primers as listed in ESM Table 7. For sequence analyses 48 candidate genes were chosen according to their annotated function.

Sequencing of candidate genes

Sequencing of candidate genes (ESM Table 3) and sequence comparison with the background strain LEW.1AR1 as well as the LEW, BN and PAR strains revealed two base exchanges leading to an amino acid exchange within the Iddm8 region, one within the Dock8 gene and one in the Vwa2 gene. In all other sequenced genes, including Cd5, Cd6, Pdcd4, Casp7 and Ins1, no base exchange was detected. All sequenced genes are listed in ESM Table 3.

A mutation from C to G in Dock8 in exon 44 at position 228, 622, 763 (RGSC Genome Assembly v3.4) on RNO1 was identified, which is unique for the LEW.1AR1-iddm rat and not present in any other tested rat strain including the background strains LEW.1AR1, the LEW, BN and PAR (Fig. 2a). Dock8 encodes for a member of the DOCK180 protein superfamily of guanine nucleotide exchange factors (GEFs) that act as activators of Rac/Rho family GTPases [17]. The base exchange leads to an amino acid substitution from glutamine to glutamate (Q1864E) (Fig. 2b). The mutation is located in β4 of the DOCK homology region (DHR)-2 involved in CDC42 binding [18] (Fig. 3a). The mutation may be quite critical for the function of DOCK8, since the negatively charged glutamate could form a salt bridge with the adjacent R1797 located in β1 and interact with S1849, which is part of a flexible loop emerging from the α6 helix (Fig. 3b, red dotted lines). The loop itself contains residues that are involved in CDC42 binding [18]. The changes in number and strength of electrostatic interactions caused by the mutation most likely have an influence on the binding properties of DOCK8 with CDC42 affecting normal function of the protein. Loss-of-function mutations in Dock8 have been reported to cause combined immunodeficiencies [19, 20]. Another 28 different rat strains were sequenced including the two type 1 diabetes rat models, the BB rat and Komeda rat (ESM Table 1), but the Q1864E mutation was found to be unique for the LEW.1AR1-iddm rat. We analysed DNA samples from LEW.1AR1-iddm directly after occurrence of the spontaneous mutation in the background strain LEW.1AR1 as well as at later time points. The [G/G] genotype could be verified in all these LEW.1AR-iddm rats.

Fig. 2
figure 2

Sequencing of the Dock8 gene. (a) Sequencing of the Dock8 gene revealed a base exchange [C/G] in exon 44 at position 228, 622, 763 which is unique for the LEW.1AR1-iddm rat. (b) The base exchange in the Dock8 gene leads to an amino acid exchange from glutamine to glutamate (Q1864E)

Full size image
Fig. 3
figure 3

Structure of the DHR2 domain of DOCK8 from mouse in complex with CDC42. (a) Mutation Q1864E in rat DOCK8 corresponds to Q1865E in mouse DOCK8. The mutated amino acid is shown in green sticks and is located in the β4-sheet of the DHR2 domain. (b) Close-up view of the area surrounding the mutation (orange). The glutamate is supposed to form a salt bridge with the adjacent arginine (R1797) of β1 and additionally interact with residue S1849, which is part of a flexible loop emerging from the α6 helix. This loop is critically involved in interactions with CDC42 [18]. The proposed new interactions (red dotted lines) with surrounding residues might have an impact on the binding properties of DOCK8 with CDC42. The figure was prepared with the PyMOL Molecular Graphics System, Version 1.5.0.4 Schrödinger LCC using the PDB-code 3VHL [15]

Full size image

In a more telomeric region, we confirmed an SNP in Vwa2 in exon 11. At position 263, 216, 228 (RGSC Genome Assembly v3.4) on RNO1 there was a base exchange from C to T in the LEW.1AR1-iddm rat, the LEW.1AR1 rat, the LEW rat and the PAR rat compared with the BN rat strain (Fig. 4a). This SNP changed the amino acid sequence from arginine (R681W) to tryptophan (Fig. 4b). The complete structure of von Willebrand factor A domain containing (VWA)-2 is not known; however, sequence alignments with the A2 domain of VWA2, whose high resolution structure has recently been solved [21], reveal the location of the mutation in a solvent-exposed loop region. We have sequenced additional rat strains and found that 17 out of 28 tested strains have the same SNP genotype (ESM Table 1). Our sequence analyses revealed that the established rat models of type 1 diabetes (LEW.1AR1-iddm, BB, Komeda and LEW.1WR1 rats) share the same SNP genotype in exon 11 of Vwa2 (Table 1).

Fig. 4
figure 4

Sequencing of the Vwa2 gene. (a) Sequencing of the Vwa2 gene revealed a SNP [C/T] in exon 11 at position 263,216,228 between the BN rats and the LEW.1AR1-iddm, LEW.1AR1 and PAR rats. (b) The amino acid exchange is located in the third VWA domain of the protein leading to an amino acid exchange from arginine to tryptophan (R681W)

Full size image
Table 1 Prerequisites for diabetes development in the different type 1 diabetes rat models
Full size table

Segregation of Dock8 alleles within a (LEW.1AR1 × LEW.1AR1-iddm) F2

A (LEW.1AR1 × LEW.1AR1-iddm) F2 (n = 110) was generated and the segregation of the Dock8 [C/G] alleles were analysed. From these 110 animals, 18 animals were homozygous for [C/C], 60 animals were heterozygous for [C/G] and 32 animals were homozygous for [G/G] (Table 2). The distribution of the genotypes is in agreement with the Mendelian rules (χ 2 test, P = 0.1). From these 32 homozygous animals with the genotype [G/G], 20 became diabetic, corresponding to the established diabetes incidence of around 60% in the LEW.1AR1-iddm colony (Table 2) [7].

Table 2 Segregation of the C/G SNP in a (LEW.1AR1 × LEW.1AR1-iddm) F2 generation
Full size table

The mutation in the LEW.1AR1-iddm rat is responsible not only for diabetes development but also for a variable CD3+ T cell frequency in peripheral blood [13]. Therefore, we analysed the CD3+ T cell frequency of all F2 rats. The CD3+ T cell frequency was slightly decreased in rats that were homozygous for the mutation [G/G] compared with animals that were heterozygous [C/G] or homozygous for the wild-type [C/C] (p < 0.001 homozygous [G/G]: 50.2 ± 2.0% vs heterozygous [C/G]: 66.0 ± 1.2% and homozygous [C/C]: 69.3 ± 2.9%). In parallel the CV was increased in animals homozygous for the mutation [G/G] (p < 0.001 homozygous [G/G]: 22.0% vs heterozygous [C/G]: 13.4% and homozygous [C/C]: 12.4%) (Table 2). The NKT frequency was slightly reduced in rats homozygous for the mutation [G/G] compared with the wild-type [C/C]. The CV was not affected (Table 2).

Differentiation of Dock8 expression in immune cells using in situ RT-PCR

Specific primers for the mutated Dock8 allele and the wild-type allele were created to distinguish by in situ RT-PCR analyses the gene expression in the immune cells of the pancreas-draining lymph nodes between the (LEW.1AR1 × LEW.1AR1-iddm) F2 rats and their genetic background strain LEW.1AR1. Expression of the mutated Dock8 mRNA transcript was observed in all rats homozygous for the mutation, while the wild-type mRNA transcript was expressed in all rats homozygous for the wild-type allele (Fig. 5).

Fig. 5
figure 5

Gene expression of Dock8 by in situ RT-PCR. mRNA expression of the mutated and the wild-type Dock8 gene in immune cells of pancreas-draining lymph nodes of the F2 generation from LEW.1AR1-iddm rats and the background strain LEW.1AR1 rats by in situ RT-PCR. Animals homozygous either for the mutation [G/G] or the wild-type [C/C] showed mRNA expression of the Dock8 gene in immune cells using the specific primers. The animals of the background strain revealed the Dock8 gene expression of the wild-type [C/C]. Scale bar, 25 μm. n = 4 animals in each group

Full size image

Quantification of Dock8 expression in different tissues using real-time RT-PCR

The expression of Dock8 was analysed by real-time RT-PCR using gene specific primers located outside the mutation (spanning from exon 14 to 15). The gene expression of different tissues from LEW.1AR1-iddm rats and rats from the background strain LEW.1AR1 was determined. Dock8 gene expression was detectable in all analysed organs, even though gene expression levels were often somewhat lower in the LEW.1AR1-iddm rats compared with the LEW.1AR1 rats, especially in the lymphatic organs (ESM Table 2).

Discussion

In the present study, we identified two base exchanges in Iddm8 on RNO1 segregating with autoimmune diabetes in the LEW.1AR1-iddm rat.

The mutation accompanying autoimmune diabetes development and a variable CD3+ T cell frequency in the LEW.1AR1-iddm rat locates within Iddm8 at the telomeric end of RNO1 from RNO1q41–RNO1q55 [9, 12, 13]. Further analyses using SNPs discovered non-recombinant blocks within Iddm8 of BN N2 and PAR N2, which might associate with the pathological symptoms of the LEW.1AR1-iddm rat. Sequencing of potential candidate genes (ESM Table 3) located within and outside these blocks revealed polymorphisms in the Dock8 gene at the proximal end (RNO1q51) of Iddm8 and the Vwa2 gene at the distal end (RNO1q55) of Iddm8. A base exchange [C/G] in Dock8 is unique for the LEW.1AR1-iddm rat and leads to an amino acid exchange from glutamine to glutamate (Q1865E). A [C/T] SNP in Vwa2 is present in several inbred rat strains (ESM Table 1) leading to an arginine/tryptophan polymorphism (R681W). The fact, that this SNP is distributed among rat inbred strains, excluded the possibility that it is the sole spontaneous mutation in the LEW.1AR1-iddm rat. Further sequence analyses revealed that all diabetes-prone rat strains such as BB and Komeda are homozygous for the [T/T] genotype (ESM Table 1). This, however, supports the assumption that Vwa2 may be involved in the pathogenesis of autoimmune diabetes, as has been reported for humans [22].

The segregation analysis of a (LEW.1AR1× LEW.1AR1-iddm) F2 colony confirmed the base exchange in the Dock8 as the crucial mutation in the LEW.1AR1-iddm rat. Only rats with the [G/G] genotype in Dock8 were affected by a variable CD3+ T cell frequency of which 62% became diabetic. Additionally, the mutation was further proven by in situ RT-PCR analyses of the (LEW.1AR1 × LEW.1AR1-iddm) F2 rats.

Dock8 is well conserved between humans and rodents. Interestingly, there is not only a high conservation in Dock8 between different species but also among members of the whole Dock family [18].

DOCK8 belongs to the DOCK180 family of atypical GEFs. All these DOCK proteins contain two characteristic domains, DHR1 and DHR2 [23]. The DHR1 domain binds to phosphatidylinositol-3,4,5-triphosphate (PtdIns[3,4,5]P3) and mediates the recruitment of DOCK proteins to the plasma membrane [24]. The GEF activity is located in the DHR2 domain, which binds to and activates GTPases of the Rac/Rho family, mainly CDC42 but also Ras-related C3 botulinum toxin substrate 1 (RAC1) [15, 25]. In particular, DOCK proteins regulate actin cytoskeleton, cell adhesion and migration [26]. DOCK proteins also play a role in the immune system. Dock2 deficient mice show T cell lymphopenia, a decreased cellularity of the thymus and secondary lymphatic organs, loss of the marginal zone B cells, a decreased lymphocyte chemotaxis and migration as well as a decreased T cell proliferation [27–30].

The mutation in the LEW.1AR1-iddm rat is located in a β4-sheet within the DHR2 domain of the DOCK8 protein. Therefore, the mutation may have an influence on the binding capacity to GTPases. Binding to the GTPases promotes integrin reorganisation and adhesion, lamellipodia formation, cell polarisation, and phagocytosis or cell fusion [23, 31]. It is known that aberrant activation of Cdc42 results in pathology, such as tumourigenesis and tumour progression, cardiovascular diseases, neuronal degenerative diseases as well as diabetes [32].

DOCK8 has already been known to play a role within the immune system. Mutations in the DOCK8 gene were associated first with the hyper-IgE syndrome [19]. In addition to the high IgE level, patients showed a limited immune response, and are susceptible to recurrent viral and sinopulmonary infections, atopy and allergic diseases. In some cases, different types of skin cancer and lymphomas have been described [19, 20]. Most patients carrying a DOCK8 mutation have a decreased number of CD4+ and CD8+ T cells [20, 33], an observation that runs in parallel to the phenotype of a variable CD3+ T cell frequency with a slight reduction more pronounced in CD4+ T cells than in CD8+ T cells, observed in peripheral blood of LEW.1AR1-iddm rats [13]. In the present study, we confirmed a variable CD3+ T cell frequency in all F2 animals homozygous for the Dock8 mutation. In addition, Dock8-deficient mice are affected by a decreased number of peripheral CD3+ T cells. In these animals, the development of the T cells is normal but mature single-positive CD4+ T cells accumulate in the thymus. The reduced CD4+ and CD8+ T cell numbers in peripheral blood are an indication for an increased turnover and a decreased survival rate [33]. Furthermore, DOCK8 has not only an effect on T cells but also on B cells and NK cells [34–37]. These findings are in agreement with those in the LEW.1AR1-iddm rat [13]. Additionally, a decrease of the NKT cell population was observed in the LEW.1AR1-iddm rat, as had been described before already in Dock8 deficient mice [38]. A defective NKT cell development and a reduced survival rate of T cells in the circulation have been reported also in Gimap5 (GTPase of the immune associated protein 5) knockout mice [39]. Likewise, the Dock8 mutation in the LEW.1AR1-iddm rat may influence development and survival of the NKT cells through changes in the GTPase activity.

Gene expression analyses in the rat showed that Dock8 was highly expressed in lymphatic organs in the background strain LEW.1AR1. The mutation in the Dock8 gene went along with a reduction of the expression level in the LEW.1AR1-iddm rat. Therefore our findings are in agreement with a high gene expression of other DOCK genes, DOCK10 and DOCK11, on human lymphocytes, both T and B cells, as well as in lymphatic organs [40, 41].

The LEW.1AR1-iddm rat [4], the BB rat [3] and the Komeda rat [42] are rat models for type 1 diabetes. What they have in common is the same MHC II haplotype B/D u [9, 42–44], which is a prerequisite for the development of autoimmune diabetes in rats. In addition, we identified in our study in all these rat models the same SNP genotype at position 263, 216, 228 (RGSC Genome Assembly v3.4) within exon 11 of Vwa2 as a factor in the genetic background, which may contribute to disease development.

A distinct mutation in each of the type 1 diabetes rat models is of crucial importance for diabetes susceptibility. The mutation in the Komeda rat is located in the Cblb gene (RNO11), a member of the Cbl/Sli family of ubiquitin-protein ligases [42]. Among a number of other functions, Cblb also regulates T cell activation [45, 46]. The mutation in the BB rat leads to a peripheral lymphopenia and was identified in the Gimap5 (Ian5) gene (RNO4), which encodes a GTPase [47, 48]. The spontaneous mutation in the LEW.1AR1-iddm rat was identified in the Dock8 gene. This mutation is accompanied by autoimmune diabetes development and a variable CD3+ T cell frequency in peripheral blood. It is known that other mutations in the Dock family also lead to an imbalance in the immune system in humans [23, 24]. Therefore, the LEW.1AR1-iddm rat is an interesting model to elucidate the importance of GEFs for the function of the different immune cell types and their relationships to the development of autoimmune diabetes in animal models and in humans.

In conclusion, the identification of the mutation in the Dock8 gene of the LEW.1AR1-iddm rat as well as of the single mutations in other genes in the BB [47, 48] and Komeda [42] rat models of human type 1 diabetes show that single mutations contribute to a polygenic threshold required for disease development. Of particular interest in this context are the similarities between BB and LEW.1AR1-iddm rats; in one case the direct loss of the GTPase function and in the other case the loss of function of the regulatory GEF of the GTPase by the respective mutation.