The JI
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     
 

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Related articles in The JI
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Ivakine, E. A.
Right arrow Articles by Danska, J. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ivakine, E. A.
Right arrow Articles by Danska, J. S.
Right arrowPubmed/NCBI databases
*Gene
*Substance via MeSH
Medline Plus Health Information
*Diabetes Type 1
The Journal of Immunology, 2006, 176: 2976-2990.
Copyright © 2006 by The American Association of Immunologists

Molecular Genetic Analysis of the Idd4 Locus Implicates the IFN Response in Type 1 Diabetes Susceptibility in Nonobese Diabetic Mice1

Evgueni A. Ivakine*,{dagger}, Omid M. Gulban*, Steven M. Mortin-Toth*, Ellen Wankiewicz*, Christopher Scott*, David Spurrell*, Angelo Canty{ddagger} and Jayne S. Danska2,*,{dagger},§

* Program in Developmental Biology, Hospital for Sick Children, Toronto, Ontario, Canada; {dagger} Department of Immunology, University of Toronto, Toronto, Ontario, Canada; {ddagger} Department of Mathematics and Statistics, McMaster University, Hamilton, Ontario, Canada; and § Department of Medical Biophysics, and Institute of Medical Sciences, University of Toronto, Toronto, Ontario, Canada


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
High-resolution mapping and identification of the genes responsible for type 1 diabetes (T1D) has proved difficult because of the multigenic etiology and low penetrance of the disease phenotype in linkage studies. Mouse congenic strains have been useful in refining Idd susceptibility loci in the NOD mouse model and providing a framework for identification of genes underlying complex autoimmune syndromes. Previously, we used NOD and a nonobese diabetes-resistant strain to map the susceptibility to T1D to the Idd4 locus on chromosome 11. Here, we report high-resolution mapping of this locus to 1.4 megabases. The NOD Idd4 locus was fully sequenced, permitting a detailed comparison with C57BL/6 and DBA/2J strains, the progenitors of T1D resistance alleles found in the nonobese diabetes-resistant strain. Gene expression arrays and quantitative real-time PCR were used to prioritize Idd4 candidate genes by comparing macrophages/dendritic cells from congenic strains where allelic variation was confined to the Idd4 interval. The differentially expressed genes either were mapped to Idd4 or were components of the IFN response pathway regulated in trans by Idd4. Reflecting central roles of Idd4 genes in Ag presentation, arachidonic acid metabolism and inflammation, phagocytosis, and lymphocyte trafficking, our combined analyses identified Alox15, Alox12e, Psmb6, Pld2, and Cxcl16 as excellent candidate genes for the effects of the Idd4 locus.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
Type 1 diabetes (T1D)3 is a complex multifactorial disease of autoimmune etiology in which susceptibility is conferred by an interaction between multiple genetic loci and environmental factors. Through genome-wide linkage studies and analysis of congenic strains, >20 insulin-dependent diabetes (Idd) loci have been identified in the diabetes-prone NOD mouse (1, 2, 3). Genome scans and association studies in human populations reveal similarly complex inheritance (4, 5, 6). Despite success in detecting many T1D susceptibility loci, identification of the corresponding genes and elucidation of their immunopathological functions have been complicated by the small effect size of individual loci on the diabetes phenotype. To date, in mice and humans, five T1D susceptibility genes have been characterized at the molecular level. The major susceptibility locus, Idd1 in mice and IDDM1 in humans, is the MHC class II haplotype that controls Ag presentation to CD4+ T cells (1). A functional variant of beta2 microglobulin has been demonstrated to function as a susceptibility allele in the murine Idd13.1 region, suggesting that variants affecting MHC class I function also are important in disease (7). A negative regulator of T cell activation, Ctla4, has been shown to be a strong candidate T1D susceptibility gene for the murine Idd5.1 and human IDDM12 loci (8, 9). The variable-length terminal repeat in the insulin gene promoter is strongly associated with T1D susceptibility in humans, although it was not detected in genetic linkage studies (10, 11, 12). A functional variant of the protein tyrosine phosphatase (PTPN22), involved in the regulation of T cell activation, was recently identified in patients with different autoimmune conditions (13) including T1D (14). The remaining T1D susceptibility genes remain undefined at the DNA sequence level.

To analyze contributions of Idd loci predicted to have strong effects on disease susceptibility, we previously compared T1D-prone NOD mice with the closely related nonobese diabetes-resistant (NOR) strain. NOR is a recombinant inbred strain that is 88% identical by descent to NOD, including the Idd1 locus (H-2g7), but is protected from T1D by genomic intervals of C57BLKS/J (BKs) origin (15). C57BLKS/J is a recombinant inbred strain derived from C57BL/6 and DBA/2 stocks (16). NOD and NOR mice share all Idd loci identified to date, with the exception of Idd4, Idd5, Idd9, and Idd13 (17, 18, 19, 20), which are sufficient to protect NOR animals from both spontaneous and cyclophosphamide-accelerated T1D (CY-T1D).

CY accelerates T1D onset in NOD males and females from months to 2–4 wk, with the cumulative incidence reaching 80–100% (20, 21). Similarly, treated NOR mice demonstrate disease protection (20). We recently showed that differential susceptibility to CY-T1D in NOD and NOR strains resides in the Idd4, Idd5, and Idd9 loci (20).

Here, we report a high-resolution physical map of the Idd4 locus predicated on a series of novel subcongenic strains that refined this interval to a 1.2- to 1.4-megabase (Mb), gene-dense region. This NOD Idd4 locus was sequenced and subjected to detailed genomic comparative analysis with C57BL/6 and DBA/2J strains, the progenitors of the BKs mouse. Gene expression arrays and quantitative real-time PCR were used to prioritize T1D candidate genes by comparing macrophages/dendritic cells (M{phi}/DCs) from congenic strains where allelic variation was confined to the Idd4 interval. The majority of the differentially expressed genes were either located within Idd4 reflecting RNA abundance variations or genes mapped outside the interval that participate in the IFN response pathway, suggesting a role in T1D pathogenesis. Integration of gene expression, sequence, and in silico analysis for all genes in this interval supports five attractive candidates for the Idd4 locus.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 Disclosures
 References
 
Mice

All mice used in this study were maintained in a specific pathogen-free barrier facility at the Hospital for Sick Children. Spontaneous diabetes incidence at age 6 mo in NOD/Jsd animals is 83% in females and 35% in males and 0% in NOR and (NODxNOR) F1 mice. All procedures performed on these mice followed the guidelines of the institutional animal care committee.

Genomic DNA preparation and genotyping

For microsatellite analysis, genomic DNA was prepared from tail snips with a DNeasy kit (Qiagen) and diluted 1/20 for use in PCR amplification. All microsatellite markers used in this study were amplified with the following conditions: 10 cycles (30 s at 94°C, 30 s at 50°C, and 1 min at 72°C), followed by 35 cycles (30 s at 94°C, 30 s at 55°C, and 1 min at 72°C in 1.5 mM MgCl2). The amplification products were electrophoresed through 2% NuSieve (American Bioanalytical) and 2% agarose (Invitrogen Life Technologies) gels and were visualized with ethidium bromide.

Generation of Idd4 subcongenic mice

Previously described NOR.NOD-Idd4 mice (20) were backcrossed with the NOR strain, their pups were intercrossed, and the resulting progeny genotyped for D11Mit74, D11Mit340, D11Mit230, D11Mit135, D11Mit217, D11Mit310, D11Mit164, D11Mit157, D11Mit177, D11Mit4, D11Mit368, D11Mit30, D11Mit90, D11Mit364, D11Mit219, D11Mit322, and D11Bhm149 to capture recombinant animals. NOR.NOD-Idd4 (R1) and NOR.NOD-Idd4 (R2) subcongenic strains were generated by backcrossing corresponding recombinants to the NOR and intercrossing the progeny to fix NOD-derived subcongenic intervals. NOR.NOD-Idd4 (R3) and NOR.NOD-Idd4 (R4) strains were generated by crossing R1 with R2 mice, intercrossing the resulting progeny, and identification of recombinants between D11Mit4 and D11Mit219 markers. Recombinant mice were backcrossed to the R2 strain, and their progeny was intercrossed to fix NOD-derived subcongenic regions.

CY treatment and diabetes assessment of congenic mice

T1D was induced by CY treatment as described previously (20). Blood glucose levels were measured on days 14, 21, 24, 27, and 35 after the first injection using a FastTake blood glucose monitor (LifeScan Canada). Mice with blood glucose levels >16 mmol/L on two subsequent measurements were considered to be diabetic. The difference in diabetes incidence between NOR mice and congenic strains was assessed using Fisher’s exact test in the statistical software package SPSS for PC version 8.01 (SPSS).

Generation and testing of novel markers

To identify novel markers within Idd4, we queried the Idd4 genomic sequence from Celera (<www.celera.org>) or Ensembl (<www.ensembl.org/mouse>) databases for long (n > 15) dinucleotide repeats using a software program that we developed called diNucleotide Tandem Repeat Finder (<http://gchelpdesk.ualberta.ca/servers/tandem_repeat_finder/cgitandem_dtrf.php>). The repeats and the corresponding 300-nt flanking sequences were used for primer design (Primer Express Version 1.5; Applied Biosystems). To identify informative markers, DNA from NOD, NOR, DBA/2J, and C57BL/6 mice was amplified for 35 cycles with the following conditions: 30 s at 94°C, 30 s at 55°C, and 1 min at 72°C in 1.5 mM MgCl2. The amplification products were electrophoresed through either 2% NuSieve (American Bioanalytical) and 2% agarose (Invitrogen Life Technologies) or 8% acrylamide gels and were visualized with ethidium bromide. Novel polymorphic microsatellite markers used in this study are presented in Table I.


View this table:
[in this window]
[in a new window]
  		Table I. Microsatellite markers used to identify Idd4 boundariesa

 
DBA/2J fragment sequences within Idd4

The sequence representing Idd4 (1.4 Mb) was downloaded from Ensembl DB and searched against Celera’s fragments database. The search output was parsed for the occurrence of the term DBA/2J in the identifier line, resulting in 2931 DBA/2J fragments. The DBA/2J fragments were re-BLASTed against the Idd4 locus using standalone BLAST. The best hit for each fragment was parsed and a genomic position assigned to the alignment relative to the C57BL/6J genome. A total of 603 DBA/2J fragments that had complete alignment with Idd4.1 and 973 fragments with 90–99% alignment were selected for analysis. To further filter false positives, the fragments were compared with the Mus musculus genome. Fragments that aligned with Idd4 and returned no other genome hits or fragments that aligned to Idd4 and had a far lower percent identity alignment at another site were selected. These selection criteria resulted in 417 DBA/2J fragments, with an average length of 746 bp. Of these, 136 aligned with 169 exons residing in 29 genes. These outputs represent all definitive DBA/2J exonic sequence available from the Celera database. These fragments were then aligned to identify potential polymorphisms among NOD, C57BL/6, and DBA/2J.

Primer design for real-time PCR gene expression analysis

All genes from the critical Idd4 interval region were queried through the Ensembl database. Primers were designed using Primer Express Version 1.5 with the following requirements: 1) primers were biased toward the 3'end of a gene; 2) one primer overlapped an exon-exon junction for ≥4 nt; 3) the predicted PCR product length was between 100 and 150 bp; and 4) primer melting temperature was 60 or 61°C. For genes with several alternatively spliced transcripts as predicted by Ensembl, we designed primers that recognized all of the different isoforms whenever possible. Alternatively, primers for all predicted individual transcripts were designed.

Tissue preparations for gene expression analysis

To analyze gene expression within the Idd4 region, the following tissues from NOR and NOR.NOD-Idd4 male mice were used: RBC-depleted bone marrow (BM), RBC-depleted spleen, thymus, lymph node (LN), LN cell suspensions cultured in medium for 5 h (no activation), LN cell suspensions cultured in medium with conconavalin A (10 mg/ml) for 5 h (activated), and BM-derived M{phi} either activated with LPS (100 ng/ml) and IFN-{gamma} (10 ng/ml) for 12 h or cultured in medium without activation.

Preparation of splenocytes, thymocytes, and LN cells. Spleen, thymus, and mesenteric LN were aseptically dissected, and single-cell suspensions were prepared in 5 ml of staining medium (1x HBSS (Invitrogen Life Technologies), 2% calf serum (Sigma-Aldrich), filtered through a 0.85-µm Nitex mesh (Sefar America), and collected by centrifugation at 400 x g for 5 min at 4°C. To mimic T cell activation, 107 LN cells were cultured in 10 ml of DMEM (supplemented with 10% FBS, 10 mM HEPES (pH 7.0), 50 nM 2-mercaptoethanol, 2 mM glutamine, 1x nonessential amino acids, and 1% penicillin/streptomycin) in the presence of conconavalin A (10 mg/ml) in T-25 Falcon flasks for 5 h of culture, then washed and collected by centrifugation.

Preparation of BM-derived M{phi}/DCs. Femurs and tibias were removed, flushed with 10 ml of DMEM, clumps dispersed by repeated aspiration, and the suspensions were filtered through a 70-µm nylon cell strainer. A total of 20 x 106 RBC-depleted cells were used for RNA extraction. The remaining cells were adjusted to 106 cells/ml in DMEM containing 10 ng/ml recombinant mouse GM-CSF (R&D Systems). A total of 20–25 x 106 cells were plated in T-75 Falcon flasks and incubated for 7 days, changing to fresh DMEM-10 with GM-CSF every 2 days. Before the medium change, the flasks were washed twice with PBS. On day 7, half of the flasks were treated with LPS (100 ng/ml; Sigma-Aldrich) and IFN-{gamma} (10 ng/ml; R&D Systems) for 12 h, and the remaining cells were incubated in fresh medium without stimulation. Flow cytometric analysis demonstrated that the cells were >95% CD11b+ and 10–20% CD11c+ at the conclusion of the cultures (data not shown).

RNA extraction and cDNA synthesis

RNA was isolated from mouse tissues and ex vivo-cultured cells using TRIzol (Invitrogen Life Technologies) according to the manufacturer’s instructions. cDNA was synthesized using Omniscript reverse transcriptase (Qiagen) according to the manufacturer’s instructions. cDNA samples were diluted 1/25 for use in a quantitative real-time PCR.

Analysis of gene expression by real-time PCR

Expression of genes within the Idd4 was analyzed under the following conditions: denaturation at 95°C for 15 s, followed by annealing/extension at 59°C for 1 min for 40 cycles. Real-time PCR was performed in 20-µl reactions containing a 4-µl volume of diluted cDNA, 0.2 µM primers (see Table II), and 1x SYBR Green buffer (Applied Biosystems) using the ABI/PRIZM 7900 HT cycler (Applied Biosystems). For each gene, serial dilutions (102–107) of a plasmid containing the gene sequence were used as a reference for the standard curve calculation. The fluorescence thresholds and the corresponding cDNA copy numbers were calculated using SDS 2.0 software (Applied Biosystems). For each gene, the reactions were performed in triplicate, and the quantity of cDNA was normalized to the cDNA amounts of both beta-actin and Hprt. Idd4-region gene sequences were obtained from the Ensembl database. The National Institute on Aging and RIKEN cDNA clones, corresponding to the genes, were identified by performing basic local alignment sequence tool (BLAST) analyses of cDNA libraries available at (<http://lgsun.grc.nia.nih.gov/cdna>) and (<http://genome.gsc.riken.go.jp>), respectively. Available clones were obtained from the Center for Applied Genomics at the Hospital for Sick Children. When plasmids were unavailable, the corresponding PCR product was purified and cloned with a PCR cloning kit (Qiagen) according to the manufacturer’s instructions.


View this table:
[in this window]
[in a new window]
  		Table II. Idd4 and IFN response genes analyzed by real-time PCRa

 
Sequencing of candidate genes

The coding regions of five genes (Psmb6, Alox15, Cxcl16, Pld2, and C1qbp) with known functions in immunological responses were compared by direct sequencing from NOD and NOR strains. cDNA prepared from activated BM-DC/M{phi} was used as a PCR template under the following conditions: 30 s at 94°C, 30 s at 65°C, and 1 min at 72°C in 1.5 mM MgSO4 for 30 cycles using High-Fidelity Taq polymerase (Invitrogen Life Technologies). The 5' regulatory (5'reg) and the untranslated (UTR) regions of the Psmb6, Alox15, and Pld2 genes also were amplified and sequenced in NOD and NOR strains, using genomic DNA as a template and the same PCR conditions. Sequencing primers were designed using Primer Express 1.5 software (Applied Biosystems) and are listed in Table III. PCR products were purified per the manufacturer’s instructions (Qiagen) and sequenced on both strands at the Center for Applied Genomics. Sequence files for NOD and NOR strains were then aligned and compared using Lasergene software (DNASTAR).


View this table:
[in this window]
[in a new window]
  		Table III. Sequencing primers used in our studya

 
Analysis of differential gene expression using microarrays

RNA from resting NOR.NOD-Idd4 (R3) (n = 3) and NOR.NOD-Idd4(R4) (n = 3) BM-derived DC/M{phi} was extracted as described above, with the additional purification step (RNeasy kit; Qiagen) according to the manufacturer’s instructions. Microarray target preparation was performed using standardized protocols as suggested by Affymetrix (<http://tcag.bioinfo.sickkids.on.ca/microarray.html>). RNA was reverse transcribed to generate dscDNA that was then used to synthesize biotin-labeled cRNA using in vitro transcription. Fragmented cRNA was hybridized to Affymetrix MGU74Av2 arrays and scanned using a confocal scanner (Agilent). Expression values for each probe set were calculated using Affymetrix Microarray Suite 5.0 software.

Statistical analysis. Quantile normalization (22) in an R package (Bioconductor; (<www.bioconductor.org>) was used to remove nonbiological variation across arrays. Robust Multiarray Average (23) was used to calculate an expression value for every probe set. To determine which genes were differentially expressed between R3 and R4 samples, a significance analysis of microarrays (SAM) (24) was performed. SAM assigns a score to each probe set based on the change in gene expression relative to the SD of the repeated measurements for the probe set and uses a significance threshold "{delta}" to control the proportion of falsely identified genes at a desired level. The threshold {delta} can be adjusted to identify sets of genes, and the false detection rate (FDR) for each set is estimated from the data.

The relative difference (d(g)) in gene expression is computed from Equation 1:

Formula 1(1)
, where the X-bars indicate the mean expressions for the g-th gene in R3 and R4 samples, S(g) indicates the gene-specific scatter (based on the SDs of repeated measurements for the gene in the R3 and R4 samples), and S0 is a constant added to stabilize the variance so that all data can be used in the analysis. SAM was used to calculate the observed relative difference for each gene in each paired comparison. SAM estimated the expected relative difference under the "null" assumption that no genes are differentially expressed. The null distribution of the relative differences d(g) was computed from multiple permutations of the data set. For each permutation, the relative differences were ranked relative to the differences averaged across all of the permutations. The observed relative differences were ranked by magnitude and plotted against the averaged relative differences from the permutations to generate a SAM plot. The expected number of false positives was found by considering each of the permutations and counting the number of genes that would be declared significant and averaging these numbers. The FDR is estimated as the expected number of false positives divided by the number of probe sets declared significant for the original data.

Sequencing of the Idd4 region in the NOD mouse

Based on our map of the Idd4 locus, this region of the NOD genome was selected for genomic sequencing at the Wellcome-Sanger Institute (Cambridge, UK) funded by a NOD genomic sequencing subcontract from the National Institutes of Health (Bethesda, MD). Ten NOD bacterial artificial chromosome (BAC) clones spanning the interval between D11Gul2537 and D11Gul2721 markers were completely sequenced, and the data posted for free access at the institute’s web site. Additional information regarding this project can be found at (<www.sanger.ac.uk/cgi-bin/projects/m_musculus/mouse_nod_clones_tpf>).


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 Disclosures
 References
 
CY-T1D in NOR.NOD-Idd4 subcongenic male mice

Previously, we demonstrated that NOR.NOD-Idd4 mice are susceptible to CY-T1D (20). To refine the location of the genes responsible for this effect, four novel recombinant subcongenic mice carrying NOD-derived Idd4 intervals were generated and assessed for T1D following CY treatment (Fig. 1). Male mice of the R1 strain were susceptible to CY-T1D, compared with NOR mice (p = 0.0006), whereas R2 mice were resistant. Double-recombinant R3 animals progressed to T1D after CY treatment in contrast to R4 mice that were resistant to the disease (p = 0.003). Importantly, the R3 and R4 strains differed only for a 2.4-Mb interval between D11Mit30 and D11Mit364, an interval sufficient to control the differential CY-T1D susceptibility phenotype.


Figure 1
View larger version (34K):
[in this window]
[in a new window]
  		FIGURE 1. Fine mapping of the Idd4 locus. Different NOR.NOD-Idd4 subcongenic strains were generated, and male mice were tested for CY-T1D as described in Materials and Methods. The strains were genotyped for multiple public (D11Mit) and novel (D11Gul) markers to determine locations of NOD- (gray) and NOR-derived (black) regions. Unresolved boundaries between NOD- and NOR-derived intervals are depicted in shaded gray ovals, and regions identical by descent between parental strains are shown in white. Percentage of diabetic males for each strain as well as a Fisher’s exact test p values vs parental NOR strain (11%, n = 36) as a measure of resistance or susceptibility to CY-T1D are illustrated. Susceptibility (S) or resistance (R) to CY-T1D, relative to the parental NOR strain, also is indicated.

 
Refinement of the Idd4-region boundaries using novel markers

We analyzed mouse genomic sequence information in the Celera and Ensembl databases across the Idd4 interval for microsatellite repeats using a Perl script and selected flanking sequences for the design of amplification primers (see Materials and Methods). PCR amplicons containing the repeats were tested for polymorphism between NOD and NOR strains, and informative markers (Table I) were used to generate a high-resolution map of the recombination boundaries in the subcongenic strains. The centromeric boundary of the Idd4 critical interval was resolved to a 12-kb region between the D11Gul2535 and D11Gul2537 markers (Fig. 1). The telomeric boundary was placed in a 200-kb region between the D11Gul2700 and D11Gul2721 markers. Neither boundary region contained known or predicted genes, according to Ensembl. Thus, Idd4 was confined to a 1.2- to 1.4-Mb region containing 52 genes (Table IV). Attempts to further refine location of the Idd4 region by analyzing 1031 additional meioses for recombination between D11Gul2537 and D11Gul2700 were not successful, representing a genetic distance of <0.3 centiMorgans (cM)/Mb between these two strains.


View this table:
[in this window]
[in a new window]
  		Table IV. Genes within the Idd4 locusa

 
Genomic sequence analysis of Idd4

Based on our mapping data, the NOD of the Idd4 locus was sequenced at the Wellcome-Sanger Institute. A tile path of 10 NOD BAC clones spanning the interval was completely sequenced and the data posted (<www.sanger.ac.uk/cgi-bin/projects/m_musculus/mouse_nod_clones_tpf>). We performed a comparative sequence analysis of the coding exonic regions for all 52 genes across the interval and noticed striking similarities between NOD and B6 strains: 47 genes showed no variation, 4 showed only silent nucleotide substitutions, and only 1, Kif1c, showed a nonsynonomous single nucleotide polymorphism (SNP) affecting amino acid sequence. Previously, we demonstrated that Idd4 is located within the DBA/2J-derived region of the NOR chromosome 11 (20). To examine sequence variation between NOD and DBA/2J, we used Celera DB to retrieve all available DBA/2J sequence fragments as described in Materials and Methods. The mined DBA/2J sequence fragment files covered only 30% of the Idd4 coding sequence. Our analysis of these data showed that, in contrast to the similarity between NOD and B6 strains in this region, DBA/2J sequence is highly divergent from both NOD and B6 with many synonymous and missense substitutions (Table V). This analysis suggested that the Idd4 locus mapped in this and our prior study (20) influenced T1D susceptibility in crosses between NOD and NOR, because the latter is DBA/2J- rather than C57BL/6-derived over this interval. These analyses reinforce the need for the high quality genomic sequence and haplotype map data for multiple mouse strains to facilitate identification of susceptibility genes underlying complex disease phenotypes, including T1D.


View this table:
[in this window]
[in a new window]
  		Table V. Summary of exonic polymorphisms among NOD, C57BL/6, and DBA/2J strains in coding regions of several ldd4 genesa

 
Comparative analysis of M{phi}/DC gene expression in R3 and R4 mice with Affymetrix arrays

We choose to examine Idd4 genotype-dependent differential genes expression in M{phi}/DCs for several reasons. First, these cells are crucial in pathogenesis of T1D, because their in vivo depletion prevents the disease (25). Second, T cells from NOR mice transfer diabetes to NOD.scid recipients (26), implicating differences in other immune cell types as a source of diabetes resistance in NOR mice. The critical role for M{phi}/DC nexus between innate immune sensing and adaptive Ag-specific immunity focused our interest on these cells. Importantly, B cells also contribute to Ag-presentation in T1D (27, 28), although they are not absolutely required for the disease in the NOD model (29) or in humans (30), sharpening our focus on critical functions of M{phi}/DC.

Affymetrix Gene Chip 430 2.0 arrays were used to profile gene expression variations between the T1D-susceptible R3 and T1D-resistant R4 strains. This comparison restricted the heritable variation between samples to the 1.4-Mb Idd4 interval. M{phi}/DCs were grown from BM precursors for 7 days in GM-CSF supplemented medium. At the conclusion of the culture, the cells were >90% CD11b+, MHC class II low, and 15–25% CD11c+ (data not shown) suggesting a mixture of monocyte/M{phi} and immature DCs. SAM identified only 12 probe sets that were differentially expressed between R3 and R4 M{phi}/DCs with a FDR ranging from 0.05 for Psmb6 to 0.55 for Usp18 (Table VI). Gene annotation was performed, and the data were displayed with probe set identification, SAM rank, fold change, gene symbol, chromosome position, and function (Table VI). Notably, four of the most highly ranked genes were located within the Idd4 locus, suggesting that these expression differences were mediated by cis-acting elements within the region. Among these, differential expression of Psmb6 and Alox15 between the R3 and R4 congenic lines were of special interest because of their roles in immune responses. Psmb6 encodes a catalytic subunit of the proteasome involved in Ag processing, and Alox15 encodes a lipo-oxygenase operative in biogenesis of leukotrienes, potent inflammatory mediators. SAM also revealed genes that were identical by descent between the T1D-susceptible R3 and T1D-resistant R4 strains that were affected in trans by allelic differences in the Idd4 region. All of these genes were up-regulated in T1D-prone R3, compared with disease-resistant R4 strain, and five of them are functionally linked to the "IFN response" pathway. Because of the role of IFN in T1D pathogenesis (reviewed in Ref.31), these data suggest that the Idd4 locus control diseases through regulating IFN-responsive genes. Differential expression of IFN-response genes between R3 and R4 strains was particularly intriguing because the comparisons were performed on M{phi}/DCs grown in parallel from myeloid precursors without experimental activation. These results suggested that inherent differences in M{phi}/DC IFN response state result from Idd4 variation between T1D-prone R3 and T1D-resistant R4 strains. To examine the behavior of these IFN-response genes under activated conditions, we stimulated BM-derived M{phi}/DCs as well as freshly explanted peritoneal M{phi} from R3 and R4 strains with LPS and IFN-{gamma} and used quantitative real-time PCR to analyze Oas12, IF28, Irf3, and Usp18 (Fig. 2). The microarray outcome in resting BM-derived M{phi}/DCs was validated for all four genes and showed the Idd4 genotype effect in both male and female mice (Fig. 2a). LPS and IFN-{gamma} treatment of BM-derived M{phi}/DCs and peritoneal M{phi} exacerbated the strain variation in expression Oas12, Irf3, and Usp18, providing additional evidence that Idd4 exerts its effect through regulation of the IFN response (Fig. 2, b and c).


View this table:
[in this window]
[in a new window]
  		Table VI. Differentially expressed genes in R3 vs R4 macrophages/DCsa

 

Figure 2
View larger version (39K):
[in this window]
[in a new window]
  		FIGURE 2. Expression analysis of Oasl2, If28, Irf3, and Usp18 in resting and activated M{phi}/DCs. Expression of Oasl2, If28, Irf3, and Usp18 genes was analyzed in BM-derived (a and b) and peritoneal M{phi} (c) from NOR.NOD-Idd4 (R3) and NOR.NOD-Idd4 (R4) strains by quantitative RT-PCR as described in Materials and Methods. Cells were either resting (a) or activated (b and c) with LPS and IFN-{gamma}. Samples from individual male (M) and female (F) mice are shown. For each gene, absolute copy numbers of the corresponding cDNA was calculated from a standard curve generated by amplifying serial dilutions of a plasmid containing the gene of interest. PCR was run in triplicate to estimate mean copy number as well as a SD for each gene. Expression of every gene was normalized to the expression of the actin gene from the same cDNA sample. Normalized data for the mean and the SD were multiplied by 1000 for better visual representation. Note that different y-axis scales were required to display three distinct tissues and conditions. Similar results were obtained from two independent experiments.

 
Expression analysis of all Idd4-region genes

SAM identified four genes mapping within the Idd4 interval that displayed differential expression (Table VI). However, given the limited dynamic range of the Affymetrix microarray platform, selection of a conservative statistical analysis of these data sets, lack of representation of all Idd4 genes on the array, and the desirability of independent validation of the SAM results, we performed real-time PCR expression analysis of all 52 genes from the Idd4 region in multiple immune cell types and conditions.

Expression of 26 genes was studied in freshly explanted BM, spleen, thymus, and LN and in BM-derived M{phi}/DCs from NOR and NOR.NOD-Idd4 mice under alternate treatment conditions (Fig. 3). Four of these 26 genes (Cldn7, Slc16a13, 4930563C04Rik, and Chrne) were expressed at low to undetectable levels in all cells/conditions tested and were assigned low priority as candidate genes. A gene was called "differentially expressed" if it displayed a steady-state mRNA level of ≥1 mRNA transcript per 1000 transcripts of actin and a >2.5-fold difference between Idd4 genotype-disparate strains. By these criteria, we confirmed differential expression of Psmb6 in BM-M{phi}/DCs, validating the microarray analysis. Interestingly, this Idd4-dependent difference was specific for M{phi}/DCs, because differential expression was not observed in heterogeneous tissues, such as whole spleen or BM that contain low frequencies of M{phi}/DCs, and not observed in lymphoid-rich thymus and LN samples (Fig. 3). Similarly, Alox12e, Pld2, and Camta2 displayed differential expression in enriched M{phi}/DCs that was undetectable in whole spleen or BM. Alox12e and Pld2, expressed at equal levels in NOR and NOR.NOD-Idd4 LN and M{phi}/DC, were attractive T1D susceptibility gene candidates because of their functions in arachidonic acid metabolism (32) and phagocytosis (33), respectively. Camta2 is a poorly annotated gene encoding a calmodulin binding transcription activator 2. A poorly annotated gene of unknown function, 6330403K07, was overexpressed in the LN, spleen, and thymus of the NOR, compared with the NOR.NOD-Idd4 strain. Two genes with immunological functions: the chemokine Cxcl16 and complement-binding protein C1qbp, were expressed at similar levels in both strains in this analysis.


Figure 3
View larger version (45K):
[in this window]
[in a new window]
  		FIGURE 3. Tissue distribution and comparative analysis of the Idd4 genes expression. Expression of 26 genes located in the Idd4 locus was analyzed in BM, thymus, spleen, LN, resting and activated lymphocytes, as well as resting and activated M{phi}/DCs from NOR and NOR.NOD-Idd4 males using quantitative real-time PCR. For each gene, absolute copy numbers of the corresponding cDNA was calculated from a standard curve generated by amplifying serial dilutions of a plasmid containing the gene of interest. For gene 4930563C04Rik, both predicted transcripts, T19065 (T1) and T19066 (T2), were evaluated. PCR was run in triplicate to estimate mean copy number as well as a SD for each gene. Expression of every gene was normalized to the expression of the actin gene from the same tissue. Normalized data for the mean and the SD were multiplied by 1000 for better visual representation. Note that different y-axis scales were required to display expression of different genes.

 
Because differential expression of four genes was best observed in enriched M{phi}/DC, we prioritized comparisons of the remaining 26 Idd4 genes in this cell population (Fig. 4). Two genes, Alox15 and Map4k6 displayed differential expression in this analysis. In agreement with the microarray analysis (Table VI), we found that Alox15 was expressed at higher levels in NOR, compared with NOR.NOD-Idd4 M{phi}/DCs. Map4k6 operative in MAPK signaling, was overexpressed in NOR.NOD-Idd4, compared with NOR samples. With this locus-comprehensive strategy, we assigned highest priority to nine genes for sequence and functional analysis: seven were differentially expressed genes between the R3 and R4 strains, and Cxcl16 and C1qbp, which were expressed at similarly high levels in both strains and were deemed good functional candidates for conveying T1D-associated effects of the Idd4 locus.


Figure 4
View larger version (26K):
[in this window]
[in a new window]
  		FIGURE 4. Analysis of gene expression in BM-derived M{phi}/DCs. Expression analysis of the 26 genes from the Idd4 region, which were not presented in Fig. 3, was performed in resting and activated BM-derived M{phi}/DCs pooled from three individual NOR and NOR.NOD-Idd4 male mice by quantitative RT-PCR as described in Materials and Methods. Resting M{phi}/DCs (1) and M{phi}/DCs activated with LPS and IFN-{gamma} (2) are shown. For every tissue, gene expression was normalized to the expression of the housekeeping gene actin. Normalized data for the mean and the SD were multiplied by 1000 for better visual representation. Note that genes with very low levels of expression in both strains and under both conditions are not presented on this graph.

 
We considered the possibility that expression analysis of some Idd4.1 genes could have been affected by differential annealing of the primers or probe sets from Affymetrix microarray to NOD and NOR sequences. The primers used for these comparative gene expression analyses were designed from the available C57BL/6J sequence that we show is extremely similar to the NOD sequence over this interval. Genomic sequence information for the NOR strain spanning the Idd4.1 region (DBA/2J origin) is currently unavailable, precluding selection of primers known to be identical between NOD and NOR strains. To mitigate this risk, the primer pairs we used were selected for similar high amplification efficiencies in both strains. In addition, we found evidence for tissue/cell type-specific differential expression as a function of Idd4.1 genotype (Figs. 3 and 4) that argues against an influence of sequence variation within the primer binding sites that would affect amplification outcomes in all samples.

Comparative sequence analysis of the candidate T1D susceptibility genes

Five of the nine priority genes were selected for comparative sequence analysis: Alox15, Cxcl16, Psmb6, Pld2, and C1qbp. For the differentially expressed Psmb6, Alox15, Pld2 we sequenced the exons, 5'UTR, 3'UTR, and 5'regulatory (~1000 bp upstream) regions from NOR DNA and compared the sequence with NOD and C57BL/6 strains from the Wellcome-Sanger Institute and Ensembl web sites. Differences in the 5' regulatory region may influence transcriptional activity, and changes in the UTR can affect RNA stability; both events leading to the different steady state level of mRNA between NOD and NOR strains.

Psmb6. Seven silent SNPs distinguishing NOD and NOR were observed in the coding region (Table VII), predicting identical PSMB6 protein sequence between the two strains. Comparative analysis of 1000 bp upstream of the ATG start codon, the 5'UTR, and 3'UTR regions also was performed. Multiple SNPs and several small deletions distinguished the Psmb6 5' regulatory region in NOD and NOR (Table VII). With exceptions at positions –190 and –319, NOD and B6 sequences were identical, consistent with their equal expression patterns in M{phi}/DCs (data not shown). Six SNPs distinguished the NOD and NOR 3'UTR and again, NOD and B6 sequences were identical at these positions (Table VII). Because of the lack of experimental evidence for the location of the Psmb6 promoter, we used computational tools to search 1000 bp upstream of the ATG start codon for potential transcription factor binding sites. Many transcription factor binding sites were predicted using Match (34) and Signal Scan (35) software, but none were well conserved in human PSMB6. The Pipmaker computational algorithm for phylogenetic conservation (36) was run for the regions 10 kb upstream of the Psmb6 gene using human and mouse Ensembl sequence. This analysis revealed phylogenetically conserved elements within a short 130-bp stretch immediately upstream of the transcription initiation site (data not shown). However, direct sequence analysis demonstrated that this sequence was identical between NOD and NOR mice (Table VII), suggesting that more distant regulatory regions are responsible for differential expression of Psmb6 between these strains.


View this table:
[in this window]
[in a new window]
  		Table VII. Polymorphisms among B6, NOD, and NOR in Pld2, 1qbp, Alox15, Cxcl16, and Psmb6 genesa

 

View this table:
[in this window]
[in a new window]
  		Table VIIA. (Table continued)

 
Alox15. Eight coding SNPs distinguishing NOD and NOR were observed. Five were synonymous SNPs and three resulted in amino acid substitutions: S210N, L335S, P616S. At least one of these substitutions (S210N) is located within a region conserved among mammalian lipo-oxygenases, suggesting that, in addition to differences in the mRNA abundance, ALOX15 enzymatic activity also may differ between NOD and NOR. Seven SNPs distinguishing NOD and NOR were observed in the 5' regulatory region where NOD was identical with B6 with the exception of a repetitive element at the position –344. A 46-nt insertion was identified in the 3'UTR of the NOR, compared with NOD Alox15, and could contribute to the much greater abundance of Alox15 mRNA in this strain (Fig. 4).

Pld2. Six coding SNPs distinguished Pld2 in NOD and NOR mice. One SNP resulted in H715Y, with potential effect on PLD2 enzymatic activity because it is located within the catalytic domain of the protein. Numerous changes, including SNP deletions and insertions, were identified in the NOR Pld2 5' regulatory and the 3'UTR, compared with both NOD and B6 strains that were identical in these regions. Additional studies are required to identify sequence variations responsible for the differential expression of Pld2 in NOD and NOR LN and M{phi}/DCs.

Cxcl16 and C1qbp. In addition to the seven differentially expressed genes within Idd4, Cxcl16 and C1qbp, which were expressed similarly in NOR and NOR.NOD-Idd4 mice, were considered good T1D susceptibility candidates because of their known involvement in immune response. CXCL16, expressed at high levels on M{phi}/DCs, induces a strong chemotactic attraction of activated CD8 cells (37) and a subset of NK-T cells (38) and mediates adhesion and phagocytosis of Gram-negative and -positive bacteria (39, 40). C1QBP is a ubiquitously expressed, multiligand binding protein involved the classical complement pathway. The coding regions of both genes were sequenced in NOD and NOR strains, yielding three synonomous SNPs (Table VII). When considered together with similarity in gene expression pattern between NOD and NOR, these results reduce interest in C1qbp as a candidate gene for Idd4. In contrast, five SNPs distinguished NOD and NOR Cxcl16 coding sequence (Table VII). A S129P substitution was observed between with predicted extracellular localization and a potential functional impact on this protein, making Cxcl16 an attractive candidate for Idd4.


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
Disclosures
 References
 
Previously, we identified the Idd4, Idd5, and Idd9 loci as crucial regulators of susceptibility to CY-accelerated and spontaneous T1D in NOD mice (20). Here, we present a high-resolution molecular genetic analysis of the Idd4 locus to derive criteria for prioritizing candidate T1D susceptibility gene(s) within this gene-rich region. Using a series of the NOR.NOD-Idd4 subcongenic strains, we refined location of the locus to a <1.4-Mb interval containing 52 genes. Complete NOD genomic sequence analysis of the region revealed a striking similarity between NOD and C57BL/6 strains and striking differences with DBA/2J, suggesting that NOR Idd4-dependent T1D resistance is mediated by alleles of DBA/2J origin that were fixed in the C57BLKS/J (BKs) progenitor of NOR. Recently, comparative SNP analysis in C57BL/6J, DBA/2J, and BKs was performed to define genomic boundaries and the genetic origins of the BKs strain. This study showed that BKs SNPs in the portion of chromosome 11 containing our Idd4.1 interval were identical with DBA/2J (41). These data are consistent with our evidence that coding region sequence in this region of NOR/Lt mice is of DBA/2J origin. In a search for Idd4 candidate genes and biological pathways associated with T1D pathogenesis, we analyzed expression of all genes in M{phi}/DCs from T1D-susceptible and T1D-resistant subcongenic strains distinguished only by the <1.4-Mb Idd4 region. Strikingly, we observed heightened expression of genes characteristic of an IFN signature in the T1D-susceptible R3, compared with T1D-resistant R4 strain. These data implicate an inherent difference in threshold for triggering the IFN pathway in NOD diabetes pathogenesis. We examined the mRNA abundance of the Idd4 genes in multiple immune cell populations and identified attractive functional candidates displaying tissue-specific differential expression. Direct sequencing of several positional candidates from NOD and NOR strains revealed sequence variation in the coding, 5' regulatory, and UTR regions of these genes. Based on our results, we have refined to five the best candidate genes for the Idd4 locus.

Idd4 was first identified as a 30-cM region in a genome-wide linkage study of (NODxB10H-2g7) x NOD animals and linked to early onset T1D (43). Subsequent analysis of NOD.B6 congenic animals refined Idd4, with the strongest T1D resistance mapping to 9.2 cM (~20 Mb) between the D11Mit30 and D11Mit41 (43). Recently, we used NOD.NOR-Idd4 mice to further narrow down the Idd4 locus to a 6.9-cM (~12-Mb) interval between D11Mit30 and D11Mit33, sufficient to protect NOD mice from both spontaneous and CY-accelerated diabetes (20). Grattan et al. (43) suggested two subloci within Idd4 (Idd4.1 and Idd4.2), but the genetic map of recombination boundaries in these NOD.B6-Idd4 mice provided limited resolution. The 1.2- to 1.4-Mb interval described here lies within this previously predicted Idd4.1. Genomic sequence analysis of the Idd4 locus presented here showed striking similarity between NOD and B6 strains and, coupled with in silico analysis of Celera data, provided strong evidence for the DBA/2J origin of NOR alleles providing protection from CY-T1D in our study. However, these analyses cannot rule out the existence of sequence differences between NOD and both DBA/2J and C57BL/6J, particularly at noncoding regulatory sites of the Idd4.1 genes. Full genomic sequence of Idd4.1 from NOR/Lt (BKs) will be required to definitively determine whether the same or different genetic variants confer T1D protection in NOR/Lt (this study) and B6 (43).

Gene expression microarray comparison of T1D-susceptible R3 and T1D-resistant R4 congenic strains identified 11 differentially expressed genes within in resting BM-derived M{phi}/DC samples. Four of these genes were located within Idd4, reflecting differences in proximal regulatory elements. These results accord with the demonstration that transcript abundance can serve as a surrogate phenotype for quantitative traits (44). The authors used differences in gene expression in maize, mouse, and human as quantitative traits to perform classical mapping studies and found that, in many cases, the observed expression differences mapped to genomic regions containing the corresponding genes. Among the differentially expressed genes in our analysis, Psmb6 and Alox15 are good candidate T1D susceptibility genes.

PSMB6, also known as proteasome subunit {delta} or proteasome subunit Y, is one of three constitutively expressed catalytic subunits of the 20S proteasome (45). Upon IFN-{gamma} treatment, the Psmb6 subunit is replaced with its inducible partner, Lmp2, forming "immunoproteasomes" (46, 47) that are believed to favor the generation of a "skewed" pool of peptides for MHC class I molecules, leading to an enhanced immune response against viral infections (reviewed in Ref.48). Contrary to other strains (46, 47), we found that expression of Psmb6 in the NOD M{phi} increases upon activation (Fig. 3), suggesting that exchange of Lmp-2 for Psmb6 in proteosomes may be strain dependent. Because Psmb6 is directly involved in the generation of peptides presented to CD8+ T cells, inefficient replacement by Lmp2 in activated APCs may favor production of distinct diabetogenic peptides contributing to NOD T1D susceptibility.

ALOX15 belongs to a family of lipid-peroxidizing enzymes responsible for the production of lipoxins from arachidonic or linoleic acid substrates (49). These substrates are elevated during inflammation (50), and multiple lines of evidence suggest that they exhibit anti-inflammatory activities (32, 51, 52). Moreover, reduced inflammation and tissue damage in a model of acute periodontitis was reported in transgenic rabbits overexpressing Alox15 (53). The greater expression of Alox15 in NOR M{phi}/DCs may, therefore, protect from T1D by production of anti-inflammatory lipoxins.

Microarray analysis and real-time PCR also identified genes that are identical by descent between R3 and R4 congenic mice and differentially regulated in trans by gene(s) within the Idd4 locus. These genes represent downstream targets of Idd susceptibility genes and reflect molecular pathways differentially regulated in T1D-susceptible vs T1D-resistant strains. Interestingly, five genes overexpressed in the T1D-susceptible R3 strains are "IFN signature response" genes. The gene encoding 2'-5'-oligoadenylate synthetase-like protein 2 (Oasl2) warrants special attention. Oasl2 belongs to the IFN-induced OAS family of genes, with critical functions in innate immune responses to viruses (54, 55). OAS proteins produce 2'-5'-oligoadenylates, required to activate latent ribonucleases, leading to viral RNA degradation and inhibition of viral replication. Recent evidence suggested that an OAS1 splice site polymorphism that affects enzyme activity (55) is associated with susceptibility to T1D (56), providing a functional link between antiviral/IFN response and autoimmunity.

We analyzed the expression pattern of all 52 genes in the Idd4.1 region in NOR and NOR.NOD-Idd4 congenic mice. Seven of these genes were differentially expressed between these two strains, including Psmb6 and Alox15, as predicted by the microarray analysis. Of the remaining five genes, Pld2 and Alox12e have confirmed roles in regulating immune responses. PLD2 catalyzes hydrolysis of phosphatidylcholine to phosphatidic acid, and choline and has been proposed to regulate phagocytosis (33, 57). Indeed, NOD M{phi} have been shown to be defective in phagocytosis of apoptotic cells, compared with the T1D-resistant strains (58). Therefore, differential expression, or activity, of PLD2 may contribute to the T1D susceptibility through the control of phagocytosis. Similarly to ALOX15, ALOX12e is involved in generation of the anti-inflammatory lipoxins (32, 51) and was cloned originally from murine epidermis, as reflected in the "e" designation (59). Differential expression of Alox12e in NOR v NOR.NOD-Idd4 M{phi}/DC is an interesting finding that requires additional functional analysis. In addition, we identified a V453L variation in ALOX12e between NOD and NOR strains (Table V) that may influence enzymatic activity.

An important finding from our gene expression studies is the dependence of outcomes on the heterogeneity and physiological status of the samples under study. For example, Psmb6, Pld2, and Alox12e displayed robust Idd4-genotype-dependent differential expressed in highly enriched M{phi}/DCs that was not detectable in whole spleen, where these cells represent a minor fraction. Previously, Eaves et al. (60) reported gene expression microarray analysis of NOD, compared with NOD.B6-Idd3, -Idd5, and -Idd9 congenic strains. These experiments compared RNA isolated from whole spleen and thymus, tissues that contain developmentally and lineage-diversified cell types. Selection of these heterogeneous tissues may have obscured identification of candidate genes at these loci or expression profiles indicative of dysregulated pathways present in restricted cell types. Thus, the use of enriched, well-characterized cell types for microarray analysis may be an important component of experimental design in studies of complex traits such as T1D.

This study provides an experimental framework for evaluating positional candidate genes responsible for complex multifactorial disease. Vital resources for high-resolution mapping of complex disease loci include penetrant disease-associated phenotypes amenable to replicated analyses in congenic strains, generation of high-density polymorphic markers across the interval of interest, and strategies for prioritizing the genes. When restricted to 1.2 Mb, we report 52 genes in the Idd4 locus that required gene-specific and genome-wide expression analysis, genomic and cDNA sequencing, and functional annotation to prioritize five best candidates (Psmb6, Pld2, Alox15, Alox12e, and Cxcl16) for in-depth functional analysis. Similar challenges were confronted in analyses of a murine systemic lupus erthyematosus model where a series of variants affecting the protein sequence and expression in the functionally related Cd2/Slam complex were shown to underlie autoimmune phenotypes in the B6.Sle1 strain (61). Identification of the causal variant(s) at this locus will require in vivo transgenic analyses. Importantly, Sle1, Idd4, and multiple other autoimmune disease loci may be due to variant haplotypes rather than single-sequence variations. Indeed, loci with sufficiently strong effects to be detected by linkage to complex disease traits may often reflect the coordinated actions of closely linked genetic variants. Finally, by combining congenic resolution of the locus with genome-wide expression analyses, our data have revealed that genetic variants in Idd4 likely influence diabetes pathogenesis through regulation of IFN response thresholds in of M{phi}/DCs and perhaps other cell types. Therapeutic manipulation of this pathway in persons with genetic risk for T1D may present a useful avenue for interference with autoimmune inflammatory progression and preservation of beta cell mass.


   Acknowledgments
 
We thank Dr. Jane Rogers, Charles Sawyer, and their colleagues at the Wellcome-Sanger Centre for BAC tiling and genomic sequencing of the NOD Idd4 locus and for many helpful discussions. We thank Drs. C. Guidos, S. Scherer, and R. McInnes for critical comments on the manuscript.


   Disclosures
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
References
 
The authors have no financial conflict of interest.


   Footnotes
 
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1 This work was supported by grants to J.S.D. from the Canadian Institutes of Health Research, the Juvenile Diabetes Research Foundation, the Canadian Genetic Disease Network, National Centers of Excellence, and Genome Canada. E.A.I. was supported by fellowships from the Banting and Best Diabetes Center, University of Toronto, and the Research and Training Center, Hospital for Sick Children. Back

2 Address correspondence and reprint requests to Dr. Jayne S. Danska, Toronto Medical Discovery Tower 14-313, 101 College Street, Toronto, Ontario, Canada M5G 1L7. E-mail address: jayne.danska{at}sickkids.ca Back

3 Abbreviations used in this paper: T1D, type 1 diabetes; NOR, nonobese diabetes-resistant; Idd, insulin-dependent diabetes; CY, cyclophosphamide; DC, dendritic cell; M{phi}, macrophage; UTR, untranslated; BM, bone marrow; LN, lymph node; SAM, significance analysis of microarrays; FDR, false detection rate; cM, centiMorgans; Mb, megabase; SNP, single nucleotide polymorphism. Back

Received for publication November 1, 2005. Accepted for publication December 19, 2005.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 

  1. Wicker, L. S., J. A. Todd, L. B. Peterson. 1995. Genetic control of autoimmune diabetes in the NOD mouse. Annu. Rev. Immunol. 13: 179-200. [Medline]
  2. Serreze, D. V., E. H. Leiter. 2001. Genes and cellular requirements for autoimmune diabetes susceptibility in nonobese diabetic mice. Curr. Dir Autoimmun. 4: 31-67. [Medline]
  3. Anderson, M. S., J. A. Bluestone. 2005. The NOD mouse: a model of immune dysregulation. Annu. Rev. Immunol. 23: 447-485. [Medline]
  4. Cox, N. J., B. Wapelhorst, V. A. Morrison, L. Johnson, L. Pinchuk, R. S. Spielman, J. A. Todd, P. Concannon. 2001. Seven regions of the genome show evidence of linkage to type 1 diabetes in a consensus analysis of 767 multiplex families. Am. J. Hum. Genet. 69: 820-830. [Medline]
  5. Pociot, F., M. F. McDermott. 2002. Genetics of type 1 diabetes mellitus. Genes Immun. 3: 235-249. [Medline]
  6. Cox, N. J., B. Wapelhorst, V. A. Morrison, L. Johnson, L. Pinchuk, R. S. Spielman, J. A. Todd, P. Concannon. 2001. Seven regions of the genome show evidence of linkage to type 1 diabetes in a consensus analysis of 767 multiplex families. Am. J. Hum. Genet. 69: 820-830.
  7. Hamilton-Williams, E. E., D. V. Serreze, B. Charlton, E. A. Johnson, M. P. Marron, A. Mullbacher, R. M. Slattery. 2001. Transgenic rescue implicates beta2-microglobulin as a diabetes susceptibility gene in nonobese diabetic (NOD) mice. Proc. Natl. Acad. Sci. USA 98: 11533-11538. [Abstract/Free Full Text]
  8. Ueda, H., J. M. Howson, L. Esposito, J. Heward, H. Snook, G. Chamberlain, D. B. Rainbow, K. M. Hunter, A. N. Smith, G. Di Genova, et al 2003. Association of the T-cell regulatory gene CTLA4 with susceptibility to autoimmune disease. Nature 423: 506-511. [Medline]
  9. Wicker, L. S., G. Chamberlain, K. Hunter, D. Rainbow, S. Howlett, P. Tiffen, J. Clark, A. Gonzalez-Munoz, A. M. Cumiskey, R. L. Rosa, et al 2004. Fine mapping, gene content, comparative sequencing, and expression analyses support Ctla4 and Nramp1 as candidates for Idd5.1 and Idd5.2 in the nonobese diabetic mouse. J. Immunol. 173: 164-173. [Abstract/Free Full Text]
  10. Bennett, S. T., A. J. Wilson, F. Cucca, J. Nerup, F. Pociot, P. A. McKinney, A. H. Barnett, S. C. Bain, J. A. Todd. 1996. IDDM2-VNTR-encoded susceptibility to type 1 diabetes: dominant protection and parental transmission of alleles of the insulin gene-linked minisatellite locus. J. Autoimmun. 9: 415-421. [Medline]
  11. Pugliese, A., M. Zeller, A. Fernandez, Jr, L. J. Zalcberg, R. J. Bartlett, C. Ricordi, M. Pietropaolo, G. S. Eisenbarth, S. T. Bennett, D. D. Patel. 1997. The insulin gene is transcribed in the human thymus and transcription levels correlated with allelic variation at the INS VNTR-IDDM2 susceptibility locus for type 1 diabetes. Nat. Genet. 15: 293-297. [Medline]
  12. Vafiadis, P., S. T. Bennett, J. A. Todd, J. Nadeau, R. Grabs, C. G. Goodyer, S. Wickramasinghe, E. Colle, C. Polychronakos. 1997. Insulin expression in human thymus is modulated by INS VNTR alleles at the IDDM2 locus. Nat. Genet. 15: 289-292. [Medline]
  13. Begovich, A. B., V. E. Carlton, L. A. Honigberg, S. J. Schrodi, A. P. Chokkalingam, H. C. Alexander, K. G. Ardlie, Q. Huang, A. M. Smith, J. M. Spoerke, et al 2004. A missense single-nucleotide polymorphism in a gene encoding a protein tyrosine phosphatase (PTPN22) is associated with rheumatoid arthritis. Am. J. Hum. Genet. 75: 330-337. [Medline]
  14. Bottini, N., L. Musumeci, A. Alonso, S. Rahmouni, K. Nika, M. Rostamkhani, J. MacMurray, G. F. Meloni, P. Lucarelli, M. Pellecchia, et al 2004. A functional variant of lymphoid tyrosine phosphatase is associated with type I diabetes. Nat. Genet. 36: 337-338. [Medline]
  15. Prochazka, M., D. V. Serreze, W. N. Frankel, E. H. Leiter. 1992. NOR/Lt mice: MHC-matched diabetes-resistant control strain for NOD mice. Diabetes 41: 98-106. [Abstract]
  16. Naggert, J. K., J. L. Mu, W. Frankel, D. W. Bailey, B. Paigen. 1995. Genomic analysis of the C57BL/Ks mouse strain. Mamm. Genome 6: 131-133. [Medline]
  17. Serreze, D. V., M. Prochazka, P. C. Reifsnyder, Bridgett, E. H. Leiter. 1994. Use of recombinant congenic and congenic strains of NOD mice to identify a new insulin-dependent diabetes resistance gene. J. Exp. Med. 180: 1553-1558. [Abstract/Free Full Text]
  18. Serreze, D. V., M. Bridgett, H. D. Chapman, E. Chen, S. D. Richard, E. H. Leiter. 1998. Subcongenic analysis of the Idd13 locus in NOD/Lt mice: evidence for several susceptibility genes including a possible diabetogenic role for beta2-microglobulin. J. Immunol. 160: 1472-1478. [Abstract/Free Full Text]
  19. Fox, C. J., A. D. Paterson, S. M. Mortin-Toth, J. S. Danska. 2000. Two genetic loci regulate T cell-dependent islet inflammation and drive autoimmune diabetes pathogenesis. Am. J. Hum. Genet. 67: 67-81. [Medline]
  20. Ivakine, E. A., C. J. Fox, A. D. Paterson, S. M. Mortin-Toth, A. Canty, D. S. Walton, K. Aleksa, S. Ito, J. S. Danska. 2005. Sex-specific effect of insulin-dependent diabetes 4 on regulation of diabetes pathogenesis in the nonobese diabetic mouse. J. Immunol. 174: 7129-7140. [Abstract/Free Full Text]
  21. Harada, M., S. Makino. 1984. Promotion of spontaneous diabetes in non-obese diabetes-prone mice by cyclophosphamide. Diabetologia 27: 604-606. [Medline]
  22. Bolstad, B. M., R. A. Irizarry, M. Astrand, T. P. Speed. 2003. A comparison of normalization methods for high density oligonucleotide array data based on variance and bias. Bioinformatics 19: 185-193. [Abstract/Free Full Text]
  23. Irizarry, R. A., B. M. Bolstad, F. Collin, L. M. Cope, B. Hobbs, T. P. Speed. 2003. Summaries of Affymetrix GeneChip probe level data. Nucleic Acids Res. 31: e15[Abstract/Free Full Text]
  24. Tusher, V. G., R. Tibshirani, G. Chu. 2001. Significance analysis of microarrays applied to the ionizing radiation response. Proc. Natl. Acad. Sci. USA 98: 5116-5121. [Abstract/Free Full Text]
  25. Charlton, B., A. Bacelj, T. E. Mandel. 1988. Administration of silica particles or anti-Lyt2 antibody prevents beta-cell destruction in NOD mice given cyclophosphamide. Diabetes 37: 930-935. [Abstract]
  26. Fox, C. J.. 1999. Genetic regulation of contingent macrophage- and T cell- dependent steps in islet inflammation. Ph.D. thesis Department of Immunology, University of Toronto, Toronto, Canada.
  27. Serreze, D. V., H. D. Chapman, D. S. Varnum, M. S. Hanson, P. C. Reifsnyder, S. D. Richard, S. A. Fleming, E. H. Leiter, L. D. Shultz. 1996. B lymphocytes are essential for the initiation of T cell-mediated autoimmune diabetes: analysis of a new "speed congenic" stock of NOD.Ig mu null mice. J. Exp. Med. 184: 2049-2053. [Abstract/Free Full Text]
  28. Noorchashm, H., Y. K. Lieu, N. Noorchashm, S. Y. Rostami, S. A. Greeley, A. Schlachterman, H. K. Song, L. E. Noto, A. M. Jevnikar, C. F. Barker, A. Naji. 1999. I-Ag7-mediated antigen presentation by B lymphocytes is critical in overcoming a checkpoint in T cell tolerance to islet beta cells of nonobese diabetic mice. J. Immunol. 163: 743-750. [Abstract/Free Full Text]
  29. Chiu, P. P., D. V. Serreze, J. S. Danska. 2001. Development and function of diabetogenic T-cells in B-cell-deficient nonobese diabetic mice. Diabetes 50: 763-770. [Abstract/Free Full Text]
  30. Martin, S., D. Wolf-Eichbaum, G. Duinkerken, W. A. Scherbaum, H. Kolb, J. G. Noordzij, B. O. Roep. 2001. Development of type 1 diabetes despite severe hereditary B-lymphocyte deficiency. N. Engl. J. Med. 345: 1036-1040. [Free Full Text]
  31. Devendra, D., G. S. Eisenbarth. 2004. Interferon {alpha}: a potential link in the pathogenesis of viral-induced type 1 diabetes and autoimmunity. Clin. Immunol. 111: 225-233. [Medline]
  32. Yoshimoto, T., Y. Takahashi. 2002. Arachidonate 12-lipoxygenases. Prostaglandins Other Lipid Mediat. 68–69: 245-262.
  33. Iyer, S. S., J. A. Barton, S. Bourgoin, D. J. Kusner. 2004. Phospholipases D1 and D2 coordinately regulate macrophage phagocytosis. J. Immunol. 173: 2615-2623. [Abstract/Free Full Text]
  34. Kel, A. E., E. Gossling, I. Reuter, E. Cheremushkin, O. V. Kel-Margoulis, E. Wingender. 2003. MATCH: a tool for searching transcription factor binding sites in DNA sequences. Nucleic Acids Res. 31: 3576-3579. [Abstract/Free Full Text]
  35. Prestridge, D. S.. 1996. SIGNAL SCAN 4.0: additional databases and sequence formats. Comput. Appl. Biosci. 12: 157-160. [Abstract/Free Full Text]
  36. Schwartz, S., Z. Zhang, K. A. Frazer, A. Smit, C. Riemer, J. Bouck, R. Gibbs, R. Hardison, W. Miller. 2000. PipMaker: a web server for aligning two genomic DNA sequences. Genome Res. 10: 577-586. [Abstract/Free Full Text]
  37. Matloubian, M., A. David, S. Engel, J. E. Ryan, J. G. Cyster. 2000. A transmembrane CXC chemokine is a ligand for HIV-coreceptor Bonzo. Nat. Immunol. 1: 298-304. [Medline]
  38. Johnston, B., C. H. Kim, D. Soler, M. Emoto, E. C. Butcher. 2003. Differential chemokine responses and homing patterns of murine TCR{alpha}beta NKT cell subsets. J. Immunol. 171: 2960-2969. [Abstract/Free Full Text]
  39. Shimaoka, T., T. Nakayama, N. Fukumoto, N. Kume, S. Takahashi, J. Yamaguchi, M. Minami, K. Hayashida, T. Kita, J. Ohsumi, et al 2003. Cell surface-anchored SR-PSOX/CXC chemokine ligand 16 mediates firm adhesion of CXC chemokine receptor 6-expressing cells. J. Leukocyte Biol. :
  40. Shimaoka, T., T. Nakayama, N. Kume, S. Takahashi, J. Yamaguchi, M. Minami, K. Hayashida, T. Kita, J. Ohsumi, O. Yoshie, S. Yonehara. 2003. Cutting edge: SR-PSOX/CXC chemokine ligand 16 mediates bacterial phagocytosis by APCs through its chemokine domain. J. Immunol. 171: 1647-1651. [Abstract/Free Full Text]
  41. Davis, R. C., E. E. Schadt, A. C. Cervino, M. Peterfy, A. J. Lusis. 2005. Ultrafine mapping of SNPs from mouse strains C57BL/6J, DBA/2J, and C57BLKS/J for loci contributing to diabetes and atherosclerosis susceptibility. Diabetes 54: 1191-1199. [Abstract/Free Full Text]
  42. Ghosh, S., S. M. Palmer, N. R. Rodrigues, H. J. Cordell, C. M. Hearne, R. J. Cornall, J. B. Prins, P. McShane, G. M. Lathrop, L. B. Peterson, et al 1993. Polygenic control of autoimmune diabetes in nonobese diabetic mice. Nat. Genet. 4: 404-409. [Medline]
  43. Grattan, M., Q. S. Mi, C. Meagher, T. L. Delovitch. 2002. Congenic mapping of the diabetogenic locus Idd4 to a 5.2-cM region of chromosome 11 in NOD mice: identification of two potential candidate subloci. Diabetes 51: 215-223. [Abstract/Free Full Text]
  44. Schadt, E. E., S. A. Monks, T. A. Drake, A. J. Lusis, N. Che, V. Colinayo, T. G. Ruff, S. B. Milligan, J. R. Lamb, G. Cavet, et al 2003. Genetics of gene expression surveyed in maize, mouse and man. Nature 422: 297-302. [Medline]
  45. Akiyama, K., S. Kagawa, T. Tamura, N. Shimbara, M. Takashina, P. Kristensen, K. B. Hendil, K. Tanaka, A. Ichihara. 1994. Replacement of proteasome subunits X and Y by LMP7 and LMP2 induced by interferon-{gamma} for acquirement of the functional diversity responsible for antigen processing. FEBS Lett. 343: 85-88. [Medline]
  46. Akiyama, K., K. Yokota, S. Kagawa, N. Shimbara, T. Tamura, H. Akioka, H. G. Nothwang, C. Noda, K. Tanaka, A. Ichihara. 1994. cDNA cloning and interferon-{gamma} down-regulation of proteasomal subunits X and Y. Science 265: 1231-1234. [Abstract/Free Full Text]
  47. Hisamatsu, H., N. Shimbara, Y. Saito, P. Kristensen, K. B. Hendil, T. Fujiwara, E. Takahashi, N. Tanahashi, T. Tamura, A. Ichihara, K. Tanaka. 1996. Newly identified pair of proteasomal subunits regulated reciprocally by interferon-{gamma}. J. Exp. Med. 183: 1807-1816. [Abstract/Free Full Text]
  48. Groettrup, M., M. van den Broek, K. Schwarz, A. Macagno, S. Khan, R. de Giuli, G. Schmidtke. 2001. Structural plasticity of the proteasome and its function in antigen processing. Crit. Rev. Immunol. 21: 339-358. [Medline]
  49. Kuhn, H., M. Walther, R. J. Kuban. 2002. Mammalian arachidonate 15-lipoxygenases structure, function, and biological implications. Prostaglandins Other Lipid Mediat. 68 69: 263-290.
  50. Zijlstra, F. J., J. H. Wilson. 1991. 15-HETE is the main eicosanoid present in mucus of ulcerative proctocolitis. Prostaglandins. Leukot. Essent. Fatty Acids 43: 55-59. [Medline]
  51. Funk, C. D., X. S. Chen, E. N. Johnson, L. Zhao. 2002. Lipoxygenase genes and their targeted disruption. Prostaglandins Other Lipid Mediat. 68–69: 303-312.
  52. Levy, B. D., C. B. Clish, B. Schmidt, K. Gronert, C. N. Serhan. 2001. Lipid mediator class switching during acute inflammation: signals in resolution. Nat. Immunol. 2: 612-619. [Medline]
  53. Serhan, C. N., A. Jain, S. Marleau, C. Clish, A. Kantarci, B. Behbehani, S. P. Colgan, G. L. Stahl, A. Merched, N. A. Petasis, et al 2003. Reduced inflammation and tissue damage in transgenic rabbits overexpressing 15-lipoxygenase and endogenous anti-inflammatory lipid mediators. J. Immunol. 171: 6856-6865. [Abstract/Free Full Text]
  54. Knapp, S., L. J. Yee, A. J. Frodsham, B. J. Hennig, S. Hellier, L. Zhang, M. Wright, M. Chiaramonte, M. Graves, H. C. Thomas, et al 2003. Polymorphisms in interferon-induced genes and the outcome of hepatitis C virus infection: roles of MxA, OAS-1 and PKR. Genes Immun. 4: 411-419. [Medline]
  55. Bonnevie-Nielsen, V., L. Leigh Field, S. Lu, D. J. Zheng, M. Li, P. M. Martensen, T. B. Nielsen, H. Beck-Nielsen, Y. L. Lau, F. Pociot. 2005. Variation in antiviral 2',5'-oligoadenylate synthetase (2'5'AS) enzyme activity is controlled by a single-nucleotide polymorphism at a splice-acceptor site in the OAS1 gene. Am. J. Hum. Genet. 76: 623-633. [Medline]
  56. Field, L. L., V. Bonnevie-Nielsen, F. Pociot, S. Lu, T. B. Nielsen, H. Beck-Nielsen. 2005. OAS1 splice site polymorphism controlling antiviral enzyme activity influences susceptibility to type 1 diabetes. Diabetes 54: 1588-1591. [Abstract/Free Full Text]
  57. Melendez, A. J., J. M. Allen. 2002. Phospholipase D and immune receptor signalling. Semin. Immunol. 14: 49-55. [Medline]
  58. O’Brien, B. A., Y. Huang, X. Geng, J. P. Dutz, D. T. Finegood. 2002. Phagocytosis of apoptotic cells by macrophages from NOD mice is reduced. Diabetes 51: 2481-2488. [Abstract/Free Full Text]
  59. Funk, C. D., D. S. Keeney, E. H. Oliw, W. E. Boeglin, A. R. Brash. 1996. Functional expression and cellular localization of a mouse epidermal lipoxygenase. J. Biol. Chem. 271: 23338-23344. [Abstract/Free Full Text]
  60. Eaves, I. A., L. S. Wicker, G. Ghandour, P. A. Lyons, L. B. Peterson, J. A. Todd, R. J. Glynne. 2002. Combining mouse congenic strains and microarray gene expression analyses to study a complex trait: the NOD model of type 1 diabetes. Genome Res. 12: 232-243. [Abstract/Free Full Text]
  61. Wandstrat, A. E., C. Nguyen, N. Limaye, A. Y. Chan, S. Subramanian, X. H. Tian, Y. S. Yim, A. Pertsemlidis, H. R. Garner, Jr, L. Morel, E. K. Wakeland. 2004. Association of extensive polymorphisms in the SLAM/CD2 gene cluster with murine lupus. Immunity 21: 769-780. [Medline]

Related articles in The JI:

IN THIS ISSUE

The JI 2006 176: 2679-2680. [Full Text]  



This article has been cited by other articles:


Home page
 J. Immunol.Home page
A. M. Marleau, K. L. Summers, and B. Singh
Differential Contributions of APC Subsets to T Cell Activation in Nonobese Diabetic Mice
J. Immunol., April 15, 2008; 180(8): 5235 - 5249.
[Abstract] [Full Text] [PDF]

Home page
 DiabetesHome page
M. McDuffie, N. A. Maybee, S. R. Keller, B. K. Stevens, J. C. Garmey, M. A. Morris, E. Kropf, C. Rival, K. Ma, J. D. Carter, et al.
Nonobese Diabetic (NOD) Mice Congenic for a Targeted Deletion of 12/15-Lipoxygenase Are Protected From Autoimmune Diabetes
Diabetes, January 1, 2008; 57(1): 199 - 208.
[Abstract] [Full Text] [PDF]

Home page
 DiabetesHome page
E. A. Ivakine, S. M. Mortin-Toth, O. M. Gulban, A. Valova, A. Canty, C. Scott, and J. S. Danska
The Idd4 Locus Displays Sex-Specific Epistatic Effects on Type 1 Diabetes Susceptibility in Nonobese Diabetic Mice
Diabetes, December 1, 2006; 55(12): 3611 - 3619.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Related articles in The JI
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Ivakine, E. A.
Right arrow Articles by Danska, J. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ivakine, E. A.
Right arrow Articles by Danska, J. S.
Right arrowPubmed/NCBI databases
*Gene
*Substance via MeSH
Medline Plus Health Information
*Diabetes Type 1

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS