- Open Access
Pancreatic islet expression profiling in diabetes-prone C57BLKS/J mice reveals transcriptional differences contributed by DBA loci, including Plagl1 and Nnt
© Anderson et al; licensee BioMed Central Ltd. 2009
Received: 04 November 2008
Accepted: 22 January 2009
Published: 22 January 2009
C57BLKS/J (BLKS) mice are susceptible to islet exhaustion in insulin-resistant states as compared with C57BL6/J (B6) mice, as observed by the presence of the leptin receptor (Lepr) allele, Leprdb/db. Furthermore, DBA2/J (DBA) mice are also susceptible to β-cell failure and share 25% of their genome with BLKS; thus the DBA genome may contribute to β-cell dysfunction in BLKS mice.
Here we show that BLKS mice exhibit elevated insulin secretion, as evidenced by improved glucose tolerance and increased islet insulin secretion compared with B6 mice, and describe interstrain transcriptional differences in glucose response. Transcriptional differences between BLKS and B6 mice were identified by expression profiling of isolated islets from both strains. Genomic mapping of gene expression differences demonstrated a significant association of expression differences with DBA loci in BLKS mice (P = 4×10-27).
Two genes, Nicotinamide nucleotide transhydrogenase (Nnt) and Pleiomorphic adenoma gene like 1 (Plagl1), were 4 and 7.2-fold higher respectively in BLKS islets, and may be major contributors to increased insulin secretion by BLKS islets. Contrary to reports for B6 mice, BLKS mice do not harbor a mutant Nnt gene. We detected 16 synonymous polymorphisms and a two-amino acid deletion in the Plagl1 gene in BLKS mice. Several inflammatory glucose-responsive genes are expressed at a higher level in BLKS, suggesting an inflammatory component to BLKS islet dysfunction. This study describes physiological differences between BLKS and B6 mice, and provides evidence for a causative role of the DBA genome in β-cell dysfunction in BLKS mice.
Type 2 diabetes mellitus can be considered as a two-stage disease. First the body becomes resistant to circulating insulin, and second, when coupled with pancreatic β-cell dysfunction, overt diabetes precipitates . Insulin resistance alone, without β-cell exhaustion, will not lead to hyperglycemia. In this case the islet compensates by increasing insulin production and β-cell populations are maintained. Differences in β-cell adaptation to increased insulin demand likely reflect an underlying genetic component. Indeed, in B6 mice, but not in genetically dissimilar BTBR mice, insulin resistance induced by the leptin ob allele is well compensated by islet expansion . The genetic alteration in BTBR mice which underpins this failure to adapt to increased insulin requirements was recently identified and localized to a gene contributing to islet vascularization .
Obese (Leprdb/db) B6 and 129/J mice are insulin resistant but compensate this potentially pathogenic process by islet hypertrophy and hyperinsulinemia [4, 5]. However, obese (Leprdb/db) DBA mice not only develop insulin resistance but also severe diabetes due to β-cell loss. Thus the DBA background may be considered a diabetogenic strain exhibiting a predisposition to β-cell failure. DBA mice copiously secrete insulin (10-fold over basal) in response to a glucose bolus compared with a modest (3-fold) increase in B6 mice . This hypersecretion phenotype (here after referred to as elevated insulin secretion) is retained in isolated islets, and confirms a functional difference between these two strains specific to the islet. These data suggest that elevated secretion of insulin may be deleterious to the β cell and results in β-cell exhaustion in DBA mice.
BLKS mice also exhibit pancreatic islet failure when made insulin resistant by the ob or db Lepr alleles [5, 7, 8], as evidenced by gross islet atrophy and β-cell loss. This strain is closely related to B6; however, it harbors a small contribution from the DBA strain. Genetic analysis of the BLKS mice has uncovered a contribution of approximately 71% from B6, 25% from DBA, and 4% from other strains in BLKS mice [9–11]. Given the clear differences in severity of diabetes between B6-Leprdb/db mice and BLKS-Leprdb/db mice, and the similarity between BLKS and DBA mice, it is tempting to speculate that islet sensitivity may be conferred by alleles from the DBA genome.
In this report we investigated islet function in non-obese B6 and BLKS mice. We found that BLKS mice are profoundly more glucose tolerant than B6 and this difference was associated with elevated secretion of insulin in islets. We investigated the differences in gene expression in isolated islets between the two strains, in both basal and stimulated (insulin secretion stimulated) conditions, and found that 75% of the gene expression differences in the BLKS islets were contributed from genes within DBA loci. Further, several genes regulated by glucose in both strains suggest early inflammatory effects associated with increased glucose (Txnip,Lnc2,Gad1 Slc7a3 and Spp1). Notably, we identified significantly higher expression of Nnt and Plagl1 (ZAC1) in BLKS islets, two genes previously associated with islet dysfunction.
Results and discussion
Glucose homeostasis in non-obese BLKS and B6 mice
Islet insulin secretion and pancreatic immunohistochemistry
Islet gene expression differences between glucose concentrations
Gene expression fold-change differences in islets from low to high glucose in BLKS and B6 mice
Interstrain fold difference (BLKS/B6)
Glucose response (fold change)
S100 calcium-binding protein A9
secreted phosphoprotein 1
solute carrier family 7 member 3
cell growth regulator with EF-hand domain 1
tweety homolog 1
tweety homolog 1
FK506-binding protein 11
FK506-binding protein 11
leucine-rich alpha-2-glycoprotein 1
neuronal pentraxin II
hydroxy-delta-5-steroid dehydrogenase, 3 beta- and steroid delta-isomerase 7
serpin peptidase inhibitor member 7
translocation-associated membrane protein 1
family with sequence similarity 46, member A
glutamate decarboxylase 1
Gene expression fold-change differences in islets from BLKS mice compared with those from B6 mice
Interstrain fold difference (BLKS/B6)
pleiomorphic adenoma gene-like 1
nicotinamide nucleotide transhydrogenase
nicotinamide nucleotide transhydrogenase
nicotinamide nucleotide transhydrogenase
secreted phosphoprotein 1
Purkinje cell protein 4
S100 calcium-binding protein A9
BTB domain-containing 9
CAP, adenylate cyclase-associated protein 1
1-acylglycerol-3-phosphate O-acyltransferase 4
thymocyte nuclear protein 1
1-acylglycerol-3-phosphate O-acyltransferase 4
serine carboxypeptidase 1
N-acetylglucosamine-1-phosphate transferase, gamma subunit
fatty acid-binding protein 3
aldolase B, fructose-bisphosphate
zinc finger, ZZ-type with EF-hand domain 1
serpin peptidase inhibitor, clade F member 1
transmembrane channel-like 7
major histocompatibility complex, class I, A
solute carrier family 5 member 10
THUMP domain-containing 1
major histocompatibility complex Q1b
Table 1 shows gene expression fold-change differences between low and high-glucose conditions in BLKS and B6 mice. We highlight the subset with BLKS glucose response of at least 2-fold. These transcriptional differences associated with glucose response in islets suggest an underlying inflammatory component. Lipocalin 2 (Lcn2) transcription increases 2.4-fold in response to higher glucose levels and is 2.07-fold higher in BLKS islets. This gene is regulated by the cytokine leptin in insulinoma cells  and interestingly BLKS mice also have 40% more of a specific leptin receptor (Leprotl1) at high glucose levels – 1.44 and 1.09-fold under high and low-glucose conditions, respectively. Lcn2 is also known to be up-regulated in rat models of type 2 diabetes, and associated with inflammatory functions  including apoptosis. The glucose-responsive gene Glutamate decarboxylase 1 (Gad1), which is an islet-derived autoantigen in insulin-dependent diabetes , may be a marker for β-cell loss or toxicity and has been shown to be glucose responsive. We observed a Gad1 decrease in both mouse models, but more so in the BLKS mice. The gene with the largest change due to glucose in BLKS mice was S100a9. This gene is expressed in monocytes and pancreatic cell lines, and is essential for pancreatic leukocyte infiltration [17, 18]. In addition to being an inflammatory marker S100a9 is associated with islet autoimmunity . Another glucose-responsive gene, Slc7a3, is likely an indirect marker for the presence of resident islet inflammatory cells. Slc7a3 is a polar amino acid transporter, which provides L-arginine that can be used in nitric oxide (NO) production. NO synthesis and overproduction has been associated with β-cell dysfunction in diabetic rats .
In addition to the inflammation-associated genes we also observed strong induction by glucose of the gene encoding Thioredoxin interacting protein (Txnip) in both BLKS and B6 mice. Txnip has been recently linked with glucose-induced β-cell loss . Chen et al showed that glucose incubation up-regulated Txnip and increased apoptosis, which they speculated may be via the intrinsic mitochondrial apoptotic pathway. They also showed that constitutive over-expression is not necessary for apoptosis, but that the glucose-induced increase is sufficient. Furthermore, Parikh et al have shown that, in humans, Txnip over-expression reduces basal and insulin-stimulated glucose uptake , is elevated by glucose and suppressed by insulin. We confirm the increase of Txnip expression in response to glucose, although we do not observe any differential response to glucose between B6 and BLKS mice. On the contrary, BLKS mice have 2-fold lower levels of Txnip than B6 mice. Thus Txnip is unlikely to underlie the islet failure in BLKS mice, but more likely to represent the typical response to glucose.
Islet gene expression differences between B6 and BLKS strains
Table 2 shows a list of genes which were more than 2.5-fold different in BLKS islets compared with B6 islets in high glucose conditions (column 4, High Gluc.). All members of this list were significantly different between strains, regardless of glucose concentration, after controlling the false discovery rate at 0.80 . In addition to its induction by glucose within each strain (Table 1) Spp1 expression is also significantly different between strains and is 3.5-fold higher in BLKS islets (Table 2). Spp1 or osteopontin is a multifunctional cytokine and a potential diagnostic predictor of diabetic end-stage renal disease . Arafat et al have shown that Spp1 averts cytokine-mediated β-cell toxicity through negative regulation of NO production . As we have shown, both BLKS and B6 mice up-regulate Slc7a3, a gene associated with NO production, but BLKS mice always express higher levels (1.49 to 1.97-fold). Thus, the excess capacity for NO production in BLKS mice might counter the protective up-regulation of Spp1. Other genes such as S100a9, Aldob, and Slc5a10 demonstrate a stronger glucose response in only one of the mouse strains. S100a9, a diabetes-associated pro-inflammatory molecule , had a much larger increase due to increasing glucose in BLKS mice (4.3-fold) vs. B6 mice (1.6-fold). Aldob, encoding a glycolytic enzyme that is decreased in human diabetic islets , and Slc5a10, a sodium/glucose co-transporter, had stronger induction due to increasing glucose in B6 mice (1.5 and 3.5-fold respectively) vs. BLKS mice (1.1 and 1.6-fold respectively).
From Table 2, 32 genes were more than 2.5-fold different in expression levels in BLKS islets compared with B6 islets; 16 genes were expressed at higher levels and 16 genes expressed at lower levels. Thirty of these genes displayed fold differences with magnitudes between 2.5 and 7.15. However, two genes had notable differences above this range.
The first, an uncharacterized transcript (6330403K07RIK), was 17.4-fold higher in BLKS islets. At this time, this gene's product has no assigned function and has no known homology to any protein. Interestingly, two protein sequences were identified from the NCBI gene database corresponding to 6330403K07RIK which had been translated from DBA genomic sequence and B6 mRNA sequence and were found to be different. The protein sequence from DBA mice (AAR87485) was different at four residues compared with the B6 sequence (BAB31072) over the 121 residues.
The second gene with profound differential expression was H2-Q1, with a 29.3-fold lower level in the islets from BLKS mice. H2-Q1 localizes to the H2 locus in mice on chromosome 17. B6 mice are known to be isogenic for the H2b haplotype and BLKS mice are a recombinant congenic strain carrying the H2d haplotype . The difference in H2-Q1 expression likely represents a difference conferred by the two haplotypes, and the H2-Q1 gene is possibly not expressed in mice with the H2d haplotype. Furthermore, another gene from the H2 locus was identified in Table 2 (H2-K1), and suggests that the influence of H2 haplotype can manifest in multiple gene expression differences.
In further studies, we focused on the two genes Plagl1 and Nnt, which have previously been implicated in islet dysfunction in both humans and mice.
Transcriptional differences in Plagl1 and Nnt
The second-most highly expressed gene in BLKS islets compared with B6 islets was Plagl1 (7.2-fold higher expression; Table 2). The expression state and magnitude was confirmed by qRT-PCR under the same conditions of high glucose in a separate experiment (Figure 3A). Plagl1 has been previously implicated in human transient neonatal diabetes mellitus (TNDM) , and can result from paternal uniparental disomy of chromosome 6, or paternal duplication of 6q24 (the TNDM locus). These genetic aberrations suggest the disease results from over-expression of an imprinted gene . If BLKS islets are indeed predisposed to failure (when confronted with insulin resistance), then over-expression of Plagl1 would be consistent with the expression observed in human TNDM. Similarly, mice which over-express the human TNDM locus (including Plagl1) exhibit impaired glucose homeostasis . Several functions for Plagl1 have been reported . Interestingly, evidence that Plagl1 can regulate apoptosis may be relevant to the β-cell dysfunction in these mice. In BLKS mice, the Plagl1 gene is within a locus predicted to be inherited from the DBA parental strain. To investigate the underlying cause for the difference in Plagl1 expression we sequenced the four exons of the gene. Contrary to the published DNA sequence of Plagl1 in B6 mice, BLKS mice have 16 synonymous polymorphisms and a two-amino acid deletion (623Glu-624Pro; Figure 3B). These polymorphisms are contained within the third exon of the gene. The 16 synonymous polymorphisms may not affect the function of the protein, but could perhaps contribute to the higher expression of the Plagl1 transcript by affecting RNA stability. Additionally the 6 bp deletion could alter the RNA stability or possibly change the function of the protein product. The 623Glu-624Pro deletion in BLKS mice lies at the end of a repetitive Pro-Glu-Gln repeat region not found in the human paralog. Despite this structural difference within the Plagl1 gene of human and mice, functional differences have not been observed . Thus it is uncertain whether the deletion of the 623Glu-624Pro is functionally relevant to the β-cell phenotype in BLKS mice.
We observed three different probes for Nnt that all showed a clear difference in expression (Table 2). On average, the Nnt was expressed at 4-fold higher levels in BLKS islets in high-glucose conditions. It was observed that the differences in Nnt expression were present in islets cultured in low-glucose conditions (Table 2, column 4); however, this was greater in islets following a stimulated pretreatment. B6 mice were recently shown to be glucose intolerant, as a result of impaired insulin secretion, and this was linked to mutation of the Nnt gene . In this study, a quantitative trait locus (QTL) analysis in B6 and C3H/HeH mice revealed that expression of Nnt was 5-fold lower in B6 islets, and is similar with the 4-fold level we observed compared with BLKS islets. Nnt is a nuclear-encoded mitochondrial protein involved in detoxification of reactive oxygen species , which is crucial for their removal and to reduce their deleterious effect on mitochondrial ATP production. Intact Nnt will likely lead to higher ATP levels and thus enhanced insulin secretion in the β cell. To confirm the role of Nnt in insulin secretion, B6 mice were rescued with an Nnt transgene which improved the insulin secretion . Furthermore,in vitro analyses using insulin-secreting Min6 cells and isolated islets with reduced Nnt function demonstrated diminished insulin secretion associated with lower ATP levels . An intact Nnt gene might be expected to lead to higher ATP levels in BLKS β cells and thus increased insulin secretion.
The observations of altered expression of Plagl1 and Nnt between these two strains could account for the phenotypic differences observed in insulin-resistant states, such as enforced by deficiency of leptin or its receptor. Their individual prior association with human and/or mouse diabetes suggest these genes are functionally relevant to the proper action or survival of the β cell. The expression differences of these two genes alone may possibly be sufficient to elicit a physiological effect through their interaction. Indeed, the elevated secretion of insulin expected in β cells with intact Nnt function may precipitate an apoptotic event in a pro-apoptotic, Plagl1-enriched cell.
Expression differences are mainly contributed by the DBA loci
We hypothesized that the DBA genome contributes to the islet dysfunction. The locations of the DBA blocks in the BLKS genome have been mapped by other groups [9–11]. We noticed that Plagl1 and Nnt were located in DBA loci of the BLKS genome, and investigated the genetic origin of the other genes in Table 2. We found that 28 out of these 32 genes were derived from DBA loci (Table 2, last three columns), based on the genetic mapping of Davis et al .
We used the Kruskal-Wallis test to determine if there was a statistical difference in global gene expression fold-change distributions between probes from the different genomes. At the level of the whole genome, we found there was very high statistical significance for fold-change differences between DBA, B6 and Other (χ2 = 126.52, P < 4× 10-27, for probes with significant [P < 1× 10-6] fold changes). We also investigated each chromosome individually using the same test, and found that many chromosomes showed significant differences in fold change between block types (Additional file 1, supplemental table). Chromosomes 6, 7, 8, 13 and 17 were all significant for fold-change distribution differences between the genomes. These data indicate an association of the DBA loci with the gene expression differences in BLKS mouse islets. In a QTL screen to identify loci which influence metabolic phenotypes between B6 and BLKS mice, Mu and colleagues  identified two suggestive LOD scores for plasma glucose on chromosomes 8 and 17, both identified in our study to harbor significant gene expression differences. Additionally, the recent report by Aston-Mourney and co-workers  identified suggestive LOD scores for insulin secretion on chromosomes 5, 6 and 7, of which two were identified in our study.
We have analyzed the physiological differences between B6 and BLKS mice with attention to their glucose homeostasis/islet phenotype. We have shown several transcriptional differences linked to diabetes and inflammatory mechanisms leading to islet damage. Glucose stimulation appears to elicit inflammatory responses, and particularly increases in the genes Lcn2, Spp1, S100a9, and Slc7a3. These changes are seen in both strains, but expression is always higher in BLKS mice. Altogether this suggests that islet dysfunction in BLKS has a strong inflammatory component. The largest islet gene expression differences were in Plagl1 and Nnt, two genes implicated in β-cell biology. We identified genetic differences in these genes between the two strains. The interaction of these two genes may account for the β-cell dysfunction commonly observed in BLKS mice. Lastly, we undertook a genomic approach to translate islet expression data, and revealed a significant and primary contribution from the DBA genome. These data demonstrate the utility of combining genomic, genetic, and physiological data to aid in the delineation of multi-factorial phenotypes.
All mice were maintained in accordance with the Institutional Animal Use and Care Committee of Amgen Inc. B6 and BLKS male mice were obtained at 6 weeks of age from The Jackson Laboratory (Bar Harbor, ME). Mice received food and water ad libitum and were maintained on a 12:12 h light-dark cycle (lights on at 6:30 am) and housed one per cage. Fasting (4 h) blood samples (150 μl) were collected from the retro-orbital sinus of non-anesthetized mice into EDTA plasma tubes and blood glucose was measured using a OneTouch Profile glucometer (LifeScan, Milpitas, CA). For GTT, mice were fasted from 9:00 pm to 9:00 am. Blood glucose was measured at 9:00 am, and an intraperitoneal glucose bolus (2 g/kg body weight) was administered to conscious, unrestrained mice. During the GTT, blood glucose was measured at 30, 60 and 120 min. Plasma insulin levels were analyzed using the LINCOplex mouse endocrine immunoassay panel following the manufacturer's instructions (Millipore, St. Charles, MO).
Pancreatic islet isolation
Pancreatic islets were isolated from 12-week-old mice. After clamping the common bile duct as it joins the intestine, the pancreas was inflated with 5 ml of collagenase type XI (0.6 mg/ml; Sigma-Aldrich, St. Louis, MO) diluted in Hanks balanced salt solution (HBSS; Sigma-Aldrich). The distended pancreas was removed and incubated at 37°C for 10 min, and the islets were dispersed by gentle shaking. Enzymatic digestion was stopped by the addition of 45 ml of cold HBSS containing 10%FBS. Two rounds of centrifugation (300× g for 2 min) and washes with fresh HBSS/FBS were carried out, and the islets were resuspended in 10 ml of HBSS/FBS buffer. The slurry was layered on top of a prepared histopaque gradient (comprised of a 10 ml lower layer of Histopaque 1.119, and 10 ml upper layer of Histopaque 1.077; Sigma-Aldrich). Following centrifugation of the gradient at 1000× g for 30 min, the islets were collected from the top of the 1.077 interface, pipetted onto a 70 μm cell strainer, and washed with HBSS/FBS buffer. Finally, the islets were rinsed into a Petri dish, isolated from any contaminating exocrine material visually using a dissecting microscope, and cultured in RPMI-1640 media (Cellgro, University of Miami, FL), supplemented with 10%FBS and 100 U/ml of penicillin, and 100 μg/ml of streptomycin at 37°C in 5% CO2. Prior to all experiments, islets were cultured for 48 h in RPMI-1640 media containing 11 mM glucose, and reselected for use based on morphology as assessed with a dissection microscope.
Glucose-stimulated insulin secretion assay
Twenty-four hours prior to the secretion assay, islets were picked into 24-well inserts (Multiwell Insert System, 8.0 μm pore size; BD Biosciences, San Jose, CA) in groups of 10 similarly sized islets and cultured in 1 ml/well RPMI-1640. For the insulin secretion assay, Krebs-Ringer Bicarbonate (KRBH) Buffer was prepared (129 mM NaCl, 4.8 mM KCl, 1.2 mM KH2PO4, 1.2 mM MgSO4.7H2O, 10 mM HEPES, 2.5 mM CaCl2.2H2O and 4.8 mM NaHCO3) and oxygenated for 15 min with 95%O2/5%CO2. Bovine serum albumin (fatty acid-free, Cat. No. A6003; Sigma-Aldrich) at a final concentration of 0.625% was added to the buffer, then the buffer was warmed to 37°C and pH adjusted to 7.4 with 5 M NaOH. The islets were starved for one hour at 37°C in KRBH + 1 mM glucose (Cat. No. 99-787-CI, Cellgro) by transferring the 24-well insert from the RPMI-1640 culture media to a new 24-well plate containing 1 ml/well KRBH + BSA and 1 mM glucose. The islets were transferred to the experimental plates containing 1 ml/well of KRBH + BSA and 1 mM, 10 mM and 16.7 mM glucose. Following a one-hour incubation at 37°C, the insert containing the islets was removed and the buffer plate frozen at -80°C. The insulin ELISA was carried out using the Ultra Sensitive Mouse Insulin ELISA kit (Cat. No. 90080; Crystal Chem Inc. Downers Grove, IL). Five microliters of each buffer sample was run in duplicate per assay protocol.
DNA and RT-PCR analysis
DNA was isolated from tail tissue and used for PCR and direct sequencing analysis of the Nnt gene. RT-PCR using the One-Step System (Invitrogen, Carlsbad, CA) and islet RNA was performed using the oligonucleotides, Nnt forward-TACAAGAGCTGCCGCTTTGGA, and Nnt reverse-AGACCCACTAAAGGTGACTCCG; the product was used for direct DNA sequencing.
RNA isolation kits from QIAGEN (micro-RNAeasy kit; Valencia, CA) were used to prepare RNA from purified islets with a final eluate of 15 μl per mouse. RNA was quantified using Agilent's 2100 Bioanalyzer (Santa Clara, CA). Real time PCR was performed using 0.2 μM of Plagl1 primer and probe (forward-TCAAGTGCTCGAAGGCTGAGT, reverse-TGTGTGGCCATGTGTCTCATC, probe-FAM-TGGCAAAGCCTTCGTCTCCAAGTATAAGC-BHQ). Primers were derived from different exons to avoid amplification of any residual genomic DNA. The RT-PCR reactions were done using QIAGEN's Quantitect Multiplex RT-PCR kit. The total reaction volume was 20 μl per well and 20 ng RNA per reaction. Eight replicates of islets isolated from B6 and four replicates of BLKS were run with cyclophilin used as the housekeeping gene. Amplification and quantification of PCR products was performed on an ABI Prism 7900 (Applied Biosystems, Foster City, CA). Data were exported into Excel and analyzed using the delta CT method.
Pancreata were removed and fixed in buffered zinc formalin and embedded in paraffin. Pancreas sections (5 μm) were deparaffinzed, hydrated and blocked for non-specific reactivity with CAS block (Zymed Lab., San Francisco, CA). Sections were incubated with guinea pig anti-swine insulin (A562; DAKO, Carpinteria, CA) at 1:3000 for 45 min at room temperature. Insulin was detected by biotinylated goat anti-guinea pig (BA-7000; Vector Laboratories Inc., Burlingame CA) at 1:100. Slides were quenched with 3% H2O2 and followed with avidin-biotin HRP Complex (Vector Lab., Burlingame CA). Reaction sites were visualized with DAB (DAKO). Next, slides were rinsed with PBS thoroughly and incubated with rabbit anti-human glucagon (A565; DAKO) at 1:800 for 1 h at room temperature. Glucagon was detected by a biotinylated goat anti-rabbit secondary (BA-1000; Vector Lab.) at 1:200 and followed with avidin-biotin alkaline phosphatase complex. Reaction sites were visualized with alkaline phosphatase substrate (Vector Lab.).
Experimental design for islet expression profiling
We investigated gene expression differences in islets from B6 and BLKS mice and between pretreatments in the presence of low (5 mM) and high glucose (20 mM) in the culture media (to simulate basal and stimulated conditions respectively). To provide >85% confidence that each fold change >1.5 was true, we used n = 5 per group. To eliminate differences in gene expression contributed by variation across mice within a group, we did not pool islets from different animals but kept them separate. Islets were then cultured separately for 24 h in RPMI-1640 media containing either 5 mM or 20 mM glucose. At the end of the culture pretreatment, the islets were picked from the culture vessel, rinsed briefly in HBSS/FBS buffer and homogenized in RLT buffer. RNA was prepared using the micro-RNAeasy kit (QIAGEN). On average, 580 ng of total RNA was recovered from the isolated islets from each individual mouse.
Pancreatic insulin content
Pancreata were excised and frozen in liquid nitrogen and stored at -80°C. Protein content was determined using the BCA Protein Assay Kit (Thermo Scientific Rockford, IL) following homogenization of the pancreata for 2 min at 25 Hz in 1 ml of 0.18 M HCl and 70% ETOH in a TissueLyser (Qiagen Gmbh, Germany). The tissue lysate was clarified for 10 min at 4000 rpm and the supernatant was transferred to a fresh microcentrifuge tube. For the protein assay, the supernatant was 10× diluted in PBS. The insulin content assay was performed using Ultra Sensitive Mouse Insulin ELISA kit (Crystal Chem). The tissue lysate was diluted 1000× and 6000× with sample diluent supplied with the kit before running the assay. All samples for both assays were run in duplicate.
Total RNA was profiled following the Agilent Two-Color Microarray-Based Gene Expression Analysis Protocol v4.0.2. Briefly, 200 ng of total RNA from each sample was separately amplified and labeled with Cy3- and Cy5-labeled CTP (Perkin Elmer; Wellesley, MA) using the Agilent low-input linear amplification kit. Labeled cRNA was purified using the Qiagen RNeasy Mini kit protocol for liquid samples (QIAGEN). Purified cRNA was quantified using the Nanodrop Fluorospectrometer ND-3000 (Wilmington, DE). Similar amounts of Cy3- and Cy5-labeled cRNA were combined and fragmented for 30 min at 60°C. The products were hybridized to Agilent Mouse Whole Genome Microarrays, and the resulting array data were extracted using Agilent Feature Extraction Software v8.1.
Raw data for each array were analyzed with Matlab software (The Mathworks Inc. Natick, MA). Data were not analyzed as dye ratios, but as dye-normalized relative fluorescence units (RFUs). Dye normalization was accomplished by pairing Cy3 and Cy5 values, and determining the effect of dye on RFU magnitude by including dye as a factor in an ANOVA analysis of the binary logarithm-transformed probe intensities, including all two-factor interactions. The other factors in the ANOVA were strain and glucose pretreatment. Found to be significant, the effect of reporter dye was subtracted in a least-squares manner. Ultimately, technical and then biological replicates were averaged prior to fold-change calculations. Raw and normalized data discussed in this publication have been deposited in National Center for Biotechnology Information's Gene Expression Omnibus (GEO, http://www.ncbi.nlm.nih.gov/geo/) and are accessible through GEO Series accession number GSE11257.
Genotyping data from Davis et al  were used to divide the genome into DBA and B6 regions. Each region is bordered by two markers for the same strain with no intervening markers of any other strain. Regions with informative SNPs, not attributable to DBA or B6 were labeled Other. Each microarray probe was mapped to a region by determining if its target gene overlapped with any identified region. Probes not mapped to one of the three regions above were labeled as Unmapped. Fold-change distributions of the four classes of probe were compared using the Kruskal-Wallis test, a nonparametric ANOVA testing the hypothesis that groups have the same median fold-change rank. Fold changes were calculated between BLKS and B6 mice under high-glucose conditions, and those with a strain effect P value of < 1 E-6 were used in the test. This test, for differences between regions, was executed for the entire genome and for each chromosome individually.
We would like to thank Dr Russ Cattley and Gwyneth Van for help with immunohistochemistry.
- Bell GI, Polonsky KS: Diabetes mellitus and genetically programmed defects in beta-cell function. Nature. 2001, 414: 788-791. 10.1038/414788a.View ArticlePubMedGoogle Scholar
- Stoehr JP, Nadler ST, Schueler KL, Rabaglia ME, Yandell BS, Metz SA, Attie AD: Genetic obesity unmasks nonlinear interactions between murine type 2 diabetes susceptibility loci. Diabetes. 2000, 49: 1946-1954. 10.2337/diabetes.49.11.1946.View ArticlePubMedGoogle Scholar
- Clee SM, Yandell BS, Schueler KM, Rabaglia ME, Richards OC, Raines SM, Kabara EA, Klass DM, Mui ET, Stapleton DS, Gray-Keller MP, Young MB, Stoehr JP, Lan H, Boronenkov I, Raess PW, Flowers MT, Attie AD: Positional cloning of Sorcs1, a type-2 diabetes quantitative trait locus. Nat Genet. 2006, 38: 688-693. 10.1038/ng1796.View ArticlePubMedGoogle Scholar
- Gapp DA, Leiter EH, Coleman DL, Schwizer RW: Temporal changes in pancreatic islet composition in C57BL/6J-db/db (diabetes) mice. Diabetologia. 1983, 25: 439-443. 10.1007/BF00282525.View ArticlePubMedGoogle Scholar
- Leiter EH, Coleman DL, Hummel KP: The influence of genetic background on the expression of mutations at the diabetes locus in the mouse. III. Effect of H-2 haplotype and sex. Diabetes. 1981, 30: 1029-1034. 10.2337/diabetes.30.12.1029.View ArticlePubMedGoogle Scholar
- Kooptiwut S, Zraika S, Thorburn AW, Dunlop ME, Darwiche R, Kay TW, Proietto J, Andrikopoulos S: Comparison of insulin secretory function in two mouse models with different susceptibility to beta-cell failure. Endocrinology. 2002, 143: 2085-2092. 10.1210/en.143.6.2085.PubMedGoogle Scholar
- Baetens D, Stefan Y, Ravazzola M, Malaisse-Lagae F, Coleman DL, Orci L: Alteration of islet cell populations in spontaneously diabetic mice. Diabetes. 1978, 27: 1-7. 10.2337/diabetes.27.1.1.View ArticlePubMedGoogle Scholar
- Kaku K, Province M, Permutt MA: Genetic analysis of obesity-induced diabetes associated with a limited capacity to synthesize insulin in C57BL/KS mice: evidence for polygenic control. Diabetologia. 1989, 32: 636-643.PubMedGoogle Scholar
- Davis RC, Schadt EE, Cervino AC, Peterfy M, Lusis AJ: Ultrafine mapping of SNPs from mouse strains C57BL/6J, DBA/2J, and C57BLKS/J for loci contributing to diabetes and atherosclerosis susceptibility. Diabetes. 2005, 54: 1191-1199. 10.2337/diabetes.54.4.1191.View ArticlePubMedGoogle Scholar
- Mao HZ, Roussos ET, Peterfy M: Genetic analysis of the diabetes-prone C57BLKS/J mouse strain reveals genetic contribution from multiple strains. Biochim Biophys Acta. 2006, 1762: 440-446.View ArticlePubMedGoogle Scholar
- Naggert JK, Mu JL, Frankel W, Bailey DW, Paigen B: Genomic analysis of the C57BL/Ks mouse strain. Mamm Genome. 1995, 6: 131-133. 10.1007/BF00303258.View ArticlePubMedGoogle Scholar
- Goren HJ, Kulkarni RN, Kahn CR: Glucose homeostasis and tissue transcript content of insulin signaling intermediates in four inbred strains of mice: C57BL/6, C57BLKS/6, DBA/2, and 129X1. Endocrinology. 2004, 145: 3307-3323. 10.1210/en.2003-1400.View ArticlePubMedGoogle Scholar
- Zraika S, Aston-Mourney K, Laybutt DR, Kebede M, Dunlop ME, Proietto J, Andrikopoulos S: The influence of genetic background on the induction of oxidative stress and impaired insulin secretion in mouse islets. Diabetologia. 2006, 49: 1254-1263. 10.1007/s00125-006-0212-9.View ArticlePubMedGoogle Scholar
- Hekerman P, Zeidler J, Korfmacher S, Bamberg-Lemper S, Knobelspies H, Zabeau L, Tavernier J, Becker W: Leptin induces inflammation-related genes in RINm5F insulinoma cells. BMC Mol Biol. 2007, 8: 41-10.1186/1471-2199-8-41.PubMed CentralView ArticlePubMedGoogle Scholar
- Homo-Delarche F, Calderari S, Irminger JC, Gangnerau MN, Coulaud J, Rickenbach K, Dolz M, Halban P, Portha B, Serradas P: Islet inflammation and fibrosis in a spontaneous model of type 2 diabetes, the GK rat. Diabetes. 2006, 55: 1625-1633. 10.2337/db05-1526.View ArticlePubMedGoogle Scholar
- Baekkeskov S, Aanstoot HJ, Christgau S, Reetz A, Solimena M, Cascalho M, Folli F, Richter-Olesen H, De Camilli P: Identification of the 64 K autoantigen in insulin-dependent diabetes as the GABA-synthesizing enzyme glutamic acid decarboxylase. Nature. 1990, 347: 151-156. 10.1038/347151a0.View ArticlePubMedGoogle Scholar
- Fanjul M, Renaud W, Merten M, Guy-Crotte O, Hollande E, Figarella C: Presence of MRP8 and MRP14 in pancreatic cell lines: differential expression and localization in CFPAC-1 cells. Am J Physiol. 1995, 268: C1241-1251.PubMedGoogle Scholar
- Schnekenburger J, Schick V, Kruger B, Manitz MP, Sorg C, Nacken W, Kerkhoff C, Kahlert A, Mayerle J, Domschke W, Lerch MM: The calcium-binding protein S100A9 is essential for pancreatic leukocyte infiltration and induces disruption of cell-cell contacts. J Cell Physiol. 2008, 216: 558-567. 10.1002/jcp.21433.View ArticlePubMedGoogle Scholar
- Ikemoto M, Matsumoto S, Egawa H, Okitsu T, Iwanaga Y, Umemoto S, Itoh H, Murayama H, Fujita M: A case with transient increases in serum S100A8/A9 levels implying acute inflammatory responses after pancreatic islet transplantation. Ann Clin Biochem. 2007, 44: 570-572. 10.1258/000456307782268156.View ArticlePubMedGoogle Scholar
- Salehi A, Meidute Abaraviciene S, Jimenez-Feltstrom J, Ostenson CG, Efendic S, Lundquist I: Excessive islet NO generation in type 2 diabetic GK rats coincides with abnormal hormone secretion and is counteracted by GLP-1. PLoS ONE. 2008, 3: e2165-10.1371/journal.pone.0002165.PubMed CentralView ArticlePubMedGoogle Scholar
- Chen J, Saxena G, Mungrue IN, Lusis AJ, Shalev A: Thioredoxin-interacting protein: a critical link between glucose toxicity and beta-cell apoptosis. Diabetes. 2008, 57: 938-944. 10.2337/db07-0715.PubMed CentralView ArticlePubMedGoogle Scholar
- Parikh H, Carlsson E, Chutkow WA, Johansson LE, Storgaard H, Poulsen P, Saxena R, Ladd C, Schulze PC, Mazzini MJ, Jensen CB, Krook A, Björnholm M, Tornqvist H, Zierath JR, Ridderstråle M, Altshuler D, Lee RT, Vaag A, Groop LC, Mootha VK: TXNIP regulates peripheral glucose metabolism in humans. PLoS Med. 2007, 4: e158-10.1371/journal.pmed.0040158.PubMed CentralView ArticlePubMedGoogle Scholar
- Benjamini Y, Hochberg Y: Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc Ser B. 1995, 57: 289-300.Google Scholar
- Yamaguchi H, Igarashi M, Hirata A, Tsuchiya H, Sugiyama K, Morita Y, Jimbu Y, Ohnuma H, Daimon M, Tominaga M, Kato T: Progression of diabetic nephropathy enhances the plasma osteopontin level in type 2 diabetic patients. Endocr J. 2004, 51: 499-504. 10.1507/endocrj.51.499.View ArticlePubMedGoogle Scholar
- Arafat HA, Katakam AK, Chipitsyna G, Gong Q, Vancha AR, Gabbeta J, Dafoe DC: Osteopontin protects the islets and beta-cells from interleukin-1 beta-mediated cytotoxicity through negative feedback regulation of nitric oxide. Endocrinology. 2007, 148: 575-584. 10.1210/en.2006-0970.View ArticlePubMedGoogle Scholar
- Johansson F, Kramer F, Barnhart S, Kanter JE, Vaisar T, Merrill RD, Geng L, Oka K, Chan L, Chait A, Heinecke JW, Bornfeldt KE: Type 1 diabetes promotes disruption of advanced atherosclerotic lesions in LDL receptor-deficient mice. Proc Natl Acad Sci USA. 2008, 105: 2082-2087. 10.1073/pnas.0709958105.PubMed CentralView ArticlePubMedGoogle Scholar
- Gunton JE, Kulkarni RN, Yim S, Okada T, Hawthorne WJ, Tseng YH, Roberson RS, Ricordi C, O'Connell PJ, Gonzalez FJ, Kahn CR: Loss of ARNT/HIF1beta mediates altered gene expression and pancreatic-islet dysfunction in human type 2 diabetes. Cell. 2005, 122: 337-349. 10.1016/j.cell.2005.05.027.View ArticlePubMedGoogle Scholar
- Fischer-Lindahl K: On naming H2 haplotypes: functional significance of MHC class Ib alleles. Immunogenetics. 1997, 46: 53-62. 10.1007/s002510050242.View ArticlePubMedGoogle Scholar
- Kamiya M, Judson H, Okazaki Y, Kusakabe M, Muramatsu M, Takada S, Takagi N, Arima T, Wake N, Kamimura K, Satomura K, Hermann R, Bonthron DT, Hayashizaki Y: The cell cycle control gene ZAC/PLAGL1 is imprinted – a strong candidate gene for transient neonatal diabetes. Hum Mol Genet. 2000, 9: 453-460. 10.1093/hmg/9.3.453.View ArticlePubMedGoogle Scholar
- Ma D, Shield JP, Dean W, Leclerc I, Knauf C, Burcelin RR, Rutter GA, Kelsey G: Impaired glucose homeostasis in transgenic mice expressing the human transient neonatal diabetes mellitus locus, TNDM. J Clin Invest. 2004, 114: 339-348.PubMed CentralView ArticlePubMedGoogle Scholar
- Abdollahi A: LOT1 (ZAC1/PLAGL1) and its family members: mechanisms and functions. J Cell Physiol. 2007, 210: 16-25. 10.1002/jcp.20835.View ArticlePubMedGoogle Scholar
- Toye AA, Lippiat JD, Proks P, Shimomura K, Bentley L, Hugill A, Mijat V, Goldsworthy M, Moir L, Haynes A, Quarterman J, Freeman HC, Ashcroft FM, Cox RD: A genetic and physiological study of impaired glucose homeostasis control in C57BL/6J mice. Diabetologia. 2005, 48: 675-686. 10.1007/s00125-005-1680-z.View ArticlePubMedGoogle Scholar
- Freeman H, Shimomura K, Cox RD, Ashcroft FM: Nicotinamide nucleotide transhydrogenase: a link between insulin secretion, glucose metabolism and oxidative stress. Biochem Soc Trans. 2006, 34: 806-810. 10.1042/BST0340806.View ArticlePubMedGoogle Scholar
- Freeman HC, Hugill A, Dear NT, Ashcroft FM, Cox RD: Deletion of nicotinamide nucleotide transhydrogenase: a new quantitative trait locus accounting for glucose intolerance in C57BL/6J mice. Diabetes. 2006, 55: 2153-2156. 10.2337/db06-0358.View ArticlePubMedGoogle Scholar
- Freeman H, Shimomura K, Horner E, Cox RD, Ashcroft FM: Nicotinamide nucleotide transhydrogenase: a key role in insulin secretion. Cell Metab. 2006, 3: 35-45. 10.1016/j.cmet.2005.10.008.View ArticlePubMedGoogle Scholar
- Aston-Mourney K, Wong N, Kebede M, Zraika S, Balmer L, McMahon JM, Fam BC, Favaloro J, Proietto J, Morahan G, Andrikopoulos S: Increased nicotinamide nucleotide transhydrogenase levels predispose to insulin hypersecretion in a mouse strain susceptible to diabetes. Diabetologia. 2007, 50: 2476-2485. 10.1007/s00125-007-0814-x.View ArticlePubMedGoogle Scholar
- Mu JL, Naggert JK, Svenson KL, Collin GB, Kim JH, McFarland C, Nishina PM, Levine DM, Williams KJ, Paigen B: Quantitative trait loci analysis for the differences in susceptibility to atherosclerosis and diabetes between inbred mouse strains C57BL/6J and C57BLKS/J. J Lipid Res. 1999, 40: 1328-1335.PubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.