Abstract
Purpose
Type-2 diabetes mellitus (T2D) is a metabolic disorder that can progress to a serious chronic disease and frequently develops in obese individuals in association with various pathogenic complications that shorten the lifespan of these patients. The liver is an important organ regulating lipid metabolism, which is damaged in both obesity and T2D; however, the specific pathways involved in these pathogenic effects remain unclear. Establishing a suitable animal model that effectively mimics the human biological condition is a critical factor to allow for precise identification of T2D-related genes.
Methods
The KK.Cg-Ay mouse strain is one such model that has offered insight into obesity-related T2D pathogenesis. To comprehensively assess the association between obesity and T2D, in the present study, we performed microarray analysis on liver tissue samples of KK.Cg-Ay and KK-α/α wild-type mice to examine differences in gene expression and methylation patterns and their related biological processes and pathways.
Results
We found that inflammation accompanied by abnormal lipid metabolism led to the spontaneous mechanism of obesity-induced diabetes, resulting in differential expression of some genes related to the terms of insulin resistance and glucose tolerance. Surprisingly, disruption of steroid biosynthesis strongly facilitated the diabetic pathogenesis. To support these findings, we highlighted some candidate genes and determined their relationships in biological networks of obesity-induced T2D.
Conclusion
These findings provide valuable reference data that can facilitate further detailed investigations to elucidate the pathogenic mechanism of obesity-induced diabetes in mice, which can be associated with the human condition to inform new prevention and treatment strategies.
Electronic supplementary material
The online version of this article (10.1007/s40200-019-00453-8) contains supplementary material, which is available to authorized users.
Keywords: Type-2 diabetes, Obesity, Liver metabolism, KK.Cg-Ay/J mouse, Microarray analysis
Introduction
The incidence of diabetes has been increasing worldwide in recent decades. The World Health Organization estimated that there were approximately 150 million patients with general diabetes in 2016 [1]. Diabetes is mainly classified into type 1, which is mostly diagnosed in children under 12 years old, and type 2, which is mainly prevalent in adults over 40 years of age. Type-2 diabetes mellitus (T2D) is much more common than type 1, and has shown a rapidly increasing trend in the USA along with an increasing rate of obesity, which has also broadened the age range of the disease [2, 3]. In particular, childhood obesity rates have dramatically increased between 1980 and 2010, with approximately 12.5 million children between 2 and 19 years old diagnosed with T2D during this time [4–6]. These trends support obesity as a risk factor of T2D development, along with the fact that more than 90% of obese adult patients are diagnosed with T2D in the USA [7]. Moreover, these prevalence estimates are considered to be reliable as they are based on modeling the worldwide population as an abundant genetic library in the USA, supported by global statistics for investigating the relation between obesity and diabetes, demonstrating that the number of obese patients is directly proportional to the number of T2D diagnoses, with dominance in the South American and western Asian populations [8].
To better understand this mechanistic link between the prevalence of obesity and T2D, genetic analysis is required to identify specifically regulated genes and the related ontological biological processes. Indeed, recent studies [9, 10] identified candidate genes that could serve as biomarkers to detect T2D. Moreover, to better understand the risk factors of the disease, several animal models have been developed to investigate changes in various tissues with the goal of elucidating the mechanism in the human condition [11–14]. In particular, gene modulation patterns have been evaluated in different organs of mouse models of obesity-induced T2D, and interactive networks have been established to grasp the underlying genetic and epigenetic mechanisms. In particular, the liver is the largest internal organ in mammals with multiple essential functions for detoxification, metabolism, and biochemical activities, including regulating hormone production, lipid metabolism, and controlling blood sugar. The liver has a substantial impact on regulating carbohydrate levels to release insulin from pancreatic cells [15–17]. Moreover, lipid metabolism, as one of the most essential processes in mammalian bodies, is mostly carried out in the liver. In the obese condition, fatty overload results in lipid accumulation in the liver, leading to metabolic disorder in both the liver and in other organs, especially with regard to abnormal glucose metabolism.
With the aim of further understanding the role of the liver in the pathogenesis of diabetes, in the present study, we performed a comparative genetic analysis of the liver tissues of KK-α/α wild-type (KK/HlJ or KK) mice without obesity and KK.Cg-Ay/J (KK-Ay or YKK) mice with obesity-induced T2D. YKK mice are heterozygous for the Ay mutation, which is lethal in homozygous form, but results in obesity within 8 weeks in the heterozygous form, ultimately leading to hyperglycemia, hyperinsulinemia, and glucose intolerance, resulting in diabetes due to polygenic causes. Therefore, the YKK mouse strain is a suitable model of diabetes in comparison with the wild-type KK strain as a control [18].
The consequence of insulin resistance and ineffective insulin production are the main check-points of T2D, and most of the organs of patients with diabetes ultimately comprise insulin-resistant cells, leading to a high concentration of non-use insulin. This in turn signals apoptosis pathways and activation of the immune system in pancreatic locations, thereby eliminating normal functional β-cells, resulting in an autoimmune reaction [19, 20]. These processes are repeatedly carried out by memorizing the faulty activity of nearby lymph nodes, which can explain why the insulin level recovers with liver transplantation [21] from an insulin-sensitive donor or by removing overactive lymph nodes instead of islet cells. Therefore, insulin resistance remains a critical issue in T2D, which is present in all abdominal organs such as the liver. In this study, we used the YKK mouse model to investigate the contribution of lipid metabolism to the insulin resistance of the liver in the pathogenesis of diabetes. Moreover, KK mice are susceptible to diabetes with increased risks during aging in obesity [22]. To examine the link between obesity and diabetes, we considered the YKK mouse strain as an obesity-induced diabetic model in comparison to its wild-type counterpart KK mice to investigate the factors contributing to the networks of biological disruption from obesity to diabetic phenotypes, by examining the status of gene expression and DNA methylation patterns at various stages of development using microarray analysis and functional annotation. This integrated analysis could identify differentially expressed genes to reveal candidate drivers of the transition from obesity to TDM. Concretely, these genes were classified into obesity-related, T2D-related, and intermediary, and a portion of the genes were further annotated with online databases to identify essential pathways that can provide novel hypotheses on T2D pathogenesis as guidance for new treatments and diagnostics. In addition to providing a valuable resource for further investigating the link between obesity and T2D in humans, differentially expressed genes in mice along with screening for diabetes candidate genes can further validate the utility of mouse models of obesity-induced diabetes that are similar to the pathogenesis of the human condition. Specifically, this work can provide insight into the spontaneous development of diabetes in obese KK.Cg-Ay/J mice without consideration of the role of early mutations.
Materials and methods
KK-α/α mice (control) and KK.Cg-Ay/J (diabetes) mice
Four male 4-week-old KK (control) and YKK (diabetic) mice were purchased from Jackson Laboratories (Bar Harbor, ME, USA) and grouped in specialized airy, transparent, hygienic plastic boxes according to type. The cages were placed in a sterile room with securely controlled temperature (22–25 °C), relative humidity (50–70%), and a natural photoperiod with nearly 12-h light/12-h dark. The mice were fed LabDiet 5 k52 (St. Louis, MO, USA) containing 19.3% protein, 13.2% total fat, and 4.3% fiber. All animal procedures were performed in the Animal Center of China Medical University (Taichung, Taiwan), and were approved by the Institutional Animal Care and Use Committee (IACUC) of China Medical University (IACUC: 102-217). The euthanasia was performed by the trained staff of the Animal Center with isoflurane (Baxter, UK) perfusion must be maintained until the heart stops. It has been reviewed by the council of Agriculture, Executive Yuan and the Animal Welfare Act using IACUC guidelines.
RNA extraction
The mixture of equally sliced liver tissues from the mice were weighed to 40 mg to extract approximately 40 μg of total RNA using RNeasy Mini Kit (Qiagen) according to the manufacturer instructions. The purity of RNA was determined in 10 mM Tris-HCl (pH 7.5) on a Nanodrop ND-1000 spectrometer (ND-1000; Labtech International, East Sussex, UK). The purified RNA was quantified to 300-ng samples and amplified by PCR. For microarray analysis, the samples were labeled with GeneChip WT Sense Target Labeling and Control Reagents (Affymetrix).
Microarray analysis: GeneChip expression 1.0 ST array
The arrays were scanned on an Affymetrix GeneChip® Scanner 3000 7G (Thermo Fisher Scientific). All of the reagents and buffers are contained in GeneChip® Hybridization, Wash, and Stain Kit (Affymetrix). The gene expression and promoter arrays were used proportionally in 1.0 ST and 1.0 R library of probes to retrieve sequence information related to specific genes. Expression analysis was performed by Transcriptome Analysis Console (TAC) of Affymetrix with the eBayes algorithm and the following filter options: fold change ≥1.2 and ≤ −1.2, P < 0.01 to enhance the enrichment of natural mutated effects, and fold change ≥1.5 and ≥ −1.5, P < 0.05 for finding the essential genes, whose modulation was influenced by mutations associated with the long-term development of diabetes.
Isolation of methylated DNA and whole-genome amplification
The MethylMiner Methylated DNA Enrichment Kit (Invitrogen) was used to isolate the methylated DNA. The MBD-magnetic beads conjugates were prepared with 3.5 mg of MBD-Biotin Protein coupled to 10 mL of Dynabeads M-280 Streptavidin according to the manufacturer’s instructions, which were then resuspended in one volume of bind/wash buffer after washing three times. The capture reaction was performed by adding 1 mg of sonicated DNA with 100–500 bp to the MBD magnetic beads on a rotating mixer for 1 h at room temperature in duplicate. Immunoprecipitated DNA and input DNA from MeDIP immunoprecipitations and MBD-Capture reactions were amplified with GenomePlex Single Cell Whole Genome Amplification Kit (Sigma), according to the manufacturer’s instructions. Fifty nanograms of DNA was used in each amplification reaction. The reactions were cleaned up using cDNA cleanup columns (Affymetrix), and 7.5 μg of amplified DNA was fragmented to be labeled using the Affymetrix Chromatin Immunoprecipitation Assay Protocol (P/N 702238 Rev. 3).
DNA methylation array
Hybridization was performed against the Affymetrix GeneChip Mouse Promoter 1.0R array for 17 h at 45 °C and 60 rpm. The arrays were subsequently washed (Affymetrix Fluidics Station 450) and stained with streptavidin-phycoerythrin (GeneChip® Hybridization, Wash, and Stain Kit, 900,720), and then scanned on Affymetrix GeneChip® Scanner 3000. The raw output data were processed with the Tiling Analysis Software of Affymetrix with the annotated library of the National Center for Biotechnology Information (NCBI) version 33. Differential DNA methylation analysis was performed with Partek software using default parameters.
Array database
The array data are accessible at the Gene Expression Omnibus (GEO) database of the NCBI (GSE120866; https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc= GSE120866). This dataset contains gene expression (GSE120864, https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE120864) and DNA methylation data (GSE120865, https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE120865).
Functional annotation
For classification and clustering of the functional ontology, the DAVID [23, 24] tool was used to investigate enrichment of differentially expressed genes in GO biological processes and KEGG pathways [25–27]. To construct the correlational networks of genes/proteins, Cytoscape [28] and Pathway Commons (http://www.pathwaycommons.org) [29] were used to retrieve the information from relevant research articles or other reliable sources. The linkage of murine to human studies was conducted with the MGI database [30, 31]. In addition, DisGeNET was used to determine the intersection of information in relationship analysis to gain insight into the biological mechanism through the relationship of gene groups and phenotypes expressed in human diseases.
Statistical analysis
Statistical analysis was conducted with SPSS software. All data used in parametric analyses are presented as the means ± standard error. SigmaPlot version 12.5 (Systat Software, Inc., San Jose, CA USA) was used for graphical presentation of the data as bar and scatter charts.
Results
Gene expression profiling reveals multiple biological functions in the diabetic pathogenesis of YKK mice
The transcriptome of the YKK (diabetic) and KK (control) mice was examined by microarray analysis to determine the fundamental molecular mechanisms of gene-induced diabetes. Because more enriched genes can support to determine the correlated networks of biological processes, differentially expressed genes were defined as those showing a fold-change in expression level of ≥1.2 or ≤ −1.2 with P < 0.05, and are illustrated in the heat map in Fig. 1a. In addition, the pattern of the total 1595 genes with differential expression across stages of diabetes development was visualized with a heat map organized in six expression cluster groups between the diabetes and control groups at 6, 16, and 42 weeks old. Phylogenetic analysis showed a similar expression pattern between the 6-week-old diabetes and control groups, and these differences became greater with age. There were 758, 2081, and 2450 up-regulated genes, and 677, 1670, and 2496 down-regulated genes detected in the comparisons of mice at 6, 16, and 42 weeks of age (Fig. 1b).
Fig. 1.
Overview of differentially expressed genes between KK.Cg-Ay/J (diabetic) and KK-α/α (control) mice at 6, 16, and 42 weeks old. a Heat map summarizing the overall clustering of significantly differentially expressed genes with levels indicated by different color bands from red (high) to blue (low). b Numbers of differentially expressed genes at each age of mice showing overlap indicating shared groups of genes
These genes were then classified into different functional types using the online tool Mouse Genome Informatics (MGI), as shown in S1 Table. To gain a deeper understanding of their biological function, the differentially expressed genes were classified into up- and down-regulated groups in each age group and subjected to Gene Ontology (GO) analysis with the DAVID tool (Fig. 2). The up-regulated genes at 6 weeks were abundantly enriched in terms related to the immune response to interferon-beta, −gamma and chemokines, and biosynthesis of cholesterol and sterol at 6 weeks old (Fig. 2a); de−/ubiquitination processes in protein, long-chain fatty acid metabolic process, and cell transport at 16 weeks old (Fig. 2c); and gene modulation as DNA methylation, nucleosome assembly, rDNA silencing, and immune response at 42 weeks old (Fig. 2e). By contrast, the downregulated genes were dominantly enriched in GO terms related to signal transduction, especially those linked to G protein coupled receptor, including the sensory perception of smell, at 6 weeks old (Fig. 2b); metabolism and biosynthesis of cholesterol, sterol, steroid, isoprenoid, and fatty acid at 16 weeks old (Fig. 2d); and redox processes, xenobiotic glucuronidation, and metabolic/biosynthetic processes in lipid, steroid, sterol, cholesterol, and flavonoid at 42 weeks old (Fig. 2f). In addition, S2 Table shows the top 20 enriched biological process GO terms and their significance levels.
Fig. 2.
Gene Ontology biological process (BP) terms enriched in differentially expressed genes between diabetic and control mice at different ages (fold change ≥1.2 or ≤ −1.2; P < 0.05) at 6 (a and b), 16 (c and d), and 42 (e and f) weeks old. The significance level of each BP was calculated with the [−log10(P value)], determined by DAVID
The related pathways of these differentially expressed genes were then analyzed by DAVID with the recently updated Kyoto Encyclopedia of Genes and Genomes (KEGG) database (https://www.genome.jp/kegg/pathway.html) (Fig. 3). The up-regulated genes were dominantly enriched in sphingolipid metabolism, cytokine/chemokine, and NF-κB signaling pathways at the 6-week-old stage (Fig. 3a); the activities of cascades, lysosome formation, insulin resistance, sphingolipid/FoxO/p53 signaling pathways, and retinol/inositol metabolism at the 16-week-old stage (Fig. 3c); and gene modulation of cancer cells, immunity activation, and NF-κ-B/MAPK/TNF signaling pathways at the 42-week-old stage (Fig. 3e). However, the down-regulated genes were significantly enriched in pathways related to olfactory transduction, transcription of cancer cells, and regulation of pluripotency at 6 weeks (Fig. 3b); steroid and terpenoid backbone biosynthesis, and metabolism of carbon, pyruvate, glutathione, glyoxylate, and dicarboxylate at 16 weeks (Fig. 3d); and carbon, ascorbate, aldarate, propanoate, retinol, and pyruvate metabolism, and biosynthesis of metabolites for antibiotics and enzymes at 42 weeks (Fig. 3f).
Fig. 3.
Enriched KEGG pathways (PWs) of the differentially expressed genes between diabetes and control mice (fold change ≥1.2 or ≤ −1.2; P < 0.05). These PWs were separated into up- and down-regulated genes at 6 weeks (a and b), 16 weeks (c and d), and 42 weeks (e and f) old. The significance level of each PW was calculated with [−log10(P value)], determined by DAVID
Overall, these results demonstrated significant gene expression profile differences between the diabetes and control mice, highlighting multiple biological processes and pathways that may contribute to the pathogenesis of the disease.
Global DNA methylation patterns in hepatocytes of diabetes and control mice
The distribution of differentially methylated regions (DMR) with respect to the transcription start site and the frequency of DMRs were investigated to obtain a general picture of the genomic landscape in the different chromosomes of the hepatocytes of the diabetic and control mice (Fig. 4). The candidate DMRs were identified based on the results of chromatin immunoprecipitation-microarray analysis to compare methylation levels of hepatocytic genes between diabetic (YKK) and control (KK) mice. A large portion of DMRs were detected in chromosomes 2, 7, and 11 along the 24 chromosomes according to the influenced gene density. In addition, the differentially methylated genes were cataloged according to gene classes using the online repository of MGI, as protein-coding genes, microRNA, or long intervening non-coding RNA (S4 Table). Overall, 8223, 5556, and 7634 DMRs were identified, 57.05%, 46.76%, and 65.06% of which were hypomethylated, and 42.95%, 53.24%, 34.94% of which were hypermethylated, at the 6-, 16-, and 42-week-old stages, respectively (Fig. 5). These results help to modulate the genome to determine the relationship of diabetes development with stages and specific chromosomes.
Fig. 4.
Differential methylated regions (DMRs) in mouse chromosomes, showing hyper- and hypomethylated genes identified at 6, 16, and 42 weeks based on chromatin immunoprecipitation-microarray analysis
Fig. 5.
Distribution of the hyper-methylated and hypomethylated differentially methylated regions (DMRs) in the DNA sequence based on the distance to the transcription start site (TSS) of the genes, defined in the promoter, gene body, and coding sequence (CDS) regions of the associated genes. These investigations were conducted in the mice at 6, 16, and 42 weeks old
Biological mechanism and function of DMRs in diabetic mice at 6, 16, and 42 weeks old
For functional annotation, we selected the identified DMRs that were also linked to the differentially expressed genes identified between the diabetic and control mice. In total, 201, 394, and 699 genes showed significant differences in methylation-dependent expression between the groups, including 109, 254, and 208 up-regulated genes with hypomethylation, and 92, 140, and 491 down-regulated genes with hypomethylation at 6, 16, and 42 weeks old. The roles of methylation in controlling the expression of genes were assessed according to enriched GO biological process terms (S1 Figure). The top 20 biological process terms for the associated genes, along with their significance levels, are listed in S5 Table. In addition, the pathways associated with these differentially methylation-dependent genes were analyzed in the recently updated KEGG database (S2 Figure), and the top 20 enriched pathways are listed in S6 Table. The analysis of methylation-dependent expression supports the essential impacts of methylation on gene regulation to contribute to the pathogenesis of diabetes with obesity.
Consistency in the methylation-dependent expression of genes through development
Although several genes showed methylation-dependent expression, some genes showed consistent trends of up-regulation with hypomethylation and down-regulation with hypermethylation throughout development at 6, 16, and 42 weeks old (Table 1). However, at 6 weeks old, the early development stage of the mouse, many genes showed incomplete expression. In addition, based on the differential expression analysis (Fig. 1), many of the genes showed similar expression levels between diabetic and control mice at this early stage. Therefore, the selection criteria for consistent genes included all the genes that showed similar trends at 16 and 42 weeks old, ignoring insignificant changes at 6 weeks old. The selected genes were functionally annotated with GO terms and their influence on abnormal traits was assessed with Mammalian Phenotypes (MP) annotation (S7 Table).
Table 1.
The consistent genes in methylation and expression during T2D development
| GENE EXPRESSION | DNA METHYLATION | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 6 weeks | 16 weeks | 42 weeks | 6 weeks | 16 weeks | 42 weeks | ||||||||
| Gene(s) | p value | Fold change | p value | Fold change | p value | Fold change | p value | MAT-score | p value | MAT-score | p value | MAT-score | |
|
Up gene regulation Hypo-methylation |
Gpcpd1 | 5.37E-06 | 2.43 | 3.34E-05 | 2.07 | 7.39E-03 | −3.09 | 7.15E-03 | −4.15 | ||||
| Phf20l1 | 5.00E-04 | 1.58 | 4.35E-02 | 1.22 | 4.68E-03 | −3.30 | 7.91E-03 | −4.09 | |||||
| Lurap1l | 6.00E-04 | 1.55 | 1.37E-02 | 1.31 | 9.50E-03 | −2.98 | 5.30E-05 | −6.98 | |||||
| Foxo3 | 1.00E-03 | 1.57 | 9.60E-03 | 1.35 | 6.80E-04 | −4.04 | 8.36E-03 | −4.06 | |||||
| Serpina3n | 1.80E-03 | 1.58 | 9.80E-05 | 1.96 | 8.36E-03 | −3.05 | 3.99E-03 | −4.50 | |||||
| Arl13b | 1.78E-02 | 1.67 | 6.40E-03 | 1.85 | 1.93E-03 | −3.64 | 9.90E-03 | −3.97 | |||||
|
Down gene regulation Hyper-methylation |
Eif2s2 | 1.00E-04 | −1.57 | 1.00E-04 | −1.57 | 6.92E-07 | −2.20 | 8.83E-06 | 14.31 | 2.23E-03 | 3.48 | 8.83E-06 | 18.30 |
| Dlc1 | 3.07E-02 | −1.31 | 2.11E-02 | −1.34 | 7.00E-04 | −1.65 | 2.39E-03 | 4.08 | 9.84E-03 | 2.86 | 5.30E-05 | 9.54 | |
| Il22ra1 | 1.82E-02 | −1.26 | 1.97E-02 | −1.25 | 3.40E-03 | −1.44 | 1.49E-03 | −4.04 | 4.22E-03 | 3.22 | 2.70E-03 | 5.46 | |
| Skp2 | 2.00E-04 | −1.76 | 2.52E-02 | −1.3 | 8.83E-06 | −7.43 | 4.62E-03 | 3.18 | 3.09E-04 | 7.18 | |||
| Mvd | 1.60E-03 | −1.85 | 1.53E-02 | −1.52 | 8.83E-06 | 7.43 | 4.81E-03 | 3.16 | 8.83E-06 | 13.09 | |||
| Nav2 | 3.20E-03 | −1.64 | 3.11E-02 | −1.38 | 1.10E-03 | −4.16 | 4.88E-03 | 3.15 | 6.18E-05 | 9.46 | |||
| Shank2 | 3.70E-03 | −1.37 | 1.71E-02 | −1.27 | 8.83E-06 | 8.10 | 4.65E-03 | 3.17 | 5.30E-05 | 9.57 | |||
| Celsr1 | 4.10E-03 | −1.34 | 3.13E-06 | −2.02 | 9.82E-03 | −3.19 | 3.89E-03 | 3.24 | 4.63E-03 | 5.05 | |||
| P4ha2 | 5.00E-03 | −1.32 | 4.08E-02 | −1.2 | 1.82E-03 | 4.21 | 6.85E-03 | 3.03 | 5.47E-04 | 6.64 | |||
| Pqlc1 | 6.20E-03 | −1.43 | 1.38E-02 | −1.36 | 4.41E-05 | 6.13 | 5.21E-04 | 4.03 | 1.85E-04 | 7.66 | |||
| Shroom3 | 7.10E-03 | −1.4 | 1.48E-02 | −1.34 | 8.83E-06 | 8.16 | 5.42E-03 | 3.11 | 8.83E-06 | 10.80 | |||
| Irak2 | 9.90E-03 | −1.35 | 7.00E-04 | −1.57 | 2.03E-04 | 5.10 | 4.46E-03 | 3.19 | 8.32E-03 | 4.63 | |||
| Zfhx3 | 1.42E-02 | −1.26 | 3.76E-02 | −1.21 | 1.44E-03 | 4.31 | 3.62E-04 | 4.17 | 7.06E-05 | 9.19 | |||
| Rin3 | 1.58E-02 | −1.34 | 5.20E-03 | −1.43 | 1.24E-04 | 5.32 | 3.27E-04 | 4.21 | 8.83E-06 | 13.06 | |||
| Nop16 | 2.81E-02 | −1.24 | 2.26E-02 | −1.25 | 4.41E-05 | 6.16 | 5.76E-03 | 3.09 | 4.90E-03 | 5.01 | |||
| Sdr9c7 | 3.02E-02 | −1.26 | 3.00E-04 | −1.64 | 7.86E-04 | −4.35 | 7.13E-03 | 3.00 | 1.03E-03 | 6.23 | |||
Highly abundant essential genes from the origin of the mutated checkpoint to the development of diabetic phenotypes
To further identify the most essential differentially expressed genes contributing to diabetes development, we focused on those showing consistent expression in the three investigated stages (6, 16, and 42 weeks) as the input mutated genes, which were related to the most abundant genes identified at each stage. Specifically, at 6 weeks, mutant traits were identified based on the mutated checkpoint. Genes modulated consistently by a mutation and express their characteristics through all developmental stages are considered to be “central genes”. The effects of such central genes spread to other genes to form gene networks, which can be identified using the Cytoscape and Pathway Common tools. To restrict the size of the networks based on the expression data, the differentially expressed genes identified at each stage were localized and were ensured to link with the central genes. A list of genes was obtained for each correlational network of each stage, which were assigned to essential genes based on the influence of mutation on expression; the summary of the procedure is schematically described in S3 Figure. To build interactive networks of the essential genes with expression data, the nodes of data-containing genes were selected as the genes belonging to both humans and mice. Therefore, the networks for 6-week-old (S4 Figure), 16-week-old (S5 Figure), and 42-week-old (S6 Figure) mice were constructed to provide an overview of the contribution of the essential genes to diabetic pathogenesis in mice and humans. In addition, the lists of genes of interest known as essential genes were dominantly connected to other genes to identify critical impacts (i.e., white nodes) of the developmental pathogenesis. The genes of the mice at 6, 16, and 42 weeks old were functionally annotated with DAVID and also classified into novel (non-reviewed) or studied (reviewed) in relation to diabetes research according to the DisGeNET database (Table 2). Overall, these analyses identified that the steroid pathway plays critical roles in the development of diabetes with high occurrence at 6, 16, and 42 weeks old. The list of essential genes with related pathways can offer a resource for understanding the disrupted mechanism of mutated genes related to diabetic phenotypes.
Table 2.
KEGG pathways found enriched with the involvement of essential genes by DAVID, assigned to 6, 16 and 42 weeks old. Also, these genes were annotated only in mouse determined association with Type-2 Diabetes or not, proportional to reviewed and unreviewed categories. The information is sourced from DisGeNet and other resources
| Age | KEGG pathway | Enrichment score [−log10(P value)] | Gene ID(s) (Non-Reviewed) | Gene ID(s) (Reviewed) |
|---|---|---|---|---|
| 6 weeks | PI3K-Akt signaling pathway | 1.3 | Foxo3, Il2rg, Prlr | Cdkn1a, Itga2 |
| Steroid biosynthesis | 1.1 | Fdft1, Sqle | ||
| HTLV-I infection | 1.0 | Fzd7, Il2rg | Cdkn1a, Vcam1 | |
| 16 weeks | Metabolic pathways | 10.0 | Nsdhl, St 3 gal5, Acot1, Acot2, Acot3, Asns, Crls1, Csad, Cyp2b9, Cyp2c38, Cyp21a1, Cyp4a10, Cyp4a14, Cyp4a31, Cyp4a32, Cyp51, Fdps, Hykk, Idi1, Ldhb, Lss, Mvd, Oat, Gm38481, Rdh11, Sqle, Sc5d, Uprt, Upp2 | Dhcr7, Acly, Acaca, Acacb, Cers6, Fasn, Msmo1, Nnmt, Pklr, Sds |
| Steroid biosynthesis | 7.6 | Nsdhl, Cyp51, LssSqle, Sc5d | Dhcr7, Msmo1 | |
| Biosynthesis of antibiotics | 7.5 | Nsdhl, Cyp51, Fdps, Idi1, Ldhb, Lss, Mvd, Oat, Gm38481, Sqle, Sc5d | Acly, Msmo1, Pklr, Sds | |
| Retinol metabolism | 3.5 | Cyp2b9, Cyp2c38, Cyp4a10, Cyp4a14, Cyp4a31, Cyp4a32, Rdh11 | ||
| Terpenoid backbone biosynthesis | 2.8 | Fdps, Idi1, Mvd, Gm38481 | ||
| Arachidonic acid metabolism | 2.7 | Cyp2b9, Cyp2c38, Cyp4a10, Cyp4a14, Cyp4a31, Cyp4a32 | ||
| Fatty acid elongation | 2.6 | Elovl6, Acot1, Acot2, Acot3 | ||
| Biosynthesis of unsaturated fatty acids | 2.6 | Elovl6, Acot1, Acot2, Acot3 | ||
| Pyruvate metabolism | 2.1 | Ldhb | Acaca, Acacb, Pklr | |
| Fatty acid biosynthesis | 2.1 | Acaca, Acacb, Fasn | ||
| Inflammatory mediator regulation of TRP channels | 2.0 | Cyp2c38, Cyp4a10, Cyp4a14, Cyp4a31, Cyp4a32 | Il1r1 | |
| Steroid hormone biosynthesis | 1.9 | Akr1c18, Cyp2b9, Cyp2c38, Cyp21a1, Sult1e1 | ||
| Fatty acid degradation | 1.9 | Cyp4a10, Cyp4a14, Cyp4a31, Cyp4a32 | ||
| NF-kappa B signaling pathway | 1.7 | Bcl2l1, Cd14, Il1r1, Ly96, Vcam1 | ||
| Propanoate metabolism | 1.5 | Ldhb | Acaca, Acacb | |
| PPAR signaling pathway | 1.3 | Cyp4a10, Cyp4a14, Cyp4a31, Cyp4a32 | ||
| Glucagon signaling pathway | 1.1 | Ldhb, Sik1 | Acaca, Acacb | |
| Fatty acid metabolism | 1.0 | Elovl6 | Acaca, Fasn | |
| 42 weeks | Metabolic pathways | 10.3 | Hmgcs1, Nsdhl, Acot1, Acot2, Asns, Crls1, Csad, Cyp2b9, Cyp2c38, Cyp21a1, Cyp4a10, Cyp4a14, Cyp4a31, Cyp4a32, Cyp51, Dpys, Fdft1, Fdps, Hykk, Idi1, Ldhb, Lss, Mvd, Oat, Gm38481, Rdh11, Sqle, Sc5d, Uprt, Upp2 | Dhcr7, Acly, Acaca, Acacb, Cers6, Fasn, Msmo1, Nnmt, Pklr, Sds |
| Steroid biosynthesis | 9.3 | Nsdhl, Cyp51, Fdft1, Lss, Sqle, Sc5d | Dhcr7, Msmo1 | |
| Biosynthesis of antibiotics | 9.2 | Hmgcs1, Nsdhl, Cyp51, Fdft1, Fdps, Idi1, Ldhb, Lss, Mvd, Oat, Gm38481, Sqle, Sc5d | Acly, Msmo1, Pklr, Sds | |
| Terpenoid backbone biosynthesis | 4.0 | Hmgcs1, Fdps, Idi1, Mvd, Gm38481 | ||
| Retinol metabolism | 3.5 | Cyp2b9, Cyp2c38, Cyp4a10, Cyp4a14, Cyp4a31, Cyp4a32, Rdh11 | ||
| Arachidonic acid metabolism | 2.6 | Cyp2b9, Cyp2c38, Cyp4a10, Cyp4a14, Cyp4a31, Cyp4a32 | ||
| Pyruvate metabolism | 2.1 | Ldhb | Acaca, Acacb, Pklr | |
| Fatty acid biosynthesis | 2.0 | Acaca, Acacb, Fasn | ||
| Inflammatory mediator regulation of TRP channels | 2.0 | Cyp2c38, Cyp4a10, Cyp4a14, Cyp4a31, Cyp4a32, Il1r1 | ||
| Steroid hormone biosynthesis | 1.9 | Akr1c18, Cyp2b9, Cyp2c38, Cyp21a1, Sult1e1 | ||
| Fatty acid degradation | 1.8 | Cyp4a10, Cyp4a14, Cyp4a31, Cyp4a32 | ||
| Fatty acid elongation | 1.5 | Elovl6, Acot1, Acot2 | ||
| Propanoate metabolism | 1.5 | Ldhb | Acaca, Acacb | |
| Biosynthesis of unsaturated fatty acids | 1.5 | Elovl6, Acot1, Acot2 | ||
| PPAR signaling pathway | 1.3 | Cyp4a10, Cyp4a14, Cyp4a31, Cyp4a32 | ||
| NF-kappa B signaling pathway | 1.1 | Bcl2l1, Cd14, Il1r1, Ly96 | ||
| Glucagon signaling pathway | 1.0 | Ldhb, Sik1 | Acaca, Acacb |
Discussion
With the aim of gaining a better understanding of the pathogenic mechanisms of diabetes in YKK mice, we sought to assemble the fragmented information of altered biological processes and pathways, which can be logically summarized into the pre, early, and late stages of diabetes development, equivalent to 6, 16, and 42 weeks of age in the mouse. At the pre-disease stage (6 weeks), inflammation is believed to trigger the fasting-state biosynthesis of cholesterols to steroid hormones such as glucocorticoids, which immediately reduces the stress of immune activation in the liver. Autoimmunity is a key factor of diabetes, which involves long-term activation of the immune system leading to symptoms of pre-diabetes, which are found in non-obese diabetic mice as well as in humans [32, 33]. The crucial roles of the inflammatory processes at this stage were confirmed in the present study given the high activities of NF-κB and chemokine signaling pathways observed in YKK mice, as a known model of obesity. To respond to this activated immune system, G protein receptors reduce signaling transduction and cellular sensitivity to insulin, leading to hyperinsulinemia. In the next stage (early disease stage, 16 weeks), owing to the stressful condition, cellular metabolism might be continuously affected by steroid hormones such as glucocorticoids in response to the inflammation [34–37], influencing the survival of hepatocytes by activation of Akt/Foxo pathways and sphingolipid metabolism. In addition, ubiquitination occurs to degrade unnecessary proteins and other compounds, as fatty acids are used for catabolism to supply energy to cells instead of glucose to reduce the stress, leading to autophagy [38, 39]. Although this “self-eating” process ultimately decreases lipid, sterol, steroid, and cholesterol metabolism, stress is elevated owing to the development of insulin resistance at this stage, consequently disrupting the activities of metabolic pathways such as carbon/pyruvate metabolism and glycolysis/gluconeogenesis. Finally, after suffering this insufficient energy supply in a state of glucose intolerance, the hepatocytes experience uncontrolled methylation and modulation of gene expression through profoundly inducing MAPK signaling, leading to dysregulated cell proliferation under the stressful condition. Although the immune systems are activated via the NF-κB signaling pathways [40–43], at this stage, infectious and cancerous factors cannot be effectively defended against owing to deficiency of drug metabolism and antibiotic activities.
Overall, our study provides insight into the mechanisms of diabetic pathogenesis in the mouse model. The proposed mechanism is schematically described in Fig. 6, highlighting the essential factors for the development of diabetes in obesity, including cholesterol biosynthesis and autoimmunity. The cell automatically adjusts to stressful conditions by enhancing its membrane with stabilization and resistance, which incidentally interrupts normal biological processes while reducing the effects of external signals [44–46]. We found enhanced synthesis of cholesterols and sphingolipids in the diabetic mice at 6 and 16 weeks of age, which are involved in cell signaling and communication. This suggests that the activities of signal transduction and motility of membrane molecules as substrates and supportive complexes might be delayed to interfere with standard functions as long-term effects. Based on this model, we also propose a clinical treatment strategy, which was developed with the combination of hospital records for patients with T2D under long-term observation and the insight into the pathogenic mechanism of diabetes revealed in this study (S7 Figure).
Fig. 6.
Hypotheses of the connection between T2D and obesity based on our integrated analysis of gene expression and methylation
The methylation analysis indicated that most of the methylated positions were in the gene body region. In addition, there were more hypomethylated genes than hypermethylated genes identified at the 16-week-old stage, while the opposite pattern was observed at the 6- and 42-week-old stages. This finding suggests that methylation impacts gene modulation differently at each stage. Although methylation at the gene promoter directly represses transcription [47, 48], the coding regions of genes can also be methylated, which enhances gene expression; however, the regulation of these genes is suppressed when the levels of methylation exceed the balancing limits [49, 50]. Thus, methylation of the gene body might substantially control the levels of gene transcripts instead of promoters. Therefore, more detailed investigations on methylation of the gene body are warranted to better understand its contribution to diabetes development. Furthermore, analysis of the distribution of DMRs among murine chromosomes was performed to determine which chromosomes dominantly contribute to diabetes pathogenesis. As expected, the hypermethylated levels at the 42-week-old stage in each chromosome were higher than those of other stages due to the main consequences of diabetes and aging. In summary, methylation of the gene body appears to be essential to regulate the expression of genes related to murine diabetes, which might also indicate the pathogenic mechanism in the human condition.
Moreover, the analysis of the essential genes highlighted central pathways with important contributions to promoting the pathogenesis of diabetic with other related traits, collectively known as obesity. The interactive networks demonstrated the spreading relationships from mutated genes to other interacting genes, which were determined with the assistance of Cytoscape software and the Pathways Common database. The central genes emerging in all stages of development were Raly, Ralyl, Eif2s2, Sik1, Hnrnpc, and Hnrnpcl1/2, which were down-regulated for the most part. As expected, the inducing mutation disrupted the structural non-agouti gene (Raly), which confers a completely yellow coat color, and even more so, the susceptibility to diabetes. Some other essential genes such as Foxo3, Cdkn1a, Hspa8, Bcl2l1, Usmg5, Tfcp2, Map3k5, Fasn, Gadd45g, and Tsc22d1 were found to interact significantly with many genes that influenced the expression profiles at each stage, and these genes show extensive influences of other genes and biological processes, as listed in Table 2. Similar to the detailed analysis of gene expression and methylation patterns, given the early development stage of 6 weeks old, the different types of cells were mainly differentiated in spite of limited growth, but there was still a significant change in gene modulation between the diabetic and control groups of the mice. From the beginning of inducing mutations to the age of 6 weeks, the genes showing differential expression result in complex interactions of mutated and normal genes to ultimately influence murine biological activities, including those related to PI3K/Akt pathways of cell survival, steroid production of fasting fat metabolism, and HTLV infection related to immune system responses. The interactive network was then expanded in connection with other candidate genes, which were mostly related to similar KEGG pathways such as metabolic pathways, steroid synthesis, and biosynthesis of antibiotics at 16 and 42 weeks old. Surprisingly, the production of steroids consistently emerged as an enriched pathway, and also had a high number of associated genes in diabetic development. This mutation drove most of the changed pathways, including synthesis of lipid products, and metabolism of fat and pyruvate, relating to the PPAR and glucagon signaling pathways. Notably, the results demonstrated an impact on regulation of inflammation in the NF-κB pathways and steroid hormones to influence the autoimmune reaction in diabetic pathogenesis.
In conclusion, using the YKK mouse model of obesity-induced diabetes, we demonstrated that mutation of the Ay gene modulates other genes to interrupt related biological pathways. In particular, we highlighted a surprising finding in that the production of steroid hormones and inflammatory activities play important roles in the pathogenesis of diabetes. However, the interpretation of these results is limited by recent knowledge of this field and limitations of the findings in this particular animal model. Therefore, the results of this study should be interpreted in light of accumulating research to provide a reasonable explanation and elucidate the detailed mechanism with clinical translation potential.
Electronic supplementary material
(DOCX 17 kb)
(DOCX 35 kb)
(DOCX 35 kb)
(DOCX 15 kb)
(DOCX 29 kb)
(DOCX 17 kb)
(DOCX 22 kb)
(PNG 271 kb)
(PNG 162 kb)
(PNG 69 kb)
(PNG 1553 kb)
(PNG 2172 kb)
(PNG 2025 kb)
(PNG 45 kb)
Acknowledgments
We are grateful to staffs within the research and clinical teams at Genetic Center, China Medical University Hospital for help in obtaining and processing samples for this research.
Funding
This work was supported by grants from China Medical University Hospital in Taiwan (DMR-107-048 and DMR-108-121.
Compliance with ethical standards
Conflict of interest
All authors declare that they have no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Contributor Information
Shih-Yin Chen, Email: chenshihy@gmail.com.
Fuu-Jen Tsai, Email: d10704@mail.cmuh.org.tw.
References
- 1.Association AD Screening for type 2 diabetes. Diabetes Care. 2004;27:S11. doi: 10.2337/diacare.27.2007.S11. [DOI] [PubMed] [Google Scholar]
- 2.Control CfD . Prevention. National diabetes statistics report, 2017. Atlanta: Centers for Disease Control and Prevention; 2017. [Google Scholar]
- 3.Benjamin EJ, Virani SS, Callaway CW, Chamberlain AM, Chang AR, Cheng S, et al. Heart disease and stroke Statistics-2018 update: a report from the American Heart Association. Circulation. 2018;137(12):e67–e492. doi: 10.1161/CIR.0000000000000558. [DOI] [PubMed] [Google Scholar]
- 4.Organization WH. Global strategy on diet, physical activity and health. 2004.
- 5.Ogden CL, Carroll MD, Lawman HG, Fryar CD, Kruszon-Moran D, Kit BK, et al. Trends in obesity prevalence among children and adolescents in the United States, 1988-1994 through 2013-2014. JAMA. 2016;315(21):2292–2299. doi: 10.1001/jama.2016.6361. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Association AD Type 2 diabetes in children and adolescents. Pediatrics. 2000;105(3):671–680. doi: 10.1542/peds.105.3.671. [DOI] [PubMed] [Google Scholar]
- 7.Organization WH. World Health Organization obesity and overweight fact sheet. 2016.
- 8.Atlas ID. Brussels: International Diabetes Federation; 2011. International Diabetes Federation 2017.
- 9.Lees T, Nassif N, Simpson A, Shad-Kaneez F, Martiniello-Wilks R, Lin Y, et al. Recent advances in molecular biomarkers for diabetes mellitus: a systematic review. Biomarkers. 2017;22(7):604–613. doi: 10.1080/1354750X.2017.1279216. [DOI] [PubMed] [Google Scholar]
- 10.Wang L, Weinshilboum R. Metformin pharmacogenomics: biomarkers to mechanisms. Diabetes. 2014;63(8):2609–2610. doi: 10.2337/db14-0609. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Chiu HK, Tsai EC, Juneja R, Stoever J, Brooks-Worrell B, Goel A, et al. Equivalent insulin resistance in latent autoimmune diabetes in adults (LADA) and type 2 diabetic patients. Diabetes Res Clin Pract. 2007;77(2):237–244. doi: 10.1016/j.diabres.2006.12.013. [DOI] [PubMed] [Google Scholar]
- 12.Hemminki K, Liu X, Forsti A, Sundquist J, Sundquist K, Ji J. Subsequent type 2 diabetes in patients with autoimmune disease. Sci Rep. 2015;5:13871. doi: 10.1038/srep13871. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Tsai S, Clemente-Casares X, Revelo XS, Winer S, Winer DA. Are obesity-related insulin resistance and type 2 diabetes autoimmune diseases? Diabetes. 2015;64(6):1886–1897. doi: 10.2337/db14-1488. [DOI] [PubMed] [Google Scholar]
- 14.Joglekar MV, Januszewski AS, Jenkins AJ, Hardikar AA. Circulating microRNA biomarkers of diabetic retinopathy. Diabetes. 2016;65(1):22–24. doi: 10.2337/dbi15-0028. [DOI] [PubMed] [Google Scholar]
- 15.Basu R, Barosa C, Jones J, Dube S, Carter R, Basu A, et al. Pathogenesis of prediabetes: role of the liver in isolated fasting hyperglycemia and combined fasting and postprandial hyperglycemia. J Clin Endocrinol Metab. 2013;98(3):E409–E417. doi: 10.1210/jc.2012-3056. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Sharabi K, Tavares CD, Rines AK, Puigserver P. Molecular pathophysiology of hepatic glucose production. Mol Asp Med. 2015;46:21–33. doi: 10.1016/j.mam.2015.09.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Hatting M, Tavares CDJ, Sharabi K, Rines AK, Puigserver P. Insulin regulation of gluconeogenesis. Ann N Y Acad Sci. 2018;1411(1):21–35. doi: 10.1111/nyas.13435. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Iwatsuka H, Shino A, Suzuoki Z. General survey of diabetic features of yellow KK mice. Endocrinol Jpn. 1970;17(1):23–35. doi: 10.1507/endocrj1954.17.23. [DOI] [PubMed] [Google Scholar]
- 19.Donath MY, Størling J, Maedler K, Mandrup-Poulsen T. Inflammatory mediators and islet β-cell failure: a link between type 1 and type 2 diabetes. J Mol Med. 2003;81(8):455–470. doi: 10.1007/s00109-003-0450-y. [DOI] [PubMed] [Google Scholar]
- 20.Cnop M, Welsh N, Jonas JC, Jorns A, Lenzen S, Eizirik DL. Mechanisms of pancreatic beta-cell death in type 1 and type 2 diabetes: many differences, few similarities. Diabetes. 2005;54(Suppl 2):S97–107. doi: 10.2337/diabetes.54.suppl_2.s97. [DOI] [PubMed] [Google Scholar]
- 21.Kim S, Kim J, Lee K, Park T, Baek H, Yu H, et al., editors. Effect of liver transplantation on glucose levels in patients with prediabetes or type 2 diabetes. Transplantation proceedings: Elsevier; 2014. [DOI] [PubMed]
- 22.Ikeda H. KK mouse. Diabetes Res Clin Pract. 1994;24(Suppl):S313–S316. doi: 10.1016/0168-8227(94)90268-2. [DOI] [PubMed] [Google Scholar]
- 23.Pruitt KD, Tatusova T, Maglott DR. NCBI reference sequences (RefSeq): a curated non-redundant sequence database of genomes, transcripts and proteins. Nucleic Acids Res. 2007;35(Database issue):D61–D65. doi: 10.1093/nar/gkl842. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Barrett T, Wilhite SE, Ledoux P, Evangelista C, Kim IF, Tomashevsky M, et al. NCBI GEO: archive for functional genomics data sets—update. Nucleic Acids Res. 2012;41(D1):D991–D9D5. doi: 10.1093/nar/gks1193. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Kanehisa M, Goto S. KEGG: Kyoto encyclopedia of genes and genomes. Nucleic Acids Res. 2000;28(1):27–30. doi: 10.1093/nar/28.1.27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Kanehisa M, Sato Y, Kawashima M, Furumichi M, Tanabe M. KEGG as a reference resource for gene and protein annotation. Nucleic Acids Res. 2015;44(D1):D457–DD62. doi: 10.1093/nar/gkv1070. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Kanehisa M, Furumichi M, Tanabe M, Sato Y, Morishima K. KEGG: new perspectives on genomes, pathways, diseases and drugs. Nucleic Acids Res. 2017;45(D1):D353–DD61. doi: 10.1093/nar/gkw1092. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003;13(11):2498–2504. doi: 10.1101/gr.1239303. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Cerami EG, Gross BE, Demir E, Rodchenkov I, Babur O, Anwar N, et al. Pathway commons, a web resource for biological pathway data. Nucleic Acids Res. 2011;39(Database issue):D685–D690. doi: 10.1093/nar/gkq1039. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Finger JH, Smith CM, Hayamizu TF, McCright IJ, Xu J, Law M, et al. The mouse gene expression database (GXD): 2017 update. Nucleic Acids Res. 2017;45(D1):D730–D7D6. doi: 10.1093/nar/gkw1073. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Smith CL, Blake JA, Kadin JA, Richardson JE, Bult CJ, Group MGD Mouse genome database (MGD)-2018: knowledgebase for the laboratory mouse. Nucleic Acids Res. 2017;46(D1):D836–DD42. doi: 10.1093/nar/gkx1006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Kaminitz A, Yolcu ES, Stein J, Yaniv I, Shirwan H, Askenasy N. Killer Treg restore immune homeostasis and suppress autoimmune diabetes in prediabetic NOD mice. J Autoimmun. 2011;37(1):39–47. doi: 10.1016/j.jaut.2011.03.003. [DOI] [PubMed] [Google Scholar]
- 33.Grossmann V, Schmitt VH, Zeller T, Panova-Noeva M, Schulz A, Laubert-Reh D, et al. Profile of the immune and inflammatory response in individuals with prediabetes and type 2 diabetes. Diabetes Care. 2015;38(7):1356–1364. doi: 10.2337/dc14-3008. [DOI] [PubMed] [Google Scholar]
- 34.Gryglewski RJ. Steroid hormones, anti-inflammatory steroids and prostaglandins. Pharmacol Res Commun. 1976;8(4):337–348. doi: 10.1016/0031-6989(76)90034-5. [DOI] [PubMed] [Google Scholar]
- 35.Barnes PJ, Adcock I. Anti-inflammatory actions of steroids: molecular mechanisms. Trends Pharmacol Sci. 1993;14(12):436–441. doi: 10.1016/0165-6147(93)90184-L. [DOI] [PubMed] [Google Scholar]
- 36.Coutinho AE, Chapman KE. The anti-inflammatory and immunosuppressive effects of glucocorticoids, recent developments and mechanistic insights. Mol Cell Endocrinol. 2011;335(1):2–13. doi: 10.1016/j.mce.2010.04.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Cruz-Topete D, Cidlowski JA. One hormone, two actions: anti- and pro-inflammatory effects of glucocorticoids. Neuroimmunomodulation. 2015;22(1-2):20–32. doi: 10.1159/000362724. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Kaur J, Debnath J. Autophagy at the crossroads of catabolism and anabolism. Nat Rev Mol Cell Biol. 2015;16(8):461. doi: 10.1038/nrm4024. [DOI] [PubMed] [Google Scholar]
- 39.Khaminets A, Behl C, Dikic I. Ubiquitin-dependent and independent signals in selective autophagy. Trends Cell Biol. 2016;26(1):6–16. doi: 10.1016/j.tcb.2015.08.010. [DOI] [PubMed] [Google Scholar]
- 40.Baeuerle PA, Henkel T. Function and activation of NF-kappaB in the immune system. Annu Rev Immunol. 1994;12(1):141–179. doi: 10.1146/annurev.iy.12.040194.001041. [DOI] [PubMed] [Google Scholar]
- 41.Liang Y, Zhou Y, Shen P. NF-kappaB and its regulation on the immune system. Cell Mol Immunol. 2004;1(5):343–350. [PubMed] [Google Scholar]
- 42.Liu T, Zhang L, Joo D, Sun S-C. NF-κB signaling in inflammation. Signal Transd Target Ther. 2017;2:17023. doi: 10.1038/sigtrans.2017.23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Taniguchi K, Karin M. NF-κB, inflammation, immunity and cancer: coming of age. Nat Rev Immunol. 2018;18(5):309. doi: 10.1038/nri.2017.142. [DOI] [PubMed] [Google Scholar]
- 44.Simons K, Toomre D. Lipid rafts and signal transduction. Nat Rev Mol Cell Biol. 2000;1(1):31–39. doi: 10.1038/35036052. [DOI] [PubMed] [Google Scholar]
- 45.Head BP, Patel HH, Insel PA. Interaction of membrane/lipid rafts with the cytoskeleton: impact on signaling and function: membrane/lipid rafts, mediators of cytoskeletal arrangement and cell signaling. Biochim Biophys Acta. 2014;1838(2):532–545. doi: 10.1016/j.bbamem.2013.07.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Fonseca MC, Franca A, Florentino RM, Fonseca RC, Lima Filho ACM, Vidigal PTV, et al. Cholesterol-enriched membrane microdomains are needed for insulin signaling and proliferation in hepatic cells. Am J Physiol Gastrointest Liver Physiol. 2018;315(1):G80–G94. doi: 10.1152/ajpgi.00008.2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Li E, Zhang Y. DNA methylation in mammals. Cold Spring Harb Perspect Biol. 2014;6(5):a019133. doi: 10.1101/cshperspect.a019133. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Schübeler D. Function and information content of DNA methylation. Nature. 2015;517(7534):321. doi: 10.1038/nature14192. [DOI] [PubMed] [Google Scholar]
- 49.Jjingo D, Conley AB, Yi SV, Lunyak VV, Jordan IK. On the presence and role of human gene-body DNA methylation. Oncotarget. 2012;3(4):462–474. doi: 10.18632/oncotarget.497. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Yang X, Han H, De Carvalho DD, Lay FD, Jones PA, Liang G. Gene body methylation can alter gene expression and is a therapeutic target in cancer. Cancer Cell. 2014;26(4):577–590. doi: 10.1016/j.ccr.2014.07.028. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
(DOCX 17 kb)
(DOCX 35 kb)
(DOCX 35 kb)
(DOCX 15 kb)
(DOCX 29 kb)
(DOCX 17 kb)
(DOCX 22 kb)
(PNG 271 kb)
(PNG 162 kb)
(PNG 69 kb)
(PNG 1553 kb)
(PNG 2172 kb)
(PNG 2025 kb)
(PNG 45 kb)