- Email: [email protected]
hyperinsulinemic conditions in 3T3 cells
Biochimie 94 (2012) 2656e2664
Contents lists available at SciVerse ScienceDirect
Biochimie journal homepage: www.elsevier.com/locate/biochi
Research paper
Histone H3 lysine 4 monomethylation (H3K4me1) and H3 lysine 9 monomethylation (H3K9me1): Distribution and their association in regulating gene expression under hyperglycaemic/hyperinsulinemic conditions in 3T3 cells Jeena Gupta a, Sandeep Kumar a, Juntao Li b, R. Krishna Murthy Karuturi b, Kulbhushan Tikoo a, * a
Laboratory of Chromatin Biology, Department of Pharmacology and Toxicology, National Institute of Pharmaceutical Education and Research (NIPER), Mohali, Punjab, India Computational and Systems Biology, Genome Institute of Singapore, A*STAR (Agency for Science, Technology and Research), 60 Biopolis St, S138672, Singapore, Republic of Singapore
b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 21 December 2011 Accepted 2 August 2012 Available online 21 August 2012
Hyperglycemia/hyperinsulinemia are leading cause for the induction type 2 diabetes and the role of posttranslational histone modifications in dysregulating the expression of genes has emerged as potential important contributor in the progression of disease. The paradoxical nature of histone H3-Lysine 4 and Lysine 9 mono-methylation (H3K4me1 and H3K9me1) in both gene activation and repression motivated us to elucidate the functional relationship of these histone modifications in regulating expression of genes under hyperglycaemic/hyperinsulinemic condition. Chromatin immunoprecipitationemicroarray analysis (ChIP-chip) was performed with H3 acetylation, H3K4me1 and H3K9me1 antibody. CLUSTER analysis of ChIP-chip (Chromatin immunoprecipitationemicroarray analysis) data showed that mRNA expression and H3 acetylation/H3K4me1 levels on genes were inversely correlated with H3K9me1 levels on the transcribed regions, after 30 min of insulin stimulation under hyperglycaemic condition. Interestingly, we provide first evidence regarding regulation of histone de/acetylases and de/methylases; Myst4, Jmjd2b, Aof1 and Set by H3Ac, H3K4me1 and H3K9me1 under hyperinsulinemic/hyperglycaemic condition. ChIPeqPCR analysis shows association of increased H3Ac/H3K4me1 and decreased levels of H3K9me1 in up regulation of Myst4, Jmjd2, Set and Aof1 genes. We further analyse promoter occupancy of histone modifications by ChIP walking and observed increased occupancy of H3Ac/H3K4me1 on promoter region (1000 to 1) of active genes and H3K9me1 on inactive genes under hyperglycemic/ hyperinsulinemic condition. To best of our knowledge this is the first report that shows regulation of chromatin remodelling genes by alteration in the occupancy of histone H3Ac/H3K4/K9me on both promoter and transcribed regions. Ó 2012 Elsevier Masson SAS. All rights reserved.
Keywords: Hyperglycemia Insulin Histone modifications Chromatin remodelling genes Chromatin immunoprecipitation and diabetes
1. Introduction Long standing models have suggested the importance of histone modifications in altering the chromatin structure and regulating the histoneeDNA/histoneehistone interactions [1,2]. These histoneeDNA interactions set a critical stage for the regulation of the transcription of genes by influencing the DNA accessibility to the cell’s transcriptional machinery elements [3]. The transcriptional regulation of gene expression is a dynamic process and mediates the conversion of inactive heterochromatin structure to
* Corresponding author. Department of Pharmacology and Toxicology, National Institute of Pharmaceutical Education and Research (NIPER), Sector 67, S.A.S. Nagar, Mohali, Punjab-160 062, India. Tel.: þ91 172 2214682 87; fax: þ91 172 2214692. E-mail address: [email protected] (K. Tikoo). 0300-9084/$ e see front matter Ó 2012 Elsevier Masson SAS. All rights reserved. http://dx.doi.org/10.1016/j.biochi.2012.08.011
active euchromatin structure and vice versa. The elaborate collection of posttranslational modifications of histone proteins including acetylation, phosphorylation, methylation, ubiquitination and ADPribosylation, acts like a molecular code, also known as “histone code” and provides a platform to accommodate the binding of factors regulating the chromatin function [4,5]. Therefore, aberrant alterations in these epigenetic modifications (that do not involve a change in DNA sequence), can change chromatin structure and thus could lead to the deregulated gene transcription and the disease progression. Among the modifications of different histones (H2A, H2B, H3 and H4), the modifications of histone H3 are most studied and appreciated in the regulation of gene transcription [6,7]. Evidences show that the acetylation of key lysine residues of histone H3 is generally associated with gene activation [8]. However H3 lysine methylation displays a high degree of complexity as methylation of
J. Gupta et al. / Biochimie 94 (2012) 2656e2664
lysines 4, 36 and 79 of histone H3 primarily dictate gene activation whereas, methylation at lysines 9, 27, 20 are most commonly described at sites of gene repression [9,10]. But certain methylation marks can have multiple functions as H3K9, a methylation mark commonly enriched at the heterochromatin, has also been found to be present at the transcribed regions of active genes [11,12]. In addition, there are other mechanisms also by which an individual site of histone modifications can carry differential regulation of gene expression. These include the degree of processivity (mono, di or tri functional group) and the spacial distribution of these histone modifications on either the promoter region or the transcribed region of the gene. Reports have shown that in Saccharomyces cerevisiae, H3 lysine 4 dimethylation (H3K4me2) and H3 lysine 4 trimethylation (H3K4me3) have been shown to peak in the 50 transcribed region of active genes, implicating them in the early phases of the transcriptional elongation [13,14]whereas, H3 lysine 4 monomethylation (H3K4me1) is more characterized on the enhancer region of the gene promoters [15]. In contrast some studies have suggested the association of H3K4me1 with silenced euchromatin [16,17]. In Arabidopsis, another histone H3 modification dimethylation of lysine 9 (H3K9me2) is found to be a critical mark for gene silencing but not monomethylation (H3K9me1) or trimethylation (H3K9me3) [18]. However, there are reports that show the enrichment of H3K9me3 at heterochromatin region in mammalian cells, while H3K9me1 and H3K9me2 are localized to the euchromatin region of silenced genes [19]. Therefore distribution of these distinct histone H3 modifications; K4me and K9me remain unclear. Thus underlying the paradoxical nature of these histone modifications, we carried out the present study to delineate the functional properties of H3K4 and H3K9 monomethylation marks and also to study their association with H3 acetylation, a well known gene activation mark. Reports have shown that, among the different mechanisms that could lead to induction of diabetes, the role of epigenetic mechanisms in dysregulating the expression of genes, has emerged as a potential important contributor in the last decade [20]. Several studies have revealed interesting insights into the relationship between di-methylation and tri-methylation states of H3K4/H3K9 and type 2 diabetes [21,22] but none of the study has demonstrated the functional significance of mono-methylated states of H3K4/ H3K9 underlying diabetic condition (hyperglycemia/hyperinsulinemia). Adipocytes being the important contributor in the induction of the lifestyle disorders; obesity and type 2 diabetes, so in the present study we try to mimic the diabetic conditions by culturing 3T3 adipocytes under high glucose and high insulin condition [23,24]. Our aim was to functionally characterize H3K4/ H3K9 monomethylation with the alteration in gene expression in 3T3 cells under hyperglycemic/hyperinsulinemic conditions.
2657
USA). All the other chemicals were purchased from Sigma (St. Louis, MO, USA), unless otherwise mentioned. 2.2. Cell culture and treatment 3T3 cells were cultured in DMEM supplemented with 10% foetal bovine serum and antibiotics (penicillin 100 IU/ml and streptomycin 100 mg/ml) in 5% CO2 incubator at 37 C. At 80e90% confluency, cells were treated with either low glucose (LG) (5 mM þ 20 mM mannitol, for osmotic balance) or high glucose (HG) (25 mM) for 5 h in serum-free DMEM, followed by insulin stimulation (100 nM) for 30 min (denoted by LGI or HGI respectively). Cells were then washed twice with ice cold phosphate buffer saline (PBS). 2.3. Isolation of total RNA 3T3 cells (80e90% confluence) were treated with low glucose (5 mM glucose þ 20 mM mannitol) and high glucose (25 mM glucose) for 5 h and stimulated with insulin (100 nM) for 30 min. At the end of the treatment, the cells were washed with PBS and the total RNA was isolated using a protocol as described by Chomczynski and Sacchi [41] using TRIZOL reagent (Invitrogen) and was purified using RNeasy kit (Auprep RNeasy mini kit, Life Technologies) according to the manufacturer’s protocol. The RNA quality and integrity from each sample were assured using Nanodrop (ND-1000) by A260/280 absorbance ratio and using agarose gel electrophoresis respectively. The total RNA prepared from 3T3 cells was used for the quantitative PCR (qPCR) and the cDNA microarray analysis. 2.4. Probe labelling and hybridization The mouse 15K array (Microarray centre, University Health Care, Toronto) used in the present study consisted of 15,264 genes spotted in duplicate. The fluorescence labelled cDNA probes were prepared from 20 mg of total RNA using a SuperScript III (Invitrogen Life Technologies), oligo dT primer (Invitrogen Life Technologies) and Cy3 or Cy5 labelled dCTPs (Amersham biosciences) from the low glucose and the high glucose treated 3T3 cells stimulated with insulin for 30 min. The Cy3 and Cy5 probes were then mixed in an equal amount and the hybridization was carried out at 43 C for 16 h on Hyb Array 12 hybridization station (Perkin Elmer). The hybridization was repeated three times. A dye-swapped experiment was also performed to improve the accuracy of the measurement and to rule out the nonspecific signals. 2.5. Analysis of microarray data
2. Methods 2.1. Chemicals 3T3 cell line was purchased from American Type Cell Culture (ATCC). Cell culture media, antibiotic solution, foetal calf serum, and trypsineEDTA solution were purchased from GIBCO (USA). Insulin was purchased from Novartis Pharma and CM-DCF-DA from Molecular Probes (Invitrogen USA). ECL detection kit and ECL hyperfilm were obtained from Amersham Bioscience (USA). Anti-pH3 (Ser-10) antibody and horseradish peroxidase (HRP)-conjugated secondary antibody were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Anti-acetylated H3, anti-monomethyl histone H3-K4, anti-monomethyl histone H3-K9 and anti histone H3 were purchased from Upstate Biotechnology (Lake Placid, NY,
Slides were scanned with a Scanarray Gx microarray scanner (Perkin Elmer) and the images were analysed using Scanarray software. The flagged spots were excluded from the analysis. The mean intensity for each spot was taken and the log 2 value for the signal intensity of each spot was calculated for each slide. Genes with FG (foreground) >1.5 BG (Background) were retained for further analysis. The SAM (Significance Analysis of Microarrays) [25] in conjunction with SLR (stepped linear regression) [26] was used for finding significantly differentially expressed or histone modified genes with FDR (false discovery rate) <10%. The genes showing more than 1.5 fold change were taken for further analysis. Genecard (www.genecards.org) and Pubmed (www.pubmed.gov) were used for assigning the genes for specific biological processes. For clustering, we have selected the differentially expressed genes with change in any one of histone modification and cluster these
2658
J. Gupta et al. / Biochimie 94 (2012) 2656e2664
genes on the basis of correlation coefficient by Cluster and Treeview analysis (http://rana.lbl.gov/EisenSoftware.htm). 2.6. Reverse transcriptase polymerase chain reaction (RT-qPCR) To confirm the cDNA microarray, qPCR was performed for certain genes. The first strand cDNA was generated from 20 mg of total RNA by using a SuperScript III (Invitrogen, Life Technologies) and oligo dT primer (Invitrogen, Life Technologies) in a 30 ml reaction mixture. The quantitative real time PCR reaction was performed using LightCycler 2.0 (Roche Diagnostics) in a final volume of 20 ml containing 50 ng of cDNA, 4 ml of reaction buffer from LightCycler FastStart DNA master plus kit (Roche Diagnostics) and specific forward and reverse primers (Midland Certified Reagent Company Inc.). After the amplification, a melting curve analysis was performed to verify the specificity of the reaction. The analysis was performed using LightCycler software (Roche Diagnostics). The relative gene expression was assessed using the comparative Ct (DCt) method and normalized to GAPDH. 2.7. ChIP-on-chip (chromatin immunoprecipitation and microarray hybridization) ChIPs were performed by using MAGnify Chromatin immunoprecipitation kit (Invitrogen, Life Technologies). Briefly, the insulinstimulated cells under low glucose and high glucose condition were washed with PBS after the indicated time interval and were crosslinked using 1% formaldehyde for 20 min. Cross-linking was stopped by the addition of 0.1 M glycine and the cell lysates were sonicated. Consistency of equal levels of sonication of the cell lysates was maintained and before immunoprecipitation, all samples were tested for the level of DNA fragmentation, with the mean size of DNA fragments was always kept 500 bp. Immunoprecipitation with the indicated antibodies or preimmune mouse IgG (as control) was carried out for 2 h with magnetic beads. Input (20 ml of sample separated before immunoprecipitation) and immunoprecipitates were washed, eluted, and then crosslinking was reversed. The DNA fragments were recovered using magnetic beads provided with kit and was amplified for specific genes by qPCR. No DNA controls were always included; making sure that the primeredimer formation was not detectable in the experimental samples while performing the qPCR. Relative fold change was assessed using the comparative Ct (DCt) method and normalized to input. Experiments were typically repeated three times; the error bars in the figures show the standard deviations. For ChIP-chip analysis labelling of the immunoprecipitated DNA was performed using the Bioprime Labelling system (Invitrogen, Life Technologies). Labelled ChIP/Cy5 and Cy3 DNAs were combined and hybridized to the array as already described above. 3. Results 3.1. Profiling histone modifications across coding regions of genes under high glucose condition To identify the differential expression of genes under hyperglycemic/hyperinsulinemic condition we performed cDNA microarray analysis after insulin stimulation of 30 min under high glucose condition. SAM analysis in conjugation with Stepped Linear Regression analysis [26] was performed to filter differentially expressed genes with at least 1.5 fold difference and FDR <10%. With this we identified 1301 genes which show change in expression in HGI (high glucose þ insulin) as compared to LGI (low glucose þ insulin) condition (Online Appendix 1). To get further insight into the level of H3K9me1, H3K4me1 and H3Ac across the
coding regions of the mouse genome, we performed ChIPecDNA analysis using 15 K cDNA array after 30 min of the insulin stimulation under both low and high glucose condition. To detect the difference in these histone modifications between high glucose and low glucose condition, the profiles were determined as normalized ratios HGI/LGI (fold change). Applying the above mentioned statistical analysis and criterion we identified 1168 targets for H3Ac (Online appendix 2), 148 targets for H3K4me1 (Online appendix 3) and 546 targets for H3K9me1 (Online appendix 4) with differential status in HGI as compared to LGI condition in coding regions of the genes. However there were only 12 common genes between these three modifications on cDNA array. To understand the role of these histone H3 modifications in regulation of the genes under hyperglycemic/hyperinsulinemic conditions, we identified the genes that underwent changes in any of these three histone modifications along with change in their gene expression levels. To do so, we set up a criterion and select only the genes that were common in cDNA expression analysis with differential change in status in any one of either H3Ac or H3K4me1 or H3K9me1. This stringent criterion might result in false negatives but it also reduces the number of genes to a manageable size for further validation analysis and the chance of having false positives. With this criterion we identified 831 genes with significant differential H3Ac or H3K4me1 or H3K9me1 status and also change in their mRNA expression levels (Online appendix 5). Of these 608 genes were down regulated and 223 genes were up regulated. To evaluate the major effected biological process in the cells under hyperglycemia/hyperinsulinemia, we further categorized these genes to the various functional categories and observed that most of the biological processes were down regulated (Fig. 1). Also we identified that signal transduction; transcription and protein metabolism were among the major effected pathway. We further validate the change in mRNA expression and H3Ac and H3K9me1 levels on insulin signalling genes which were known to play a role in diabetes, Gapvd1, Mapk6, Ctbp1 and Nfat by qPCR and ChIPeqPCR analysis respectively and observed similar levels of change (Fig. 2). These results demonstrated that cells exposed to hyperglycemia and hyperinsulinemia (pathological conditions of type 2 diabetes), disturb most of the biologic pathways of the cell, mainly hampering the signal transduction elements by altering histone H3 modifications. 3.2. Correlation between H3Ac, H3K4me1 and H3K9me1 status with mRNA expression levels Further to delineate the functional relationship of H3Ac, H3K4me1 and H3K9me1 with each other and with mRNA expression of the genes, we performed hierarchical clustering of the genes using Pearson correlation coefficient metric on the above selected dataset (candidate genes with significant difference in any of the three modifications and having differential mRNA expression level) using publicly available software packages CLUSTER and TREEVIEW (http://rana.lbl.gov/EisenSoftware.htm) (Fig. 3). With this analysis we demonstrate that histone H3Ac levels in the coding regions of the genes very well correlate with the mRNA expression level of the respective genes signifying H3Ac as a mark of gene activation in the coding regions of the genes. Furthermore, mRNA expressions of most of the genes were inversely proportional to H3K9me1 levels, suggesting that increased H3K9me1 occupancy in the coding regions of the genes is associated with gene inactivation. However, very less number of genes was enriched for H3K4me1 in the coding regions and we also failed to observe much overlap between H3K4me1 and mRNA expression levels (Fig. 3). These results indicate increased occupancy of H3Ac and H3K9me1 in the coding region of the genes and not for H3K4me1.
J. Gupta et al. / Biochimie 94 (2012) 2656e2664
2659
Fig. 1. cDNA microarray and ChIP-chip analysis compares the genes regulated by the insulin (100 nM) stimulation under high glucose as compared to low glucose conditions: Bar diagram shows differentially expressed genes towards the various functional pathways as analysed by the microarray analysis along with changes in any of these three histone modifications (H3Ac, H3K4me1 or H3K9me1) as analysed by ChIP-chip analysis after 30 min of insulin stimulation under high glucose as compared to the low glucose condition. Data shows statistically significant change in gene expression with FDR <10% and fold change >1.5. The figure is derived from the Supplementary data (Online appendices 1e5) and assigned to the corresponding biological process as described in the “Materials and methods”.
3.3. Hyperglycemia/hyperinsulinemia induced changes in expression of chromatin modifying genes and their regulation by histone modifications Out of differentially expressed genes identified by cDNA microarray and ChIP-chip analysis, we observed significant change in the expression of genes that are responsible for mediating chromatin remodelling by insulin under high glucose condition. These include down regulation of Myst4 and Ep400 (histone acetyltransferases, HAT), Jmjd2b and Jarid2 (histone methyl-transferases,
HMT) and Dyrk2 (histone kinase). In addition to above mentioned genes, Brdt gene which is involved in reorganization of acetylated chromatin was also found to be down regulated. Increase in the expression of Set gene (HAT inhibitor) and also genes responsible for histone H3K4 de-methylation (Jarid1a and Aof1) further supports our earlier observation. The change in expression of these genes observed in the present study was in accordance with our previous findings that show decrease in levels of H3Ac, H3K4me1 and H3K9me1 after 30 min of insulin stimulation under high glucose condition [27].
Fig. 2. Change in gene expression, histone H3 acetylation and lysine 9 monomethylation levels on coding regions of the insulin signalling genes: 3T3 cells were cultured as described in the “Materials and methods”. For ChIPeqPCR cells were cross linked with 1% formaldehyde and the ChIP was performed with respective antibody. Bar graph shows change in the mRNA expression, H3Ac and H3K9me1 levels on a: Gapvd1; b: Mapk6; c: Ctbp1 and d: Nfat, where LGI represents cells cultured under low glucose condition and treated with insulin for 30 min and HGI represents cells cultured under high glucose condition and treated with insulin for 30 min. Relative fold change was calculated after normalization with mouse 18S gene and input for qPCR and ChIPeqPCR respectively. Similar results were obtained in the three independent sets of experiments. All the values were represented as mean S.E.M. (n ¼ 3), ***p < 0.001, **p < 0.01 and *p < 0.05, Vs LGI.
2660
J. Gupta et al. / Biochimie 94 (2012) 2656e2664
Fig. 3. Hierarchical cluster analysis of mRNA, H3Ac, H3K4me1 and H3K9me1 profiles on coding regions of genes altered by the insulin (100 nM) stimulation under high glucose as compared to low glucose conditions: genes were selected for <10% FDR and >1.5 fold enrichment in ChIP DNA under high glucose as compared to low glucose after 30 min of insulin stimulation. Lane 1 shows mRNA expression profile, lane 2: H3Ac profile, lane 3: H3K4me1 profile and lane 4 shows H3K9me1 profile, where red indicates increased mRNA expression or increased H3Ac/H3K4me1/H3K9me1 levels, green indicates decreased mRNA expression or decreased H3Ac/H3K4me1/H3K9me1 levels and black indicates no change. Intensity of colour correlates to the magnitude of change. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Out of the above nine chromatin remodelling genes that show expression change in our microarray data, four genes were selected for further validation by qPCR. The four selected chromatin remodelling genes were each from different class of histone modifications, i.e., Myst4 is basically known to be a histone acetyltransferase, Set is a HAT inhibitor, Jmjd2b is a histone demethylase (specifically demethylate H3K9) and Aof1 is a histone H3K4 demethylase [34e37]. Similar changes in expression of these genes as observed by both microarray and qPCR analysis (Fig. 4) further prompted us to confirm the change in H3Ac, H3K4me1 and H3K9me1 levels on their coding regions by performing ChIPeqPCR analysis (Fig. 5). We observed a decrease in the level of H3Ac on Myst4 and Jmjd2b and an increase on Set and Aof1 genes confirming our ChIP-chip data. However, we failed to observe any change in H3K9me1 levels on the coding regions of histone H3K9 demethylase (Jmjd2b) and H3K4 demethylase (Aof1). Decreased H3K4me1 levels on Myst4 and Jmjd2b and increased H3K4me1 levels on Set and Aof1 further confirmed our ChIP-chip analysis (Fig. 5). These results suggest the presence of crosstalk between these histone modifications under hyperinsulinemic/hyperglycemic conditions. However, levels of H3K9me1 were only changed on histone
Fig. 4. Change in gene expression of chromatin remodelling genes: the fold change in mRNA expression levels of chromatin remodelling genes determined by microarray analysis was further validated by qPCR using a light cycler 2.0 on RNA extracted from 3T3 cells cultured as described in the “Materials and methods”. Bar graphs show change in the mRNA expression of the histone modification regulating genes; Jmjd2b, Aof1, Myst4 and Set under the high glucose condition (HGI) as compared to the low glucose condition (LGI) after insulin treatment of 30 min, a: by microarray analysis and b: by qPCR analysis. Relative fold change was calculated after normalization with mouse 18S gene for qPCR analysis. Similar results were obtained in the three independent sets of experiments. All the values were represented as mean S.E.M. (n ¼ 3), ***p < 0.001, **p < 0.01 and *p < 0.05, Vs LGI.
acetylase (Myst4) and deacetylase (Set), highlighting the role of this modification in regulating histone acetylation only. 3.4. Relative abundance of histone modifications on promoters of chromatin remodelling genes Studies have shown that histone modifications adopt characteristic and specific distribution for regulating gene expression, some of which occur in the vicinity of promoters, while others occur across the coding regions of the genes. Hence, to get further insight into the functional role of H3Ac, H3K4me1 and H3K9me1 in regulating gene expression by their presence/absence on the promoter region of the differentially expressed genes, we performed ChIP-walking. We focused mainly on the regulation of promoter regions (nt 2000 to 1 from transcription start site) of the chromatin remodelling genes Myst4, Jmjd2b, Set and Aof1. Our data shows similar pattern of change for H3Ac and H3K4me1 on Myst4, Jmjd2b, Aof1 and Set gene promoter regions under both low glucose and high glucose condition after insulin stimulation. Comparing the relative abundance of these histone modifications, we observed an increased occurrence of H3Ac and H3K4me1 between regions 1000 to 500 nt on Myst4 and Jmjd2b promoter under low glucose condition which on the other hand get decreased under high glucose condition. H3K9me1 shows an opposite trend suggesting the specific distribution of these histone modifications
J. Gupta et al. / Biochimie 94 (2012) 2656e2664
2661
Fig. 5. Histone H3 acetylation, lysine 4 monomethylation and lysine 9 monomethylation levels on coding regions of the chromatin modification regulating genes: 3T3 cells were cultured as described in the “Materials and methods”. Bar graph shows H3Ac, H3K4me1 and H3K9me1 levels on a: Myst4; b: Set; c: Jmjd2b and d: Aof1, where LGI represents cells cultured under low glucose condition and treated with insulin for 30 min and HGI represents cells cultured under high glucose condition and treated with insulin for 30 min. Relative fold change was calculated after normalization with input. Similar results were obtained in the three independent sets of experiments. All the values were represented as mean S.E.M. (n ¼ 3), ***p < 0.001, **p < 0.01 and *p < 0.05, Vs LGI.
on the promoter regions of histone acetylase Myst4 (Fig. 6) and histone demethylase Jmjd2b (Fig. 7). The decreased mRNA expression of Myst4 and Jmjd2b under high glucose as observed by qPCR (see Section 3.4), further emphasizes the role of H3Ac and H3K4me1 in gene activation and H3K9me1 in gene silencing even on the promoter region of the chromatin remodelling genes. Histone deacetylase Set and histone demethylase Aof1 promoter regions showed increased H3Ac and H3K4me1 at regions 800 to 300 nt and 1000 to 600 nt respectively under high glucose condition as compared to low glucose condition. However, H3K9me1 methylation got decreased at these regions (Figs. 6 and 7). Comparison of these results with the increased mRNA expression of Set and Aof1 under high glucose condition further confirms the significance of H3K9me1 as gene inactivation mark at the promoter region of these genes. To check the specificity of our ChIP assays, we also performed ChIPs with mouse IgGs and did not observe any significant amplification. 4. Discussion Despite the knowledge that high glucose alters multiple histone modifications and associated gene expression [28,29] including the sustainable epigenetic changes by long term memory of chronic exposure to high glucose [30], the impact of chronic insulin treatment under high glucose condition on histone monomethylation status has not been previously explored on a genome wide scale. By profiling of histone modifications with alterations in gene expression under hyperglycemic/hyperinsulinemic condition (mimicking diabetic condition), we for the first time show the differential regulation of gene expression by H3Ac, H3K4me1 and H3K9me1. CLUSTER analysis of our ChIPecDNA microarray data demonstrates direct correlation between H3Ac levels and mRNA expression levels suggesting H3Ac as mark of gene activation in accordance with earlier reports [8]. However, the low frequency of occurrence of H3K4me1 on the coding regions of the transcriptionally up regulated genes as observed by ChIPecDNA microarray
analysis, suggests that monomethylation of H3K4 may not be always required for transcriptional elongation of active genes by RNA polymerase, as established previously for H3K4 di/tri methylation [14,31]. van Dijk et al. have shown H3K4 monomethylation as a mark for silenced euchromatin in Chlamydomonas [16]. However our data shows an increase in relative abundance of H3K4me1 along with H3Ac on the promoters (region between 0 and 1000) of the chromatin remodelling genes (Myst4, Set, Jmjd2b and Aof1) under hyperglycemic/hyperinsulinemic condition. Increase in relative abundance of H3Ac and H3K4me1 matches very well with the increased level of mRNA expression of these genes. Our data is in agreement with earlier report by Lin et al., which shows higher abundance of H3K4 monomethylation on the enhancer regions of active genes in B cell development [15]. Thus our results emphasize the role of H3Ac and H3K4me1 in regulating the expression of chromatin remodelling genes under diabetic condition by altering its relative abundance on the promoter region of these genes. Apart from histone H3Ac and H3K4me1 we further studied the role of other histone modification (H3K9me1) which may act as a mark or signature for inactive/active chromatin. Correlating the levels of H3K9me1 with mRNA expression of the genes by CLUSTER analysis depicts that most of the genes with increased occupancy of H3K9me1 as well as decreased H3 acetylation levels across the coding regions were down regulated. Earlier reports have also shown the association of H3K9 mono/di/tri methylation with transcriptionally silent chromatin [18,32]. But we for the first time provide a global profile of H3K9me1 across the transcribed regions of the genes under hyperglycemic/hyperinsulinemic condition. Rice et al. have shown that promoter regions of silenced genes in euchromatin are largely mono or di-methylated at histone H3K9 residue, whereas pericentric heterochromatin is predominantly trimethylated at H3K9 residue [33]. Consistent with this report we also provide evidence that there is an increased distribution of H3K9me1 across the promoter region (1 to 1000 bp) of the chromatin remodelling genes down regulated under hyperinsulinemic/hyperglycemic condition. Collectively these results
2662
J. Gupta et al. / Biochimie 94 (2012) 2656e2664
Fig. 6. Relative abundance of histone modifications on promoter of histone de/acetylases Myst4 and Set: 3T3 cells were cultured as described in the “Materials and methods”. qPCR was performed across the promoter region of Myst4 and Set. All distances are relative to transcription start site. Plot shows H3Ac, H3K4me1 and H3K9me1 levels on a: Myst4 promoter under low glucose condition after 30 min of insulin stimulation, b: Myst4 promoter under high glucose condition after 30 min of insulin stimulation, c: Set promoter under low glucose condition after 30 min of insulin stimulation and d: Set promoter under high glucose condition after 30 min of insulin stimulation. Similar results were obtained in the three independent sets of experiments.
highlight the role of histone H3K9me1 in gene inactivation on both promoter as well as transcribed regions. Interestingly our data also shows the cross-talk between the genes involved in altering histone H3 modifications (Fig. 8). Acetylation and methylation of histones are highly dynamic processes and are mediated by various enzymes called histone de/ acetylases and de/methylases. We here for the first time report the regulation of histone de/acetylases and de/methylases; Myst4, Jmjd2b, Aof1 and Set by H3Ac, H3K4me1 and H3K9me1 under hyperinsulinemic/hyperglycemic condition. Functionally, Myst4 is a histone acetyl transferase, Set is an inhibitor of histone acetylation, Jmjd2b is a histone demethylase (specifically demethylate H3K9) and Aof1 is a histone H3K4 demethylase [34e37]. The decreased mRNA expression of Myst4 and Jmjd2b and increased expression of Aof1 and Set under hyperinsulinemic/hyperglycemic condition support our previous findings that there is global down regulation of H3 acetylation/H3K4me1 levels and an increase in H3K9me1 levels both in vitro and in vivo [27,38e40]. Furthermore, ChIP-chip and ChIPeqPCR data demonstrates the role of histone modifications (H3Ac, H3K4me1 and H3K9me1) in regulating the expression of genes responsible for chromatin remodelling, suggesting existence of a cross-talk between different histone modifications. Further, comparing the relative abundance of H3Ac, H3K4me1 and H3K9me1 across the promoter region of
chromatin remodelling genes, between the high glucose and low glucose condition after insulin treatment, reveals a characteristic pattern for these histone modifications. Interestingly, similar pattern of H3Ac and H3K4me1 was observed at the promoter regions of the genes responsible for histone modifications, but H3K9me1 showed an opposite pattern. The characteristic increase in the H3Ac and H3K4me1 and decrease in H3K9me1 on the promoter region (1 to 1000 bp) from the transcription start site for the up-regulated genes Aof1 and Set and vice-versa for the down-regulated genes Myst4 and Jmjd2b, suggest the critical role of these histone modifications in regulating the levels of these genes under hyperglycemic/hyperinsulinemic conditions. These findings highlight the role of histone modifications (H3Ac, H3K4me1 and H3K9me1) in regulating expression of these genes by altering its relative abundance at their promoter regions. Further, comparison of the gene expression related to transcriptional category by coclustering reveals the crosstalk between different histone modifications on the promoter region of these genes. In most of the down regulated genes, decreased pattern of H3k4me1 and H3Ac was found, whereas increased H3K9me1 abundance was associated with decreased mRNA level. Only for a small set of genes, H3k9me1 abundance was associated with increased mRNA expression (Supplementary Fig. 1).
J. Gupta et al. / Biochimie 94 (2012) 2656e2664
2663
Fig. 7. Relative abundance of histone modifications on promoter of histone de/methylases Jmjd2b and Aof1: 3T3 cells were cultured as described in the “Materials and methods”. qPCR was performed across the promoter region of Jmjd2b and Aof1. All distances are relative to transcription start site. Plot shows H3Ac, H3K4me1 and H3K9me1 levels on a: Jmjd2b promoter under low glucose condition after 30 min of insulin stimulation, b: Jmjd2b promoter under high glucose condition after 30 min of insulin stimulation, c: Aof1 promoter under low glucose condition after 30 min of insulin stimulation and d: Aof1 promoter under high glucose condition after 30 min of insulin stimulation. Similar results were obtained in the three independent sets of experiments.
In summary, our work highlights the changes in H3K4me1 and H3K9me1 on the transcribed regions of the genes after providing diabetes related insults and also their relative abundance across the promoter region of the genes responsible for altering histone
modifications. Moreover, we also provide evidence for the presence of cross-talk between different histone modifications under diabetic conditions. The understanding of these epigenetic alterations may therefore lead to additional therapeutic strategies.
I
High Glucose Condition
I
Plasma Membrane IR 30 minutes
H3 A Acetylation t l ti H3K4 de/methylase
H3 de/acetylase
H3K9 H3K9me
H3K4 H3K4me
Nucleus
H3K9 de/methylase
Chromatin compaction
Altered gene expression
Fig. 8. Flowchart summarizing the cross-talk between insulin-induced epigenetic modifications under hyperglycemic condition.
2664
J. Gupta et al. / Biochimie 94 (2012) 2656e2664
Author contributions J.G. researched data, wrote manuscript. S.K. contributed in gene co-clustering and preparing gene association network. J.L. helps in data analysis. R.K.M. contributed to discussion, reviewed/edited manuscript. K.T. contributed to discussion, reviewed/edited manuscript and provides final approval of the version to be published. Duality of interest The authors declare that no conflict of interest exists with the publication of this work. Acknowledgements This work was supported in part by a grant from National Institute of Pharmaceutical Education and Research (NIPER) and a grant from Department of Biotechnology, Government of India. Mrs. Jeena Gupta was supported by Senior Research Fellowship (SRF) from Council of Scientific and Industrial Research, India (CSIR) (Award No. 09/727(0041)/2005-EMR-I). Appendix A. Supplementary material Supplementary data related to this article can be found online at http://dx.doi.org/10.1016/j.biochi.2012.08.011. References [1] A.P. Wolffe, J.J. Hayes, Chromatin disruption and modification, Nucleic Acids Res. 27 (1999) 711e720. [2] B.D. Strahl, C.D. Allis, The language of covalent histone modifications, Nature 403 (2000) 41e45. [3] M. Iizuka, M.M. Smith, Functional consequences of histone modifications, Curr. Opin. Genet. Dev. 13 (2003) 154e160. [4] T. Jenuwein, C.D. Allis, Translating the histone code, Science 293 (2001) 1074e1080. [5] B.M. Turner, Cellular memory and the histone code, Cell 111 (2002) 285e291. [6] M. Grunstein, Histone acetylation in chromatin structure and transcription, Nature 389 (1997) 349e352. [7] M. Lachner, T. Jenuwein, The many faces of histone lysine methylation, Curr. Opin. Cell. Biol. 14 (2002) 286e298. [8] K. Struhl, Histone acetylation and transcriptional regulatory mechanisms, Genes Dev. 12 (1998) 599e606. [9] P. Chi, C.D. Allis, G.G. Wang, Covalent histone modifications e miswritten, misinterpreted and mis-erased in human cancers, Nat. Rev. Cancer 10 (2010) 457e469. [10] B.M. Lee, L.C. Mahadevan, Stability of histone modifications across mammalian genomes: implications for ‘epigenetic’ marking, J. Cell. Biochem. 108 (2009) 22e34. [11] C.R. Vakoc, S.A. Mandat, B.A. Olenchock, G.A. Blobel, Histone H3 lysine 9 methylation and HP1gamma are associated with transcription elongation through mammalian chromatin, Mol. Cell. 19 (2005) 381e391. [12] C.R. Vakoc, M.M. Sachdeva, H. Wang, G.A. Blobel, Profile of histone lysine methylation across transcribed mammalian chromatin, Mol. Cell. Biol. 26 (2006) 9185e9195. [13] A.J. Bannister, R. Schneider, F.A. Myers, A.W. Thorne, C. Crane-Robinson, T. Kouzarides, Spatial distribution of di- and tri-methyl lysine 36 of histone H3 at active genes, J. Biol. Chem. 280 (2005) 17732e17736. [14] R. Schneider, A.J. Bannister, F.A. Myers, A.W. Thorne, C. Crane-Robinson, T. Kouzarides, Histone H3 lysine 4 methylation patterns in higher eukaryotic genes, Nat. Cell. Biol. 6 (2004) 73e77. [15] Y.C. Lin, S. Jhunjhunwala, C. Benner, S. Heinz, E. Welinder, R. Mansson, M. Sigvardsson, J. Hagman, C.A. Espinoza, J. Dutkowski, T. Ideker, C.K. Glass, C. Murre, A global network of transcription factors, involving E2A, EBF1 and Foxo1, that orchestrates B cell fate, Nat. Immunol. 11 (2010) 635e643. [16] K. van Dijk, K.E. Marley, B.R. Jeong, J. Xu, J. Hesson, R.L. Cerny, J.H. Waterborg, H. Cerutti, Monomethyl histone H3 lysine 4 as an epigenetic mark for silenced euchromatin in Chlamydomonas, Plant Cell 17 (2005) 2439e2453. [17] C.D. Carvin, M.P. Kladde, Effectors of lysine 4 methylation of histone H3 in Saccharomyces cerevisiae are negative regulators of PHO5 and GAL1-10, J. Biol. Chem. 279 (2004) 33057e33062.
[18] J.P. Jackson, L. Johnson, Z. Jasencakova, X. Zhang, L. PerezBurgos, P.B. Singh, X. Cheng, I. Schubert, T. Jenuwein, S.E. Jacobsen, Dimethylation of histone H3 lysine 9 is a critical mark for DNA methylation and gene silencing in Arabidopsis thaliana, Chromosoma 112 (2004) 308e315. [19] A.H. Peters, S. Kubicek, K. Mechtler, R.J. O’Sullivan, A.A. Derijck, L. PerezBurgos, A. Kohlmaier, S. Opravil, M. Tachibana, Y. Shinkai, J.H. Martens, T. Jenuwein, Partitioning and plasticity of repressive histone methylation states in mammalian chromatin, Mol. Cell. 12 (2003) 1577e1589. [20] A. Lomba, F.I. Milagro, D.F. Garcia-Diaz, A. Marti, J. Campion, J.A. Martinez, Obesity induced by a pair-fed high fat sucrose diet: methylation and expression pattern of genes related to energy homeostasis, Lipids Health Dis. 9 (2010) 60. [21] C.P. Mack, An epigenetic clue to diabetic vascular disease, Circ. Res. 103 (2008) 568e570. [22] F. Miao, R. Natarajan, Mapping global histone methylation patterns in the coding regions of human genes, Mol. Cell. Biol. 25 (2005) 4650e4661. [23] Z. He, T. Jiang, Z. Wang, M. Levi, J. Li, Modulation of carbohydrate response element-binding protein gene expression in 3T3-L1 adipocytes and rat adipose tissue, Am. J. Physiol. Endocrinol. Metab. 287 (2004) E424eE430. [24] I. Matias, M.P. Gonthier, P. Orlando, V. Martiadis, L. De Petrocellis, C. Cervino, S. Petrosino, L. Hoareau, F. Festy, R. Pasquali, R. Roche, M. Maj, U. Pagotto, P. Monteleone, V. Di Marzo, Regulation, function, and dysregulation of endocannabinoids in models of adipose and beta-pancreatic cells and in obesity and hyperglycemia, J. Clin. Endocrinol. Metab. 91 (2006) 3171e3180. [25] V.G. Tusher, R. Tibshirani, G. Chu, Significance analysis of microarrays applied to the ionizing radiation response, Proc. Natl. Acad. Sci. U S A 98 (2001) 5116e5121. [26] J. Li, J. Liu, R.K.M. Karuturi, Stepped Linear Regression to Accurately Assess Statistical Significance in Batch Confounded Differential Expression Analysis, in: Bioinformatics Research and Applications, Lecture Notes in Computer Science, Springer. 4983 (2008) 481e491. [27] D.G. Kabra, J. Gupta, K. Tikoo, Insulin induced alteration in post-translational modifications of histone H3 under a hyperglycemic condition in L6 skeletal muscle myoblasts, Biochim. Biophys. Acta 1792 (2009) 574e583. [28] F. Miao, D.D. Smith, L. Zhang, A. Min, W. Feng, R. Natarajan, Lymphocytes from patients with type 1 diabetes display a distinct profile of chromatin histone H3 lysine 9 dimethylation: an epigenetic study in diabetes, Diabetes 57 (2008) 3189e3198. [29] F. Miao, X. Wu, L. Zhang, Y.C. Yuan, A.D. Riggs, R. Natarajan, Genome-wide analysis of histone lysine methylation variations caused by diabetic conditions in human monocytes, J. Biol. Chem. 282 (2007) 13854e13863. [30] A. El-Osta, D. Brasacchio, D. Yao, A. Pocai, P.L. Jones, R.G. Roeder, M.E. Cooper, M. Brownlee, Transient high glucose causes persistent epigenetic changes and altered gene expression during subsequent normoglycemia, J. Exp. Med. 205 (2008) 2409e2417. [31] D. Schubeler, D.M. MacAlpine, D. Scalzo, C. Wirbelauer, C. Kooperberg, F. van Leeuwen, D.E. Gottschling, L.P. O’Neill, B.M. Turner, J. Delrow, S.P. Bell, M. Groudine, The histone modification pattern of active genes revealed through genome-wide chromatin analysis of a higher eukaryote, Genes Dev. 18 (2004) 1263e1271. [32] J.K. Sims, S.I. Houston, T. Magazinnik, J.C. Rice, A trans-tail histone code defined by monomethylated H4 Lys-20 and H3 Lys-9 demarcates distinct regions of silent chromatin, J. Biol. Chem. 281 (2006) 12760e12766. [33] J.C. Rice, S.D. Briggs, B. Ueberheide, C.M. Barber, J. Shabanowitz, D.F. Hunt, Y. Shinkai, C.D. Allis, Histone methyltransferases direct different degrees of methylation to define distinct chromatin domains, Mol. Cell. 12 (2003) 1591e1598. [34] N. Pelletier, N. Champagne, S. Stifani, X.J. Yang, MOZ and MORF histone acetyltransferases interact with the Runt-domain transcription factor Runx2, Oncogene 21 (2002) 2729e2740. [35] S.B. Seo, P. McNamara, S. Heo, A. Turner, W.S. Lane, D. Chakravarti, Regulation of histone acetylation and transcription by INHAT, a human cellular complex containing the set oncoprotein, Cell 104 (2001) 119e130. [36] Y. Katoh, M. Katoh, Comparative integromics on JMJD2A, JMJD2B and JMJD2C: preferential expression of JMJD2C in undifferentiated ES cells, Int. J. Mol. Med. 20 (2007) 269e273. [37] Z. Yang, J. Jiang, D.M. Stewart, S. Qi, K. Yamane, J. Li, Y. Zhang, J. Wong, AOF1 is a histone H3K4 demethylase possessing demethylase activity-independent repression function, Cell. Res. 20 (2010) 276e287. [38] K. Tikoo, R.L. Meena, D.G. Kabra, A.B. Gaikwad, Change in post-translational modifications of histone H3, heat-shock protein-27 and MAP kinase p38 expression by curcumin in streptozotocin-induced type I diabetic nephropathy, Br. J. Pharmacol. 153 (2008) 1225e1231. [39] K. Tikoo, D.N. Tripathi, D.G. Kabra, V. Sharma, A.B. Gaikwad, Intermittent fasting prevents the progression of type I diabetic nephropathy in rats and changes the expression of Sir2 and p53, FEBS Lett. 581 (2007) 1071e1078. [40] A.B. Gaikwad, J. Gupta, K. Tikoo, Epigenetic changes and alteration of Fbn1 and Col3A1 gene expression under hyperglycaemic and hyperinsulinaemic conditions, Biochem. J. 432 (2010) 333e341. [41] P. Chomczynski, N. Sacchi, Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction, Anal. Biochem. 162 (1987) 156e159.
Contents lists available at SciVerse ScienceDirect
Biochimie journal homepage: www.elsevier.com/locate/biochi
Research paper
Histone H3 lysine 4 monomethylation (H3K4me1) and H3 lysine 9 monomethylation (H3K9me1): Distribution and their association in regulating gene expression under hyperglycaemic/hyperinsulinemic conditions in 3T3 cells Jeena Gupta a, Sandeep Kumar a, Juntao Li b, R. Krishna Murthy Karuturi b, Kulbhushan Tikoo a, * a
Laboratory of Chromatin Biology, Department of Pharmacology and Toxicology, National Institute of Pharmaceutical Education and Research (NIPER), Mohali, Punjab, India Computational and Systems Biology, Genome Institute of Singapore, A*STAR (Agency for Science, Technology and Research), 60 Biopolis St, S138672, Singapore, Republic of Singapore
b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 21 December 2011 Accepted 2 August 2012 Available online 21 August 2012
Hyperglycemia/hyperinsulinemia are leading cause for the induction type 2 diabetes and the role of posttranslational histone modifications in dysregulating the expression of genes has emerged as potential important contributor in the progression of disease. The paradoxical nature of histone H3-Lysine 4 and Lysine 9 mono-methylation (H3K4me1 and H3K9me1) in both gene activation and repression motivated us to elucidate the functional relationship of these histone modifications in regulating expression of genes under hyperglycaemic/hyperinsulinemic condition. Chromatin immunoprecipitationemicroarray analysis (ChIP-chip) was performed with H3 acetylation, H3K4me1 and H3K9me1 antibody. CLUSTER analysis of ChIP-chip (Chromatin immunoprecipitationemicroarray analysis) data showed that mRNA expression and H3 acetylation/H3K4me1 levels on genes were inversely correlated with H3K9me1 levels on the transcribed regions, after 30 min of insulin stimulation under hyperglycaemic condition. Interestingly, we provide first evidence regarding regulation of histone de/acetylases and de/methylases; Myst4, Jmjd2b, Aof1 and Set by H3Ac, H3K4me1 and H3K9me1 under hyperinsulinemic/hyperglycaemic condition. ChIPeqPCR analysis shows association of increased H3Ac/H3K4me1 and decreased levels of H3K9me1 in up regulation of Myst4, Jmjd2, Set and Aof1 genes. We further analyse promoter occupancy of histone modifications by ChIP walking and observed increased occupancy of H3Ac/H3K4me1 on promoter region (1000 to 1) of active genes and H3K9me1 on inactive genes under hyperglycemic/ hyperinsulinemic condition. To best of our knowledge this is the first report that shows regulation of chromatin remodelling genes by alteration in the occupancy of histone H3Ac/H3K4/K9me on both promoter and transcribed regions. Ó 2012 Elsevier Masson SAS. All rights reserved.
Keywords: Hyperglycemia Insulin Histone modifications Chromatin remodelling genes Chromatin immunoprecipitation and diabetes
1. Introduction Long standing models have suggested the importance of histone modifications in altering the chromatin structure and regulating the histoneeDNA/histoneehistone interactions [1,2]. These histoneeDNA interactions set a critical stage for the regulation of the transcription of genes by influencing the DNA accessibility to the cell’s transcriptional machinery elements [3]. The transcriptional regulation of gene expression is a dynamic process and mediates the conversion of inactive heterochromatin structure to
* Corresponding author. Department of Pharmacology and Toxicology, National Institute of Pharmaceutical Education and Research (NIPER), Sector 67, S.A.S. Nagar, Mohali, Punjab-160 062, India. Tel.: þ91 172 2214682 87; fax: þ91 172 2214692. E-mail address: [email protected] (K. Tikoo). 0300-9084/$ e see front matter Ó 2012 Elsevier Masson SAS. All rights reserved. http://dx.doi.org/10.1016/j.biochi.2012.08.011
active euchromatin structure and vice versa. The elaborate collection of posttranslational modifications of histone proteins including acetylation, phosphorylation, methylation, ubiquitination and ADPribosylation, acts like a molecular code, also known as “histone code” and provides a platform to accommodate the binding of factors regulating the chromatin function [4,5]. Therefore, aberrant alterations in these epigenetic modifications (that do not involve a change in DNA sequence), can change chromatin structure and thus could lead to the deregulated gene transcription and the disease progression. Among the modifications of different histones (H2A, H2B, H3 and H4), the modifications of histone H3 are most studied and appreciated in the regulation of gene transcription [6,7]. Evidences show that the acetylation of key lysine residues of histone H3 is generally associated with gene activation [8]. However H3 lysine methylation displays a high degree of complexity as methylation of
J. Gupta et al. / Biochimie 94 (2012) 2656e2664
lysines 4, 36 and 79 of histone H3 primarily dictate gene activation whereas, methylation at lysines 9, 27, 20 are most commonly described at sites of gene repression [9,10]. But certain methylation marks can have multiple functions as H3K9, a methylation mark commonly enriched at the heterochromatin, has also been found to be present at the transcribed regions of active genes [11,12]. In addition, there are other mechanisms also by which an individual site of histone modifications can carry differential regulation of gene expression. These include the degree of processivity (mono, di or tri functional group) and the spacial distribution of these histone modifications on either the promoter region or the transcribed region of the gene. Reports have shown that in Saccharomyces cerevisiae, H3 lysine 4 dimethylation (H3K4me2) and H3 lysine 4 trimethylation (H3K4me3) have been shown to peak in the 50 transcribed region of active genes, implicating them in the early phases of the transcriptional elongation [13,14]whereas, H3 lysine 4 monomethylation (H3K4me1) is more characterized on the enhancer region of the gene promoters [15]. In contrast some studies have suggested the association of H3K4me1 with silenced euchromatin [16,17]. In Arabidopsis, another histone H3 modification dimethylation of lysine 9 (H3K9me2) is found to be a critical mark for gene silencing but not monomethylation (H3K9me1) or trimethylation (H3K9me3) [18]. However, there are reports that show the enrichment of H3K9me3 at heterochromatin region in mammalian cells, while H3K9me1 and H3K9me2 are localized to the euchromatin region of silenced genes [19]. Therefore distribution of these distinct histone H3 modifications; K4me and K9me remain unclear. Thus underlying the paradoxical nature of these histone modifications, we carried out the present study to delineate the functional properties of H3K4 and H3K9 monomethylation marks and also to study their association with H3 acetylation, a well known gene activation mark. Reports have shown that, among the different mechanisms that could lead to induction of diabetes, the role of epigenetic mechanisms in dysregulating the expression of genes, has emerged as a potential important contributor in the last decade [20]. Several studies have revealed interesting insights into the relationship between di-methylation and tri-methylation states of H3K4/H3K9 and type 2 diabetes [21,22] but none of the study has demonstrated the functional significance of mono-methylated states of H3K4/ H3K9 underlying diabetic condition (hyperglycemia/hyperinsulinemia). Adipocytes being the important contributor in the induction of the lifestyle disorders; obesity and type 2 diabetes, so in the present study we try to mimic the diabetic conditions by culturing 3T3 adipocytes under high glucose and high insulin condition [23,24]. Our aim was to functionally characterize H3K4/ H3K9 monomethylation with the alteration in gene expression in 3T3 cells under hyperglycemic/hyperinsulinemic conditions.
2657
USA). All the other chemicals were purchased from Sigma (St. Louis, MO, USA), unless otherwise mentioned. 2.2. Cell culture and treatment 3T3 cells were cultured in DMEM supplemented with 10% foetal bovine serum and antibiotics (penicillin 100 IU/ml and streptomycin 100 mg/ml) in 5% CO2 incubator at 37 C. At 80e90% confluency, cells were treated with either low glucose (LG) (5 mM þ 20 mM mannitol, for osmotic balance) or high glucose (HG) (25 mM) for 5 h in serum-free DMEM, followed by insulin stimulation (100 nM) for 30 min (denoted by LGI or HGI respectively). Cells were then washed twice with ice cold phosphate buffer saline (PBS). 2.3. Isolation of total RNA 3T3 cells (80e90% confluence) were treated with low glucose (5 mM glucose þ 20 mM mannitol) and high glucose (25 mM glucose) for 5 h and stimulated with insulin (100 nM) for 30 min. At the end of the treatment, the cells were washed with PBS and the total RNA was isolated using a protocol as described by Chomczynski and Sacchi [41] using TRIZOL reagent (Invitrogen) and was purified using RNeasy kit (Auprep RNeasy mini kit, Life Technologies) according to the manufacturer’s protocol. The RNA quality and integrity from each sample were assured using Nanodrop (ND-1000) by A260/280 absorbance ratio and using agarose gel electrophoresis respectively. The total RNA prepared from 3T3 cells was used for the quantitative PCR (qPCR) and the cDNA microarray analysis. 2.4. Probe labelling and hybridization The mouse 15K array (Microarray centre, University Health Care, Toronto) used in the present study consisted of 15,264 genes spotted in duplicate. The fluorescence labelled cDNA probes were prepared from 20 mg of total RNA using a SuperScript III (Invitrogen Life Technologies), oligo dT primer (Invitrogen Life Technologies) and Cy3 or Cy5 labelled dCTPs (Amersham biosciences) from the low glucose and the high glucose treated 3T3 cells stimulated with insulin for 30 min. The Cy3 and Cy5 probes were then mixed in an equal amount and the hybridization was carried out at 43 C for 16 h on Hyb Array 12 hybridization station (Perkin Elmer). The hybridization was repeated three times. A dye-swapped experiment was also performed to improve the accuracy of the measurement and to rule out the nonspecific signals. 2.5. Analysis of microarray data
2. Methods 2.1. Chemicals 3T3 cell line was purchased from American Type Cell Culture (ATCC). Cell culture media, antibiotic solution, foetal calf serum, and trypsineEDTA solution were purchased from GIBCO (USA). Insulin was purchased from Novartis Pharma and CM-DCF-DA from Molecular Probes (Invitrogen USA). ECL detection kit and ECL hyperfilm were obtained from Amersham Bioscience (USA). Anti-pH3 (Ser-10) antibody and horseradish peroxidase (HRP)-conjugated secondary antibody were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Anti-acetylated H3, anti-monomethyl histone H3-K4, anti-monomethyl histone H3-K9 and anti histone H3 were purchased from Upstate Biotechnology (Lake Placid, NY,
Slides were scanned with a Scanarray Gx microarray scanner (Perkin Elmer) and the images were analysed using Scanarray software. The flagged spots were excluded from the analysis. The mean intensity for each spot was taken and the log 2 value for the signal intensity of each spot was calculated for each slide. Genes with FG (foreground) >1.5 BG (Background) were retained for further analysis. The SAM (Significance Analysis of Microarrays) [25] in conjunction with SLR (stepped linear regression) [26] was used for finding significantly differentially expressed or histone modified genes with FDR (false discovery rate) <10%. The genes showing more than 1.5 fold change were taken for further analysis. Genecard (www.genecards.org) and Pubmed (www.pubmed.gov) were used for assigning the genes for specific biological processes. For clustering, we have selected the differentially expressed genes with change in any one of histone modification and cluster these
2658
J. Gupta et al. / Biochimie 94 (2012) 2656e2664
genes on the basis of correlation coefficient by Cluster and Treeview analysis (http://rana.lbl.gov/EisenSoftware.htm). 2.6. Reverse transcriptase polymerase chain reaction (RT-qPCR) To confirm the cDNA microarray, qPCR was performed for certain genes. The first strand cDNA was generated from 20 mg of total RNA by using a SuperScript III (Invitrogen, Life Technologies) and oligo dT primer (Invitrogen, Life Technologies) in a 30 ml reaction mixture. The quantitative real time PCR reaction was performed using LightCycler 2.0 (Roche Diagnostics) in a final volume of 20 ml containing 50 ng of cDNA, 4 ml of reaction buffer from LightCycler FastStart DNA master plus kit (Roche Diagnostics) and specific forward and reverse primers (Midland Certified Reagent Company Inc.). After the amplification, a melting curve analysis was performed to verify the specificity of the reaction. The analysis was performed using LightCycler software (Roche Diagnostics). The relative gene expression was assessed using the comparative Ct (DCt) method and normalized to GAPDH. 2.7. ChIP-on-chip (chromatin immunoprecipitation and microarray hybridization) ChIPs were performed by using MAGnify Chromatin immunoprecipitation kit (Invitrogen, Life Technologies). Briefly, the insulinstimulated cells under low glucose and high glucose condition were washed with PBS after the indicated time interval and were crosslinked using 1% formaldehyde for 20 min. Cross-linking was stopped by the addition of 0.1 M glycine and the cell lysates were sonicated. Consistency of equal levels of sonication of the cell lysates was maintained and before immunoprecipitation, all samples were tested for the level of DNA fragmentation, with the mean size of DNA fragments was always kept 500 bp. Immunoprecipitation with the indicated antibodies or preimmune mouse IgG (as control) was carried out for 2 h with magnetic beads. Input (20 ml of sample separated before immunoprecipitation) and immunoprecipitates were washed, eluted, and then crosslinking was reversed. The DNA fragments were recovered using magnetic beads provided with kit and was amplified for specific genes by qPCR. No DNA controls were always included; making sure that the primeredimer formation was not detectable in the experimental samples while performing the qPCR. Relative fold change was assessed using the comparative Ct (DCt) method and normalized to input. Experiments were typically repeated three times; the error bars in the figures show the standard deviations. For ChIP-chip analysis labelling of the immunoprecipitated DNA was performed using the Bioprime Labelling system (Invitrogen, Life Technologies). Labelled ChIP/Cy5 and Cy3 DNAs were combined and hybridized to the array as already described above. 3. Results 3.1. Profiling histone modifications across coding regions of genes under high glucose condition To identify the differential expression of genes under hyperglycemic/hyperinsulinemic condition we performed cDNA microarray analysis after insulin stimulation of 30 min under high glucose condition. SAM analysis in conjugation with Stepped Linear Regression analysis [26] was performed to filter differentially expressed genes with at least 1.5 fold difference and FDR <10%. With this we identified 1301 genes which show change in expression in HGI (high glucose þ insulin) as compared to LGI (low glucose þ insulin) condition (Online Appendix 1). To get further insight into the level of H3K9me1, H3K4me1 and H3Ac across the
coding regions of the mouse genome, we performed ChIPecDNA analysis using 15 K cDNA array after 30 min of the insulin stimulation under both low and high glucose condition. To detect the difference in these histone modifications between high glucose and low glucose condition, the profiles were determined as normalized ratios HGI/LGI (fold change). Applying the above mentioned statistical analysis and criterion we identified 1168 targets for H3Ac (Online appendix 2), 148 targets for H3K4me1 (Online appendix 3) and 546 targets for H3K9me1 (Online appendix 4) with differential status in HGI as compared to LGI condition in coding regions of the genes. However there were only 12 common genes between these three modifications on cDNA array. To understand the role of these histone H3 modifications in regulation of the genes under hyperglycemic/hyperinsulinemic conditions, we identified the genes that underwent changes in any of these three histone modifications along with change in their gene expression levels. To do so, we set up a criterion and select only the genes that were common in cDNA expression analysis with differential change in status in any one of either H3Ac or H3K4me1 or H3K9me1. This stringent criterion might result in false negatives but it also reduces the number of genes to a manageable size for further validation analysis and the chance of having false positives. With this criterion we identified 831 genes with significant differential H3Ac or H3K4me1 or H3K9me1 status and also change in their mRNA expression levels (Online appendix 5). Of these 608 genes were down regulated and 223 genes were up regulated. To evaluate the major effected biological process in the cells under hyperglycemia/hyperinsulinemia, we further categorized these genes to the various functional categories and observed that most of the biological processes were down regulated (Fig. 1). Also we identified that signal transduction; transcription and protein metabolism were among the major effected pathway. We further validate the change in mRNA expression and H3Ac and H3K9me1 levels on insulin signalling genes which were known to play a role in diabetes, Gapvd1, Mapk6, Ctbp1 and Nfat by qPCR and ChIPeqPCR analysis respectively and observed similar levels of change (Fig. 2). These results demonstrated that cells exposed to hyperglycemia and hyperinsulinemia (pathological conditions of type 2 diabetes), disturb most of the biologic pathways of the cell, mainly hampering the signal transduction elements by altering histone H3 modifications. 3.2. Correlation between H3Ac, H3K4me1 and H3K9me1 status with mRNA expression levels Further to delineate the functional relationship of H3Ac, H3K4me1 and H3K9me1 with each other and with mRNA expression of the genes, we performed hierarchical clustering of the genes using Pearson correlation coefficient metric on the above selected dataset (candidate genes with significant difference in any of the three modifications and having differential mRNA expression level) using publicly available software packages CLUSTER and TREEVIEW (http://rana.lbl.gov/EisenSoftware.htm) (Fig. 3). With this analysis we demonstrate that histone H3Ac levels in the coding regions of the genes very well correlate with the mRNA expression level of the respective genes signifying H3Ac as a mark of gene activation in the coding regions of the genes. Furthermore, mRNA expressions of most of the genes were inversely proportional to H3K9me1 levels, suggesting that increased H3K9me1 occupancy in the coding regions of the genes is associated with gene inactivation. However, very less number of genes was enriched for H3K4me1 in the coding regions and we also failed to observe much overlap between H3K4me1 and mRNA expression levels (Fig. 3). These results indicate increased occupancy of H3Ac and H3K9me1 in the coding region of the genes and not for H3K4me1.
J. Gupta et al. / Biochimie 94 (2012) 2656e2664
2659
Fig. 1. cDNA microarray and ChIP-chip analysis compares the genes regulated by the insulin (100 nM) stimulation under high glucose as compared to low glucose conditions: Bar diagram shows differentially expressed genes towards the various functional pathways as analysed by the microarray analysis along with changes in any of these three histone modifications (H3Ac, H3K4me1 or H3K9me1) as analysed by ChIP-chip analysis after 30 min of insulin stimulation under high glucose as compared to the low glucose condition. Data shows statistically significant change in gene expression with FDR <10% and fold change >1.5. The figure is derived from the Supplementary data (Online appendices 1e5) and assigned to the corresponding biological process as described in the “Materials and methods”.
3.3. Hyperglycemia/hyperinsulinemia induced changes in expression of chromatin modifying genes and their regulation by histone modifications Out of differentially expressed genes identified by cDNA microarray and ChIP-chip analysis, we observed significant change in the expression of genes that are responsible for mediating chromatin remodelling by insulin under high glucose condition. These include down regulation of Myst4 and Ep400 (histone acetyltransferases, HAT), Jmjd2b and Jarid2 (histone methyl-transferases,
HMT) and Dyrk2 (histone kinase). In addition to above mentioned genes, Brdt gene which is involved in reorganization of acetylated chromatin was also found to be down regulated. Increase in the expression of Set gene (HAT inhibitor) and also genes responsible for histone H3K4 de-methylation (Jarid1a and Aof1) further supports our earlier observation. The change in expression of these genes observed in the present study was in accordance with our previous findings that show decrease in levels of H3Ac, H3K4me1 and H3K9me1 after 30 min of insulin stimulation under high glucose condition [27].
Fig. 2. Change in gene expression, histone H3 acetylation and lysine 9 monomethylation levels on coding regions of the insulin signalling genes: 3T3 cells were cultured as described in the “Materials and methods”. For ChIPeqPCR cells were cross linked with 1% formaldehyde and the ChIP was performed with respective antibody. Bar graph shows change in the mRNA expression, H3Ac and H3K9me1 levels on a: Gapvd1; b: Mapk6; c: Ctbp1 and d: Nfat, where LGI represents cells cultured under low glucose condition and treated with insulin for 30 min and HGI represents cells cultured under high glucose condition and treated with insulin for 30 min. Relative fold change was calculated after normalization with mouse 18S gene and input for qPCR and ChIPeqPCR respectively. Similar results were obtained in the three independent sets of experiments. All the values were represented as mean S.E.M. (n ¼ 3), ***p < 0.001, **p < 0.01 and *p < 0.05, Vs LGI.
2660
J. Gupta et al. / Biochimie 94 (2012) 2656e2664
Fig. 3. Hierarchical cluster analysis of mRNA, H3Ac, H3K4me1 and H3K9me1 profiles on coding regions of genes altered by the insulin (100 nM) stimulation under high glucose as compared to low glucose conditions: genes were selected for <10% FDR and >1.5 fold enrichment in ChIP DNA under high glucose as compared to low glucose after 30 min of insulin stimulation. Lane 1 shows mRNA expression profile, lane 2: H3Ac profile, lane 3: H3K4me1 profile and lane 4 shows H3K9me1 profile, where red indicates increased mRNA expression or increased H3Ac/H3K4me1/H3K9me1 levels, green indicates decreased mRNA expression or decreased H3Ac/H3K4me1/H3K9me1 levels and black indicates no change. Intensity of colour correlates to the magnitude of change. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Out of the above nine chromatin remodelling genes that show expression change in our microarray data, four genes were selected for further validation by qPCR. The four selected chromatin remodelling genes were each from different class of histone modifications, i.e., Myst4 is basically known to be a histone acetyltransferase, Set is a HAT inhibitor, Jmjd2b is a histone demethylase (specifically demethylate H3K9) and Aof1 is a histone H3K4 demethylase [34e37]. Similar changes in expression of these genes as observed by both microarray and qPCR analysis (Fig. 4) further prompted us to confirm the change in H3Ac, H3K4me1 and H3K9me1 levels on their coding regions by performing ChIPeqPCR analysis (Fig. 5). We observed a decrease in the level of H3Ac on Myst4 and Jmjd2b and an increase on Set and Aof1 genes confirming our ChIP-chip data. However, we failed to observe any change in H3K9me1 levels on the coding regions of histone H3K9 demethylase (Jmjd2b) and H3K4 demethylase (Aof1). Decreased H3K4me1 levels on Myst4 and Jmjd2b and increased H3K4me1 levels on Set and Aof1 further confirmed our ChIP-chip analysis (Fig. 5). These results suggest the presence of crosstalk between these histone modifications under hyperinsulinemic/hyperglycemic conditions. However, levels of H3K9me1 were only changed on histone
Fig. 4. Change in gene expression of chromatin remodelling genes: the fold change in mRNA expression levels of chromatin remodelling genes determined by microarray analysis was further validated by qPCR using a light cycler 2.0 on RNA extracted from 3T3 cells cultured as described in the “Materials and methods”. Bar graphs show change in the mRNA expression of the histone modification regulating genes; Jmjd2b, Aof1, Myst4 and Set under the high glucose condition (HGI) as compared to the low glucose condition (LGI) after insulin treatment of 30 min, a: by microarray analysis and b: by qPCR analysis. Relative fold change was calculated after normalization with mouse 18S gene for qPCR analysis. Similar results were obtained in the three independent sets of experiments. All the values were represented as mean S.E.M. (n ¼ 3), ***p < 0.001, **p < 0.01 and *p < 0.05, Vs LGI.
acetylase (Myst4) and deacetylase (Set), highlighting the role of this modification in regulating histone acetylation only. 3.4. Relative abundance of histone modifications on promoters of chromatin remodelling genes Studies have shown that histone modifications adopt characteristic and specific distribution for regulating gene expression, some of which occur in the vicinity of promoters, while others occur across the coding regions of the genes. Hence, to get further insight into the functional role of H3Ac, H3K4me1 and H3K9me1 in regulating gene expression by their presence/absence on the promoter region of the differentially expressed genes, we performed ChIP-walking. We focused mainly on the regulation of promoter regions (nt 2000 to 1 from transcription start site) of the chromatin remodelling genes Myst4, Jmjd2b, Set and Aof1. Our data shows similar pattern of change for H3Ac and H3K4me1 on Myst4, Jmjd2b, Aof1 and Set gene promoter regions under both low glucose and high glucose condition after insulin stimulation. Comparing the relative abundance of these histone modifications, we observed an increased occurrence of H3Ac and H3K4me1 between regions 1000 to 500 nt on Myst4 and Jmjd2b promoter under low glucose condition which on the other hand get decreased under high glucose condition. H3K9me1 shows an opposite trend suggesting the specific distribution of these histone modifications
J. Gupta et al. / Biochimie 94 (2012) 2656e2664
2661
Fig. 5. Histone H3 acetylation, lysine 4 monomethylation and lysine 9 monomethylation levels on coding regions of the chromatin modification regulating genes: 3T3 cells were cultured as described in the “Materials and methods”. Bar graph shows H3Ac, H3K4me1 and H3K9me1 levels on a: Myst4; b: Set; c: Jmjd2b and d: Aof1, where LGI represents cells cultured under low glucose condition and treated with insulin for 30 min and HGI represents cells cultured under high glucose condition and treated with insulin for 30 min. Relative fold change was calculated after normalization with input. Similar results were obtained in the three independent sets of experiments. All the values were represented as mean S.E.M. (n ¼ 3), ***p < 0.001, **p < 0.01 and *p < 0.05, Vs LGI.
on the promoter regions of histone acetylase Myst4 (Fig. 6) and histone demethylase Jmjd2b (Fig. 7). The decreased mRNA expression of Myst4 and Jmjd2b under high glucose as observed by qPCR (see Section 3.4), further emphasizes the role of H3Ac and H3K4me1 in gene activation and H3K9me1 in gene silencing even on the promoter region of the chromatin remodelling genes. Histone deacetylase Set and histone demethylase Aof1 promoter regions showed increased H3Ac and H3K4me1 at regions 800 to 300 nt and 1000 to 600 nt respectively under high glucose condition as compared to low glucose condition. However, H3K9me1 methylation got decreased at these regions (Figs. 6 and 7). Comparison of these results with the increased mRNA expression of Set and Aof1 under high glucose condition further confirms the significance of H3K9me1 as gene inactivation mark at the promoter region of these genes. To check the specificity of our ChIP assays, we also performed ChIPs with mouse IgGs and did not observe any significant amplification. 4. Discussion Despite the knowledge that high glucose alters multiple histone modifications and associated gene expression [28,29] including the sustainable epigenetic changes by long term memory of chronic exposure to high glucose [30], the impact of chronic insulin treatment under high glucose condition on histone monomethylation status has not been previously explored on a genome wide scale. By profiling of histone modifications with alterations in gene expression under hyperglycemic/hyperinsulinemic condition (mimicking diabetic condition), we for the first time show the differential regulation of gene expression by H3Ac, H3K4me1 and H3K9me1. CLUSTER analysis of our ChIPecDNA microarray data demonstrates direct correlation between H3Ac levels and mRNA expression levels suggesting H3Ac as mark of gene activation in accordance with earlier reports [8]. However, the low frequency of occurrence of H3K4me1 on the coding regions of the transcriptionally up regulated genes as observed by ChIPecDNA microarray
analysis, suggests that monomethylation of H3K4 may not be always required for transcriptional elongation of active genes by RNA polymerase, as established previously for H3K4 di/tri methylation [14,31]. van Dijk et al. have shown H3K4 monomethylation as a mark for silenced euchromatin in Chlamydomonas [16]. However our data shows an increase in relative abundance of H3K4me1 along with H3Ac on the promoters (region between 0 and 1000) of the chromatin remodelling genes (Myst4, Set, Jmjd2b and Aof1) under hyperglycemic/hyperinsulinemic condition. Increase in relative abundance of H3Ac and H3K4me1 matches very well with the increased level of mRNA expression of these genes. Our data is in agreement with earlier report by Lin et al., which shows higher abundance of H3K4 monomethylation on the enhancer regions of active genes in B cell development [15]. Thus our results emphasize the role of H3Ac and H3K4me1 in regulating the expression of chromatin remodelling genes under diabetic condition by altering its relative abundance on the promoter region of these genes. Apart from histone H3Ac and H3K4me1 we further studied the role of other histone modification (H3K9me1) which may act as a mark or signature for inactive/active chromatin. Correlating the levels of H3K9me1 with mRNA expression of the genes by CLUSTER analysis depicts that most of the genes with increased occupancy of H3K9me1 as well as decreased H3 acetylation levels across the coding regions were down regulated. Earlier reports have also shown the association of H3K9 mono/di/tri methylation with transcriptionally silent chromatin [18,32]. But we for the first time provide a global profile of H3K9me1 across the transcribed regions of the genes under hyperglycemic/hyperinsulinemic condition. Rice et al. have shown that promoter regions of silenced genes in euchromatin are largely mono or di-methylated at histone H3K9 residue, whereas pericentric heterochromatin is predominantly trimethylated at H3K9 residue [33]. Consistent with this report we also provide evidence that there is an increased distribution of H3K9me1 across the promoter region (1 to 1000 bp) of the chromatin remodelling genes down regulated under hyperinsulinemic/hyperglycemic condition. Collectively these results
2662
J. Gupta et al. / Biochimie 94 (2012) 2656e2664
Fig. 6. Relative abundance of histone modifications on promoter of histone de/acetylases Myst4 and Set: 3T3 cells were cultured as described in the “Materials and methods”. qPCR was performed across the promoter region of Myst4 and Set. All distances are relative to transcription start site. Plot shows H3Ac, H3K4me1 and H3K9me1 levels on a: Myst4 promoter under low glucose condition after 30 min of insulin stimulation, b: Myst4 promoter under high glucose condition after 30 min of insulin stimulation, c: Set promoter under low glucose condition after 30 min of insulin stimulation and d: Set promoter under high glucose condition after 30 min of insulin stimulation. Similar results were obtained in the three independent sets of experiments.
highlight the role of histone H3K9me1 in gene inactivation on both promoter as well as transcribed regions. Interestingly our data also shows the cross-talk between the genes involved in altering histone H3 modifications (Fig. 8). Acetylation and methylation of histones are highly dynamic processes and are mediated by various enzymes called histone de/ acetylases and de/methylases. We here for the first time report the regulation of histone de/acetylases and de/methylases; Myst4, Jmjd2b, Aof1 and Set by H3Ac, H3K4me1 and H3K9me1 under hyperinsulinemic/hyperglycemic condition. Functionally, Myst4 is a histone acetyl transferase, Set is an inhibitor of histone acetylation, Jmjd2b is a histone demethylase (specifically demethylate H3K9) and Aof1 is a histone H3K4 demethylase [34e37]. The decreased mRNA expression of Myst4 and Jmjd2b and increased expression of Aof1 and Set under hyperinsulinemic/hyperglycemic condition support our previous findings that there is global down regulation of H3 acetylation/H3K4me1 levels and an increase in H3K9me1 levels both in vitro and in vivo [27,38e40]. Furthermore, ChIP-chip and ChIPeqPCR data demonstrates the role of histone modifications (H3Ac, H3K4me1 and H3K9me1) in regulating the expression of genes responsible for chromatin remodelling, suggesting existence of a cross-talk between different histone modifications. Further, comparing the relative abundance of H3Ac, H3K4me1 and H3K9me1 across the promoter region of
chromatin remodelling genes, between the high glucose and low glucose condition after insulin treatment, reveals a characteristic pattern for these histone modifications. Interestingly, similar pattern of H3Ac and H3K4me1 was observed at the promoter regions of the genes responsible for histone modifications, but H3K9me1 showed an opposite pattern. The characteristic increase in the H3Ac and H3K4me1 and decrease in H3K9me1 on the promoter region (1 to 1000 bp) from the transcription start site for the up-regulated genes Aof1 and Set and vice-versa for the down-regulated genes Myst4 and Jmjd2b, suggest the critical role of these histone modifications in regulating the levels of these genes under hyperglycemic/hyperinsulinemic conditions. These findings highlight the role of histone modifications (H3Ac, H3K4me1 and H3K9me1) in regulating expression of these genes by altering its relative abundance at their promoter regions. Further, comparison of the gene expression related to transcriptional category by coclustering reveals the crosstalk between different histone modifications on the promoter region of these genes. In most of the down regulated genes, decreased pattern of H3k4me1 and H3Ac was found, whereas increased H3K9me1 abundance was associated with decreased mRNA level. Only for a small set of genes, H3k9me1 abundance was associated with increased mRNA expression (Supplementary Fig. 1).
J. Gupta et al. / Biochimie 94 (2012) 2656e2664
2663
Fig. 7. Relative abundance of histone modifications on promoter of histone de/methylases Jmjd2b and Aof1: 3T3 cells were cultured as described in the “Materials and methods”. qPCR was performed across the promoter region of Jmjd2b and Aof1. All distances are relative to transcription start site. Plot shows H3Ac, H3K4me1 and H3K9me1 levels on a: Jmjd2b promoter under low glucose condition after 30 min of insulin stimulation, b: Jmjd2b promoter under high glucose condition after 30 min of insulin stimulation, c: Aof1 promoter under low glucose condition after 30 min of insulin stimulation and d: Aof1 promoter under high glucose condition after 30 min of insulin stimulation. Similar results were obtained in the three independent sets of experiments.
In summary, our work highlights the changes in H3K4me1 and H3K9me1 on the transcribed regions of the genes after providing diabetes related insults and also their relative abundance across the promoter region of the genes responsible for altering histone
modifications. Moreover, we also provide evidence for the presence of cross-talk between different histone modifications under diabetic conditions. The understanding of these epigenetic alterations may therefore lead to additional therapeutic strategies.
I
High Glucose Condition
I
Plasma Membrane IR 30 minutes
H3 A Acetylation t l ti H3K4 de/methylase
H3 de/acetylase
H3K9 H3K9me
H3K4 H3K4me
Nucleus
H3K9 de/methylase
Chromatin compaction
Altered gene expression
Fig. 8. Flowchart summarizing the cross-talk between insulin-induced epigenetic modifications under hyperglycemic condition.
2664
J. Gupta et al. / Biochimie 94 (2012) 2656e2664
Author contributions J.G. researched data, wrote manuscript. S.K. contributed in gene co-clustering and preparing gene association network. J.L. helps in data analysis. R.K.M. contributed to discussion, reviewed/edited manuscript. K.T. contributed to discussion, reviewed/edited manuscript and provides final approval of the version to be published. Duality of interest The authors declare that no conflict of interest exists with the publication of this work. Acknowledgements This work was supported in part by a grant from National Institute of Pharmaceutical Education and Research (NIPER) and a grant from Department of Biotechnology, Government of India. Mrs. Jeena Gupta was supported by Senior Research Fellowship (SRF) from Council of Scientific and Industrial Research, India (CSIR) (Award No. 09/727(0041)/2005-EMR-I). Appendix A. Supplementary material Supplementary data related to this article can be found online at http://dx.doi.org/10.1016/j.biochi.2012.08.011. References [1] A.P. Wolffe, J.J. Hayes, Chromatin disruption and modification, Nucleic Acids Res. 27 (1999) 711e720. [2] B.D. Strahl, C.D. Allis, The language of covalent histone modifications, Nature 403 (2000) 41e45. [3] M. Iizuka, M.M. Smith, Functional consequences of histone modifications, Curr. Opin. Genet. Dev. 13 (2003) 154e160. [4] T. Jenuwein, C.D. Allis, Translating the histone code, Science 293 (2001) 1074e1080. [5] B.M. Turner, Cellular memory and the histone code, Cell 111 (2002) 285e291. [6] M. Grunstein, Histone acetylation in chromatin structure and transcription, Nature 389 (1997) 349e352. [7] M. Lachner, T. Jenuwein, The many faces of histone lysine methylation, Curr. Opin. Cell. Biol. 14 (2002) 286e298. [8] K. Struhl, Histone acetylation and transcriptional regulatory mechanisms, Genes Dev. 12 (1998) 599e606. [9] P. Chi, C.D. Allis, G.G. Wang, Covalent histone modifications e miswritten, misinterpreted and mis-erased in human cancers, Nat. Rev. Cancer 10 (2010) 457e469. [10] B.M. Lee, L.C. Mahadevan, Stability of histone modifications across mammalian genomes: implications for ‘epigenetic’ marking, J. Cell. Biochem. 108 (2009) 22e34. [11] C.R. Vakoc, S.A. Mandat, B.A. Olenchock, G.A. Blobel, Histone H3 lysine 9 methylation and HP1gamma are associated with transcription elongation through mammalian chromatin, Mol. Cell. 19 (2005) 381e391. [12] C.R. Vakoc, M.M. Sachdeva, H. Wang, G.A. Blobel, Profile of histone lysine methylation across transcribed mammalian chromatin, Mol. Cell. Biol. 26 (2006) 9185e9195. [13] A.J. Bannister, R. Schneider, F.A. Myers, A.W. Thorne, C. Crane-Robinson, T. Kouzarides, Spatial distribution of di- and tri-methyl lysine 36 of histone H3 at active genes, J. Biol. Chem. 280 (2005) 17732e17736. [14] R. Schneider, A.J. Bannister, F.A. Myers, A.W. Thorne, C. Crane-Robinson, T. Kouzarides, Histone H3 lysine 4 methylation patterns in higher eukaryotic genes, Nat. Cell. Biol. 6 (2004) 73e77. [15] Y.C. Lin, S. Jhunjhunwala, C. Benner, S. Heinz, E. Welinder, R. Mansson, M. Sigvardsson, J. Hagman, C.A. Espinoza, J. Dutkowski, T. Ideker, C.K. Glass, C. Murre, A global network of transcription factors, involving E2A, EBF1 and Foxo1, that orchestrates B cell fate, Nat. Immunol. 11 (2010) 635e643. [16] K. van Dijk, K.E. Marley, B.R. Jeong, J. Xu, J. Hesson, R.L. Cerny, J.H. Waterborg, H. Cerutti, Monomethyl histone H3 lysine 4 as an epigenetic mark for silenced euchromatin in Chlamydomonas, Plant Cell 17 (2005) 2439e2453. [17] C.D. Carvin, M.P. Kladde, Effectors of lysine 4 methylation of histone H3 in Saccharomyces cerevisiae are negative regulators of PHO5 and GAL1-10, J. Biol. Chem. 279 (2004) 33057e33062.
[18] J.P. Jackson, L. Johnson, Z. Jasencakova, X. Zhang, L. PerezBurgos, P.B. Singh, X. Cheng, I. Schubert, T. Jenuwein, S.E. Jacobsen, Dimethylation of histone H3 lysine 9 is a critical mark for DNA methylation and gene silencing in Arabidopsis thaliana, Chromosoma 112 (2004) 308e315. [19] A.H. Peters, S. Kubicek, K. Mechtler, R.J. O’Sullivan, A.A. Derijck, L. PerezBurgos, A. Kohlmaier, S. Opravil, M. Tachibana, Y. Shinkai, J.H. Martens, T. Jenuwein, Partitioning and plasticity of repressive histone methylation states in mammalian chromatin, Mol. Cell. 12 (2003) 1577e1589. [20] A. Lomba, F.I. Milagro, D.F. Garcia-Diaz, A. Marti, J. Campion, J.A. Martinez, Obesity induced by a pair-fed high fat sucrose diet: methylation and expression pattern of genes related to energy homeostasis, Lipids Health Dis. 9 (2010) 60. [21] C.P. Mack, An epigenetic clue to diabetic vascular disease, Circ. Res. 103 (2008) 568e570. [22] F. Miao, R. Natarajan, Mapping global histone methylation patterns in the coding regions of human genes, Mol. Cell. Biol. 25 (2005) 4650e4661. [23] Z. He, T. Jiang, Z. Wang, M. Levi, J. Li, Modulation of carbohydrate response element-binding protein gene expression in 3T3-L1 adipocytes and rat adipose tissue, Am. J. Physiol. Endocrinol. Metab. 287 (2004) E424eE430. [24] I. Matias, M.P. Gonthier, P. Orlando, V. Martiadis, L. De Petrocellis, C. Cervino, S. Petrosino, L. Hoareau, F. Festy, R. Pasquali, R. Roche, M. Maj, U. Pagotto, P. Monteleone, V. Di Marzo, Regulation, function, and dysregulation of endocannabinoids in models of adipose and beta-pancreatic cells and in obesity and hyperglycemia, J. Clin. Endocrinol. Metab. 91 (2006) 3171e3180. [25] V.G. Tusher, R. Tibshirani, G. Chu, Significance analysis of microarrays applied to the ionizing radiation response, Proc. Natl. Acad. Sci. U S A 98 (2001) 5116e5121. [26] J. Li, J. Liu, R.K.M. Karuturi, Stepped Linear Regression to Accurately Assess Statistical Significance in Batch Confounded Differential Expression Analysis, in: Bioinformatics Research and Applications, Lecture Notes in Computer Science, Springer. 4983 (2008) 481e491. [27] D.G. Kabra, J. Gupta, K. Tikoo, Insulin induced alteration in post-translational modifications of histone H3 under a hyperglycemic condition in L6 skeletal muscle myoblasts, Biochim. Biophys. Acta 1792 (2009) 574e583. [28] F. Miao, D.D. Smith, L. Zhang, A. Min, W. Feng, R. Natarajan, Lymphocytes from patients with type 1 diabetes display a distinct profile of chromatin histone H3 lysine 9 dimethylation: an epigenetic study in diabetes, Diabetes 57 (2008) 3189e3198. [29] F. Miao, X. Wu, L. Zhang, Y.C. Yuan, A.D. Riggs, R. Natarajan, Genome-wide analysis of histone lysine methylation variations caused by diabetic conditions in human monocytes, J. Biol. Chem. 282 (2007) 13854e13863. [30] A. El-Osta, D. Brasacchio, D. Yao, A. Pocai, P.L. Jones, R.G. Roeder, M.E. Cooper, M. Brownlee, Transient high glucose causes persistent epigenetic changes and altered gene expression during subsequent normoglycemia, J. Exp. Med. 205 (2008) 2409e2417. [31] D. Schubeler, D.M. MacAlpine, D. Scalzo, C. Wirbelauer, C. Kooperberg, F. van Leeuwen, D.E. Gottschling, L.P. O’Neill, B.M. Turner, J. Delrow, S.P. Bell, M. Groudine, The histone modification pattern of active genes revealed through genome-wide chromatin analysis of a higher eukaryote, Genes Dev. 18 (2004) 1263e1271. [32] J.K. Sims, S.I. Houston, T. Magazinnik, J.C. Rice, A trans-tail histone code defined by monomethylated H4 Lys-20 and H3 Lys-9 demarcates distinct regions of silent chromatin, J. Biol. Chem. 281 (2006) 12760e12766. [33] J.C. Rice, S.D. Briggs, B. Ueberheide, C.M. Barber, J. Shabanowitz, D.F. Hunt, Y. Shinkai, C.D. Allis, Histone methyltransferases direct different degrees of methylation to define distinct chromatin domains, Mol. Cell. 12 (2003) 1591e1598. [34] N. Pelletier, N. Champagne, S. Stifani, X.J. Yang, MOZ and MORF histone acetyltransferases interact with the Runt-domain transcription factor Runx2, Oncogene 21 (2002) 2729e2740. [35] S.B. Seo, P. McNamara, S. Heo, A. Turner, W.S. Lane, D. Chakravarti, Regulation of histone acetylation and transcription by INHAT, a human cellular complex containing the set oncoprotein, Cell 104 (2001) 119e130. [36] Y. Katoh, M. Katoh, Comparative integromics on JMJD2A, JMJD2B and JMJD2C: preferential expression of JMJD2C in undifferentiated ES cells, Int. J. Mol. Med. 20 (2007) 269e273. [37] Z. Yang, J. Jiang, D.M. Stewart, S. Qi, K. Yamane, J. Li, Y. Zhang, J. Wong, AOF1 is a histone H3K4 demethylase possessing demethylase activity-independent repression function, Cell. Res. 20 (2010) 276e287. [38] K. Tikoo, R.L. Meena, D.G. Kabra, A.B. Gaikwad, Change in post-translational modifications of histone H3, heat-shock protein-27 and MAP kinase p38 expression by curcumin in streptozotocin-induced type I diabetic nephropathy, Br. J. Pharmacol. 153 (2008) 1225e1231. [39] K. Tikoo, D.N. Tripathi, D.G. Kabra, V. Sharma, A.B. Gaikwad, Intermittent fasting prevents the progression of type I diabetic nephropathy in rats and changes the expression of Sir2 and p53, FEBS Lett. 581 (2007) 1071e1078. [40] A.B. Gaikwad, J. Gupta, K. Tikoo, Epigenetic changes and alteration of Fbn1 and Col3A1 gene expression under hyperglycaemic and hyperinsulinaemic conditions, Biochem. J. 432 (2010) 333e341. [41] P. Chomczynski, N. Sacchi, Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction, Anal. Biochem. 162 (1987) 156e159.