Beta cell dysfunction induced by bone morphogenetic protein (BMP)-2 is associated with histone modifications and decreased NeuroD1 chromatin binding

Abstract

Insufficient insulin secretion is a hallmark of type 2 diabetes and has been attributed to beta cell identity loss characterized by decreased expression of several key beta cell genes. The pro-inflammatory factor BMP-2 is upregulated in islets of Langerhans from individuals with diabetes and acts as an inhibitor of beta cell function and proliferation. Exposure to BMP-2 induces expression of Id1-4, Hes-1, and Hey-1 which are transcriptional regulators associated with loss of differentiation. The aim of this study was to investigate the mechanism by which BMP-2 induces beta cell dysfunction and loss of cell maturity. Mouse islets exposed to BMP-2 for 10 days showed impaired glucose-stimulated insulin secretion and beta cell proliferation. BMP-2-induced beta cell dysfunction was associated with decreased expression of cell maturity and proliferation markers specific to the beta cell such as Ins1, Ucn3, and Ki67 and increased expression of Id1-4, Hes-1, and Hey-1. The top 30 most regulated proteins significantly correlated with corresponding mRNA expression. BMP-2-induced gene expression changes were associated with a predominant reduction in acetylation of H3K27 and a decrease in NeuroD1 chromatin binding activity. These results show that BMP-2 induces loss of beta cell maturity and suggest that remodeling of H3K27ac and decreased NeuroD1 DNA binding activity participate in the effect of BMP-2 on beta cell dysfunction.

Fig. 1. BMP-2 induced beta cell dysfunction and inhibition of proliferation.

Mouse islets were exposed to BMP-2 (50 ng/ml) for 10 days or left non-exposed. A Insulin secretion was measured by static batch incubations for 30 min in response to 2 mmol/l glucose followed by 20 mmol/l and normalized to DNA content. Data were shown as means N = 5, **p < 0.01 two-sided paired t-test vs. ctr. B Total insulin content was measured post-GSIS. Data were shown as means for N = 5, two-sided t-test vs ctr. C Beta cell proliferation was examined in whole mouse islets exposed to BMP-2 for 10 days. Proliferation was determined by immunocytochemical staining for Pdx-1 and Click-iT detection of EdU. Results are shown as the percentage of proliferating Pdx-1 positive cells. Data were shown as means for N = 4, *p < 0.05, two-sided paired t-test vs. ctr. D Immunocytochemical and Click-iT staining of whole mouse islets. Cells were stained for Pdx-1 (green) and EdU (red). The data shown is representative. E Apoptosis was measured by cytoplasmic nucleosome levels in lysates from mouse islets exposed as described above. Two-day stimulation with a cytokine mix of IL-1β (300 pg/ml) and IFN-γ (10 ng/ml) served as a positive control for the induction of apoptosis. Data were shown as means for N = 3, two-sided paired t-test vs ctr. F Caspase 3/7 activity was measured in lysates from mouse islets exposed as described above. Data were shown as means for N = 4.

Fig. 2. BMP-2 induced changes of key beta cell gene expression.

Mouse islets were exposed to BMP-2 (50 ng/ml) or left non-exposed for 10 days. Relative mRNA expression of A Ins1, Ins2, Ucn3, MafA, Pdx1, Glp1r, and Hes-1, N = 5–12, B Ki67 and Cdk1, N = 5, and C Hey-1, N = 9 was analyzed using real-time PCR. Expression levels are normalized to control Ppia and Ctr data were set to 1 represented by the dotted horizontal line. Data were shown as means ± SD, *p < 0.05; **p < 0.01, ***p < 0.001 2-sided paired t-test vs. ctr on log10 transformed data.

Fig. 3. BMP-2 exposure affects global RNA expression.

Mouse islets were exposed to BMP-2 (50 pg/ml) or left non-exposed for 10 days. A Volcano plot representing the differential expressed RNA’s between BMP-2 vs. ctr exposed mouse islets. X-axis: log2 gene expression fold-change, Y-axis: −log10 p value. Each dot represents an individual RNA. Gray: not significantly regulated genes, green: FDR <0.05 upregulated genes, and red FDR <0.05 downregulated genes. Selected regulated genes by BMP-2 are marked. B Table showing the summary of genes regulated by BMP-2. C, D Top ten most significantly enriched biological processes GO of BMP-2 upregulated and downregulated genes, respectively. X-axis: Percent of genes in the GO significantly regulated by BMP-2. Circle size: Number of significant BMP-2 regulated genes in the GO.

Fig. 4. BMP-2 regulated gene expression correlates with protein expression.

Mouse islets were exposed to BMP-2 (50 pg/ml) or left non-exposed for 10 days. Mass spectrometry was performed on extracts and A show significantly regulated (up or down) proteins (proteins >2 peptides, FDR <0.05, FC >0.3)) in BMP-2 exposed vs. ctr islets, N = 5. B Correlation plot of Protein vs. RNA regulated by BMP-2. X-axis: Log2 (FC protein BMP-2/ctr), y-axis: Log2 (FC RNA BMP-2/ctr). The top 35 significantly up- and down-regulated proteins are shown. Gray dots represent proteins where the corresponding mRNA was not significantly regulated by BMP-2. Green dots represent proteins where the corresponding mRNA was significantly upregulated by BMP-2 and red dots represent proteins where the corresponding mRNA was significantly downregulated by BMP-2. Spearman correlation test ***p < 0.0001. C Nptx2 and D Syt7 protein expression was measured in whole cell extracts from mouse islets exposed as described above using Western blotting. Representative blots are shown, and Tbp was used as a loading control. Band intensities were quantified using Image Studio and data were shown as fold increase relative to Ctr (Syt7) or as relative A.U. expression (Nptx2). Data show means ± SD of N = 4. *p < 0.05, **p < 0.01, two-sided paired t-test vs ctr.

Fig. 5. BMP-2 regulation of histone modifications.

Analysis of chromatin histone modification in mouse islets exposed to BMP-2 (50 pg/ml) or left non-exposed for 10 days. Heatmap of A H3K27 acetylation, B H3K4 tri-methylation, and C H3K27 tri-methylation from Ctr (left) and BMP-2 exposed islets (right) from −3 kb to +3 kb surrounding the center of the peak. Each horizontal line represents a single histone 3 binding site (peak) and the color scale indicates the H3K27ac, H3K4me3, and H3K27me3 ChIP-seq tags per bp, respectively. H3K27ac scale is 5E-5 cm per peak while for H3K4me3 and H3K27me3 the scale is 0.01 cm per peak. Upper panel: control enriched peaks (downregulated by BMP-2). Lower panel: BMP-2 enriched peaks (upregulated by BMP-2). D Table showing the summary of the number of peaks regulated by BMP-2. E Genome browser snapshot of ChIP-seq for selected genes; Hey-1, Id1, Ucn3, Ki67, and Top2a for the 3 histone modifications (H3K27ac, H3K4me3, and H3K27me3) from ctr and BMP-2 exposed islets.

Fig. 6. BMP-2-induced changes in RNA expression are associated with changes in histone modifications.

Graph showing the correlation between RNA fold-change and A H3K27ac Ctr enriched peaks. B H3K4me3 Ctr enriched peaks. C H3K27ac BMP-2 enriched peaks. D H3K4me3 BMP-2 enriched peaks. X-axis: Log2 (FC RNA) Y-axis: Log2 (FC histone modification). *p < 0.05, ***p < 0.001 Spearman correlation test. E Biological process GO analysis on H3K27ac downregulated peaks associated with TSS of genes. F Biological process GO on H3K27ac BMP-2 upregulated peaks associated with TSS of genes. X-axis: Percent of genes in the GO with a differential peak associated with TSS. Circle: Number of genes in the GO with a BMP-2 regulated differential peak associated with their TSS.

Fig. 7. BMP-2 induced regulation of NeuroD1 chromatin binding.

A Table depicting the top three enriched motifs in H3K27ac Control enriched (downregulated by BMP-2) peaks. B ChIP analysis showing the number of NeuroD1 binding sites (peaks) upregulated and downregulated by BMP-2. C Distribution of NeuroD1 occupied binding sites in the genome and differential NeuroD1 binding sites in response to BMP-2. D Biological processes GO analysis on BMP-2 downregulated NeuroD1 binding sites. X-axis: Percent of all genes in GO, that are near a differentially enriched binding site. Circle: Number of genes in the GO close to a differentially enriched binding site.