Fig 2: NKX6.1 binds directly to the miR-184 promoter.a–c ChIP-qPCR of NKX6.1 binding to the human miR-184 promoter sequence; data are shown as fold change in miR-184 promoter sequence bound by anti-NKX6.1 antibody compared to isotypic IgG negative control (a). ChIP-qPCR of NKX6.1 binding (b) or histone H3 acetylation (AcH3) (c) at the miR-184 promoter, with or without NKX6.1 siRNA and normalised to IgG control; values are shown as fold change in EndoC-ßH1 cells transfected with NKX6.1 siRNA compared to scrambled siRNA; statistics using Mann–Whitney U test (n = 4–5 independent experiments). d–f ChIP-qPCR of NKX6.1 binding to the selected sequence of the murine miR-184 promoter; data are shown as fold change of anti-NKX6.1 antibody compared to isotypic negative control IgG (d); statistics using Mann–Whitney U test (n = 4 independent experiments). ChIP-qPCR of NKX6.1 binding (e) or histone H3 acetylation (AcH3) (f) at the murine miR-184 promoter, with or without 100 µM H2O2 treatment in MIN6 cells; values are shown as fold change of 100 µM H2O2 treated samples compared to untreated samples; statistics using Mann–Whitney U test (n = 4 independent experiments).

Fig 3: NKX6.1-miR-184-3p-CRTC1 axis.Working model of upstream and downstream miR-184-3p molecular mechanisms in ß-cells protection.

Fig 4: NKX6.1 regulates human and murine miR-184-3p expression.a Left panel: table showing predicted binding sites of NKX6.1 in the human and mouse miR-184 gene promoter using MatInspector algorithm. For each predicted binding site, the following parameters are given: species, start/end nucleotide (Start Pos, End Pos) of the TF matrix (with respect to –500 bp of miR-184 TSS, see Sequence in Supplementary material), sequence strand, matrix similarity (SIM.), optimised matrix similarity threshold (OPT.) and target sequence (see Supplementary File 2). Right panel: representative schematic showing human and mouse MIR184 gene promoter and NKX6.1 binding sites (in red). b RT-Real-time PCR expression analysis of NKX6.1 in human pancreatic islets with T2D (n = 6; dark grey squares) compared to non-diabetic donors (n = 10; light grey circles); data are shown as 2-?Ct relative expression; statistics with Mann–Whitney U test. c Correlation analysis between expression of NKX6.1 and miR-184-3p in human pancreatic islets from T2D and non-diabetic donors (n = 16). Values are expressed as 2-?Ct (log10) relative expression; p value and r value by Spearman R correlation test. d, e 48 h silencing of NKX6.1 with siRNA (25 nM) in EndoC-ßH1 cell line showed reduction in mRNA (d) and protein (e) expression; statistics by Mann–Whitney U test (n = 4). f RT-Real-time PCR of miR-184-3p expression in EndoC-ßH1 after 48 h NKX6.1 siRNA transfection; data are expressed as mean ± SD of 2-?Ct normalised values; statistics with Mann–Whitney U test (n = 4 independent experiments). g Immunofluorescence analysis of Nkx6.1 (green) and nuclei (DAPI, blue) in MIN6 cells treated or not with 100 µM H2O2; panel A (Nkx6.1) to panel D show MIN6 cells not treated; panel E to panel H show MIN6 cells treated with H2O2; scale bar = 25 µm. h Analysis of colocalisation coefficient for quantification of NKX6.1 and nuclear staining with or without treatment with 100 µM of H2O2. Data are expressed as mean ± SD of the percentage of NKX6.1 signal overlapping nuclear staining with DAPI. Statistics with Mann–Whitney U test (n = 5 independent experiments). i RT-Real-time PCR analysis of miR-184-3p expression in MIN6 cells after 100 µM of H2O2 treatment. Data are presented as mean ± SD of fold change of MIN6 cells treated with H2O2 compared to untreated samples; statistics with Mann–Whitney U test (n = 5 independent experiments).

Fig 5: Assays of β-cell–specifying transcription factor Nkx6.1. (a) Relative Nkx6.1 to β-actin ratio in PI of Wistar (black) and GK rats (gray). (c,d) Nkx6.1/β-actin transcript ratio in PI vs. blood glucose (b) or body weight (c) in Wistar (black) and GK rats (gray and white) at age 6 (white squares; black diamonds); 15 (circles) and 37 weeks (triangles). ***P < 0.001. (b) Protein amounts for Nkx6.1 as derived from five Western blots. Wistar and GK rat PI samples were always run and are displayed together (as illustrated), while the alternative development for β-actin is displayed below. Typical full-length blots and Ponceau Red-stained membranes are included in a Supplementary Information file).