Fig 2: Granule acidification is blocked in HID-1 KO cells.Indirect immunofluorescence was performed on dispersed ß islet cells, after 1 hr of incubation with 30 µM DAMP in Hanks balanced salt solutions (4 mM glucose). Anti-DNP was used to detect DAMP. Anti-insulin was used to label insulin-containing granules. (A) 3D-SIM images of WT ß cells, showing colocalization of DAMP (green) with insulin granules (red). Scale bar, 2 µm. (B) Comparison of DAMP intensity between WT (n = 2443) and KO (n = 2117) insulin granules (imaged by confocal microscopy, 50 cells for each group and three biological replicates with six mice). The difference is significant (p < 0.001, Kolmogorov-Smirnov test). (C) Mean fluorescence intensity of DAMP from insulin-immunoreactive cells was analyzed by flow cytometry (four biological replicates with eight mice, p = 0.015, Wilcoxon test).DOI: http://dx.doi.org/10.7554/eLife.18134.018

Fig 3: HID-1 deficiency disrupts homotypic fusion of ISGs.(A) Representative sections from FIB/SEM reconstructions containing ISG fusion events. The arrowed fusion events are segmented and rendered in 3D (top corner). Scale bar, 500 nm. (B) Comparison of the frequency of fusion events. There were 8.2 ± 0.8 fusion events per 1,000 SGs in WT (19/2632, 4/532, 22/2253, n = 3 cells) and 1.2 ± 0.4 in KO (3/3024, 1/1043, 3/1539, n = 3 cells), p = 0.002, t-test. (C) An overview flowchart of the in vitro fusion assay. (D) Representative images of fusion reactions using either WT cells or HID-1 KO INS-1 cells. Fusion events are marked with white circles. Scale bar, 3 μm. (E) A typical fusion event as shown in panel (D). The graph shows the fluorescence intensity profiles of green SGs (ISG) and red SGs (PNS). (F) ISGs and PNS obtained from both WT and HID-1 KO INS-1 cells were subjected to in vitro fusion assays (n = 3) with or without ATP at 37°C or on ice. The fusion rate was calculated by dividing the number of fusion events by the total number of ISGs. (p = 0.03, 0.007 and 0.009 for WT37°C-ATP, WTice+ATP and KO37°C+ATP vs WT37°C+ATP, respectively, t-test. N.S = not significant.)DOI: http://dx.doi.org/10.7554/eLife.18134.016

Fig 4: Hid1-betaKO mice exhibit glucose intolerance.(A) A schematic diagram showing the endogenous mouse Hid-1 gene and the targeting floxed construct. (B) Western blot analysis of mouse HID-1 and actin in isolated islets from WT and Hid1-betaKO (KO) mice (five biological replicates from ten mice). (C) Quantitative analysis of the reduction of HID-1 protein in KO islets. (D) Immunofluorescent histochemical analysis of pancreas sections of WT and KO mice, showing glucagon (green), insulin (red) and HID-1 (cyan). The white squares were enlarged in Figure 1—figure supplement 1A for a magnified view. Scale bar: 20 µm. (E) Comparison of body weight (n = 6, three males and three females in each group) and (F) blood glucose during GTT (i.p. administration of 2 g/kg body weight glucose. WT, n = 8, four males and four females; KO, n = 11, five males and six females. p=0.003, 0.005, 0.001, 0.02 and 0.002 at 0, 15, 30, 60 and 120 min, respectively, t-test.) and (G) normalized blood glucose during ITT (i.p. administration of 0.5 U/kg body weight insulin. WT, n = 9, four males and five females; KO, n = 10, five males and five females) between WT and KO mice.DOI: http://dx.doi.org/10.7554/eLife.18134.002

Fig 5: HID-1 deficiency causes a defect in proinsulin processing.(A) Western blotting of islets from WT and KO mice with an anti-insulin polyclonal antibody. The proinsulin level in KO mice is higher than that in WT mice. Results are representative of three experiments from six mice. (B) Pulse-chase experiment in WT and KO mice islets using radioactive [35S]-cysteine labeling. Insulin and proinsulin-related proteins were immunoprecipitated with anti-insulin polyclonal antibody coupled to Dynabeads. Densitometry of 35S-labeled insulin species was performed using a PhosphorImager. Results are representative of three experiments with three mice per group. (C) Kinetics of proinsulin processing in WT and KO mice islets (three biological replicates from nine mice, p = 0.006 at 180 min, t-test). (D) Insulin/proinsulin ratio during proinsulin processing in WT and KO mice islets (three biological replicates from nine mice, p = 0.006 at 180 min, t-test). (E) Western blotting with anti-HID-1 polyclonal antibody and anti-insulin polyclonal antibody in WT and HID-1 KO rat INS-1 cells. Absence of HID-1 confirms the KO effect. Proinsulin is significantly higher in HID-1 KO cells than in WT mice. (F) Kinetics of proinsulin1 processing analyzed with pulse-chase experiments followed by LC-MS quantification using stable isotopic [13C,15N]-isoleucine labeling in WT and HID-1 KO rat INS-1 cells (WT, n = 4; KO, n = 6, p = 0.008, <0.001 and 0.002 at 50, 100, 150 min, respectively, t-test). (G) Kinetics of insulin1 production in WT and HID-1 KO rat INS-1 cells (p = 0.005, 0.036 and 0.038 at 50, 100 and 150 min, respectively, t-test). (H) Insulin1/proinsulin1 ratio during proinsulin processing in WT and HID-1 KO rat INS-1 cells (p = 0.01, 0.01, and 0.019 at 50, 100 and 150 min, respectively, Mann-Whitney U Test).DOI: http://dx.doi.org/10.7554/eLife.18134.008