XBP1 maintains beta cell identity, represses beta-to-alpha cell transdifferentiation and protects against diabetic beta cell failure during metabolic stress in mice

Aims/hypothesis Pancreatic beta cell dedifferentiation, transdifferentiation into other islet cells and apoptosis have been implicated in beta cell failure in type 2 diabetes, although the mechanisms are poorly defined. The endoplasmic reticulum stress response factor X-box binding protein 1 (XBP1) is a major regulator of the unfolded protein response. XBP1 expression is reduced in islets of people with type 2 diabetes, but its role in adult differentiated beta cells is unclear. Here, we assessed the effects of Xbp1 deletion in adult beta cells and tested whether XBP1-mediated unfolded protein response makes a necessary contribution to beta cell compensation in insulin resistance states. Methods Mice with inducible beta cell-specific Xbp1 deletion were studied under normal (chow diet) or metabolic stress (high-fat diet or obesity) conditions. Glucose tolerance, insulin secretion, islet gene expression, alpha cell mass, beta cell mass and apoptosis were assessed. Lineage tracing was used to determine beta cell fate. Results Deletion of Xbp1 in adult mouse beta cells led to beta cell dedifferentiation, beta-to-alpha cell transdifferentiation and increased alpha cell mass. Cell lineage-specific analyses revealed that Xbp1 deletion deactivated beta cell identity genes (insulin, Pdx1, Nkx6.1, Beta2, Foxo1) and derepressed beta cell dedifferentiation (Aldh1a3) and alpha cell (glucagon, Arx, Irx2) genes. Xbp1 deletion in beta cells of obese ob/ob or high-fat diet-fed mice triggered diabetes and worsened glucose intolerance by disrupting insulin secretory capacity. Furthermore, Xbp1 deletion increased beta cell apoptosis under metabolic stress conditions by attenuating the antioxidant response. Conclusions/interpretation These findings indicate that XBP1 maintains beta cell identity, represses beta-to-alpha cell transdifferentiation and is required for beta cell compensation and prevention of diabetes in insulin resistance states. Graphical abstract Supplementary Information The online version of this article 10.1007/s00125-022-05669-7 contains peer-reviewed but unedited supplementary material.


Islet isolation
Pancreas was perfused via the common bile duct with cold Liberase solution (Krebs-Ringer HEPES buffer, KRB-HEPES: 125 mmol/l NaCl, 4.8 mmol/l KCl, 1 mmol/l CaCl2, 1.2 mmol/l KH2(PO)4, 1.18 mmol/l MgSO4, 5 mmol/l NaHCO3 and 25 mmol/l HEPES, supplemented with 2.8 mmol/l glucose and 0.25 mg/ml Liberase (Roche). The pancreas was removed and incubated at 37°C for 16 min. The digestion was stopped by addition of KRB-HEPES containing 10 % newborn calf serum followed by mechanical disruption of the pancreas and washing of the islets. Islets were further separated using a Ficoll-Paque PLUS (GE Healthcare, Uppsala, Sweden) density gradient with centrifugation followed by handpicking under a stereomicroscope.

Glucose-stimulated insulin secretion ex vivo
Experiments were performed in batches of 5 islets in triplicate. Islets were pre-incubated in 100 µl of KRB-HEPES supplemented with 0.1 % BSA and 2 mmol/l glucose for 30 min at 37°C in a 96-well V-bottom microplate. Then they were incubated in 100 µl of fresh KRB-HEPES supplemented with 0.1 % BSA containing either 2 or 20 mmol/l glucose for 1 h at 37°C. An aliquot of the buffer was taken, and secreted insulin determined using the Insulin Ultra-Sensitive Assay (Cisbio, Codolet, France). The islets were lysed for measurement of insulin and DNA content.

Histology and immunohistochemistry
Pancreases were weighed and fixed in 10 % neutral-buffered formalin solution overnight at 4°C, and then stored in 70 % ethanol at 4°C. The pancreata were embedded in paraffin and sections were cut at 4-5 µm for histology and immunohistochemistry. Sections were deparaffinised and rehydrated using Leica Autostainer (Leica ST5010, Leica Microsystems, Wetzlar, Germany). Slides were baked for 4 min at 60°C and dewaxed in Xylene twice for 3 min. Then, slides were washed in 100 % ethanol for 3 min and rehydrated by washing in 95 %, 70 % and 50 % ethanol for 3 min each. Slides were then rinsed in distilled water. Heatinduced antigen retrieval was performed using Targeted retrieval solution (S1699) and heating the slides to 125°C for 1 min and 95°C for 10 sec in a pressure cooker. Alternatively, enzymatic antigen retrieval with proteinase K at 37°C for 30 min was used for glucagon staining. Slides were then washed in PBS and TBST (0.1 % Tween-20) for 5 min and permeabilised in TBS + 1 % Triton X-100 for 5 min. They were then rinsed in TBST 3 times for 5 min to remove the triton solution. The slides were then blocked with PBS containing 2 % BSA and 5 % goat serum for 30 min at room temperature and then incubated with respective primary antibody overnight at 4°C (ESM Table 1). The primary antibody was removed and slides washed with TBST 3 times, followed by incubation with the secondary antibody for 1 h at room temperature avoiding light (ESM Table 1). The secondary antibody was washed with TBST 3 times and the slides mounted with a DAPI mounting media (ProLong Gold, Life Technologies, Carlsbad, CA, USA), cover slipped and allowed to dry. For in situ apoptosis detection, formalin-fixed, paraffin-embedded pancreatic sections were stained using mouse anti-insulin antibody and the Click-iT TM Plus TUNEL Assay (ThermoFisher Scientific, Scoresby, VIC, Australia) according to manufacturer's instructions. Sections were mounted with DAPI mounting media (ProLong Gold), cover slipped and allowed to dry. Slides were imaged using a Leica DM5500 fluorescent microscope or Leica DM6000 Power Mosaic microscope (Leica Microsystems).

Immuno-morphometry
Immunostaining was quantified using semi-automated scripts and plugin commands written for Image J/FIJI (ImageJ Developers and FIJI Contributors). Immunofluorescent images of insulin and glucagon staining were used to outline an insulin-defined and glucagon-defined region of interest (ROI), respectively. ROIs were used to calculate beta cell and alpha cell area (µm 2 ) for each section, and beta cell and alpha cell mass calculated using the area and total pancreas mass. Beta cell proliferation and apoptosis were assessed by overlaying the insulindefined ROI with images of DAPI, Ki-67 and TUNEL stained sections. Three pancreas sections separated by at least 100 µm were used for quantification.

Electron microscopy
Freshly isolated islets were fixed in 2 % glutaraldehyde in 0.1 M cacodylate buffer, pH 7.4, overnight at 4°C. After consecutive washes in 0.1 M cacodylate buffer, the islet pellets were postfixed in 4 % osmium in the cacodylate buffer for 1 h at room temperature. Two washes in distilled water were followed by a block staining with 2 % aqueous uranyl acetate. The islets were then dehydrated through a series of ethanol and embedded in TAAB epoxy resin. Ultrathin sections (90 nm) were cut, mounted on 150-mesh copper/palladium grids and stained with Reynolds lead citrate. 10-12 representative digital electron micrographs of individual beta cells were captured from 5-6 different islets under transmission electron microscope (JEOL 1011; JEOL Ltd, Tokyo, Japan) using a digital camera (MegaView G2; Olympus, Münster, Germany). Images were analyzed with iTEM software v.5.2 (Olympus). Mature and immature insulin granules (IG) were identified based on core density and presence of characteristic halo.

RNA analysis
Total RNA from islets was extracted using RNeasy Mini Kit according to manufacturer's instructions (Qiagen, Doncaster, VIC, Australia). RNA concentration was determined using a Nanodrop spectrophotometer. RNA (200 ng) was reverse transcribed to cDNA using the QuantiTect Reverse Transcription Kit (Qiagen) according to manufacturer's instructions. Realtime PCR was performed in a 384-well plate on the 7900 HT Real Time PCR System (Applied Biosystems, Foster City, CA, USA) using standard reaction cycle conditions. The 10 μl reaction volume consisting of cDNA, 0.6 µmol/l oligonucleotide primers (ESM Table 2) and PowerSYBR Green master mix was prepared using automated pipetting on the epMotion 5070 (Eppendorf, Hamburg, Germany). The value obtained for each specific gene product was normalised to a housekeeping gene (cyclophilin A) and expressed as a fold-change of the value in control extracts.

Translating ribosome affinity purification
Translating ribosome affinity purification (TRAP) was performed based on the previously published protocol with some modifications [2,3]. Islets isolated from β-Xbp1 +/+ Gt and β-Xbp1 -/-Gt mice were homogenized in 500 µl lysis buffer (20 mM HEPES-KOH [pH 7.4], 5 mmol/l MgCl2, 150 mmol/l KCl) supplemented with 0.5 mmol/l DTT, fresh protease inhibitor (1 tablet/ml; Roche Mini Complete, EDTA-Free), 100 µg/ml cycloheximide and 40 U/ml Rnasin (Promega, Madison, WI, USA). After lysis, homogenates were centrifuged for 10 min at 2000 x G at 4°C to remove pellet nuclei and cell debris. 50 µl of 10 % NP-40 working solution (Biochemica) was added to the supernatant and mixed by gentle inversion. Next, 50 µl of DHPC (300 mmol/l) was added to the supernatant and mixed by gentle inversion and incubated on ice for 5 min. The lysate was centrifuged for 10 min at 13,000 x G. 20% (100 µl) of the supernatant was kept as input. For preparation of the GFP antibody-Dynabeads solution, 50 µl of protein G Dynabeads (Invitrogen, Carlsbad, CA, USA) was rinsed three times with 1 ml of 0.15 mol/l KCl buffer (20 mmol/l HEPES-KOH [pH 7.4], 5 mmol/l MgCl2, 150 mmol/l KCl, 1 % NP-40, 0.5 mmol/l DTT, 100 µg/ml cycloheximide) at room temperature. Then 5 µl of anti-GFP antibody (2 µg/µl; Invitrogen) was added to the beads and incubated with 275 µl of 0.15 mol/l KCl buffer for 1 h at room temperature with slow end-to-end rotation. Next, the antibody-bound beads were collected using a magnetic rack and the supernatant discarded. The collected beads were washed three times with 0.15 mol/l KCl buffer. The antibody-bead complex was resuspended in 200 µl of 0.15 mol/l KCl buffer before use. The beads were then mixed with the cell-lysate supernatant, and the mixture was incubated at 4°C with slow end-to-end rotation overnight. The complex-bound Dynabeads were collected with a magnetic rack, washed 3 times with 0.35 mol/l KCl buffer (20 mmol/l HEPES-KOH [pH 7.4], 5 mmol/l MgCl2, 350 mmol/l KCl, 1 % NP-40, 0.5 mmol/l DTT, 100 µg/ml cycloheximide) at 4°C, and immediately resuspended in 350 µl of RLT buffer supplemented with 10 % 2mercaptoethanol and incubated for 5 min at room temperature. The beads were removed using the magnetic rack and the RLT-containing RNA was precipitated with equal parts of 70% ethanol and then purified using the RNeasy Micro Kit (Qiagen) according to manufacturer's instructions. RNA concentration was determined using a Nanodrop spectrophotometer. 40 ng of RNA was used to synthesize cDNA using the QuantiTect Reverse Transcription Kit (Qiagen) according to manufacturer's instructions for subsequent RT-PCR analysis.

Statistical analysis
All data are represented as means ± SEM. Unpaired two-tailed t-test was used to compare differences between two groups. Differences between more than two groups were calculated using two-way ANOVA with Tukey's post-hoc test.  Fig. 1. XBP1 protein levels in islets of β-Xbp1 +/+ and β-Xbp1 -/mice fed a chow diet.