Insulin-degrading enzyme ablation in mouse pancreatic alpha cells triggers cell proliferation, hyperplasia and glucagon secretion dysregulation

Aims/hypothesis Type 2 diabetes is characterised by hyperglucagonaemia and perturbed function of pancreatic glucagon-secreting alpha cells but the molecular mechanisms contributing to these phenotypes are poorly understood. Insulin-degrading enzyme (IDE) is present within all islet cells, mostly in alpha cells, in both mice and humans. Furthermore, IDE can degrade glucagon as well as insulin, suggesting that IDE may play an important role in alpha cell function in vivo. Methods We have generated and characterised a novel mouse model with alpha cell-specific deletion of Ide, the A-IDE-KO mouse line. Glucose metabolism and glucagon secretion in vivo was characterised; isolated islets were tested for glucagon and insulin secretion; alpha cell mass, alpha cell proliferation and α-synuclein levels were determined in pancreas sections by immunostaining. Results Targeted deletion of Ide exclusively in alpha cells triggers hyperglucagonaemia and alpha cell hyperplasia, resulting in elevated constitutive glucagon secretion. The hyperglucagonaemia is attributable in part to dysregulation of glucagon secretion, specifically an impaired ability of IDE-deficient alpha cells to suppress glucagon release in the presence of high glucose or insulin. IDE deficiency also leads to α-synuclein aggregation in alpha cells, which may contribute to impaired glucagon secretion via cytoskeletal dysfunction. We showed further that IDE deficiency triggers impairments in cilia formation, inducing alpha cell hyperplasia and possibly also contributing to dysregulated glucagon secretion and hyperglucagonaemia. Conclusions/interpretation We propose that loss of IDE function in alpha cells contributes to hyperglucagonaemia in type 2 diabetes. Graphical abstract Supplementary Information The online version contains peer-reviewed but unedited supplementary material available at 10.1007/s00125-022-05729-y.

for 1 h, and afterwards in 16 mmol/l glucose secretion buffer for 1 h.
To evaluate insulin paracrine effect on glucagon secretion ex vivo, isolated islets were pre-incubated for 1 h at 1 mmol/l glucose "secretion buffer"; then islets were treated with insulin 100 nmol/l (Sigma, USA) in 1 mmol/l glucose "secretion buffer" (control group was treated with vehicle instead of insulin). To evaluate the effect of pharmacological inhibition of IDE on glucagon secretion ex vivo, we used the IDE-specific activity inhibitor 6bK (Tocris Bio-techne, USA). Isolated islets were pre-incubated with 10 µmol/l 6bK and afterwards incubated at 1 and 16 mmol/l glucose "secretion buffer" as described above.
The extracellular medium was collected after each incubation, and glucagon and insulin concentration were measured by ELISA as described above. To determine pancreas glucagon and insulin content, after dissection, whole pancreas was incubated overnight in acid-ethanol buffer (1.5% v/v HCl in EtOH) at 4°C and hormones were measured from supernatant by ELISA.

Flow cytometry analysis in isolated islet cells
After islet isolation, islet cells were dispersed by trypsin, isolated islet cells (IICs) were washed with Flow cytometry analysis (FACS) buffer (PBS containing 1% BSA, 1 mmol/l EDTA and 0.01% sodium azide), and then labelled with Live/Dead Fixable Near-IR Dead Cell stain (Invitrogen, USA) according to kit instructions and fixed with 2% paraformaldehyde 10 min at 4 °C. Cells were permeabilised with 0.5% Triton X100 in FACS buffer for 10 min at room temperature, and then incubated for 30 min with blocking solution (2% normal goat serum in FACS Buffer). IICs were incubated for 30 min at 4 °C in dark with antibodies in blocking solution, 1:1,000 rabbit anti-IDE (Millipore, USA) and 1:1,000 mouse anti-glucagon conjugated with Alexa Fluor® 594 (R&D Systems, USA), and then incubated in dark for 30 min at 4 °C in blocking solution with 1:1,000 goat antirabbit conjugated with Alexa Fluor® 488 (Invitrogen, USA). Cells were stored in FACS buffer at 4 ºC until spectral flow cytometric analysis. Fluorescence minus one (FMO) controls were used to determine positive staining. See ESM Table 2 for antibodies information.
Cells were acquired on the Aurora Spectral Flow Cytometer (CYTEK) and analyzed using Kaluza software (Beckman Coulter) version 2.1.1.

RNA isolation and qRT-PCR
Islet, liver, kidney, muscle and fat RNA was extracted using TRIZol® Reagent (Thermo Fisher Scientific, USA), according to the manufacturer's instructions. Quantification of mRNA levels was determined using NanoDrop™ N-D1000 spectrophotometer. These samples were treated with RapidOut DNA Removal Kit (Thermo Fisher Scientific, USA).
First strand cDNA was synthesized with iScript™ cDNA synthesis kit (Bio-Rad, USA) as described by manufacturer. qPCR was carried out on equal amounts of cDNA in triplicates for each sample using Maxima Probe qPCR Master Mix (Thermo Fisher Scientific, USA) with corresponding TaqMan® Gene Expression Assays (Applied Biosystems, USA) in a thermal cycler (Lightcycler 480, Roche, Switzerland). Data were normalized with the housekeeping gene Rpl18 and relative expression was quantified using the comparative 2 -ΔΔCT method. See ESM Table 1 for primers and TaqMan assays.

Pancreas histomorphometry
Four-month-old mice (A-IDE-KO mice administered tamoxifen at 8 weeks and Ide f/f controls) were euthanized, then their pancreata were dissected, weighed and fixed in 10% neutral buffer formalin overnight at 4°C, embedded into paraffin blocks then cut into 5 µm sections and mounted on glass slides. Staining was performed in two sections per mouse pancreas spaced at least 200 µm apart. To analyze pancreas histomorphometry, sections were incubated with anti-insulin antibody (Abcam, UK) for beta-cell area and anti-glucagon antibody (Abcam, UK) for alpha-cell area, washed, then incubated with HRP-conjugated secondary antibodies, washed and then stained using 3,3′diaminobenzidine tetrahydrochloride (Sigma, USA) as substrate. Finally, sections were counterstained with hematoxylin as previously reported [13]. Images were acquired using a Nikon Eclipse 90i microscope fitted with a Nikon CCD camera (DSRi1), using a 20X objective with transmitted light. beta-cell area, alpha-cell area and islet number were calculated using Image J 1.52p software (NIH, USA) [12,17]. To analyze alpha cell peripheral location, cells not found in the outer layer of the islet or adjacent to it were counted as "delocalised." See ESM Table 2 for antibodies information.

Pancreas immunostaining
A-IDE-KO and control pancreas sections were stained with the following antibodies previously validated by the manufacturer and previous publications [2,12]. Sections were counterstained with nuclear DAPI staining. Fluorescence images of the sections were acquired using the aforementioned Nikon microscope and camera using a 40X objective.
All pictures were obtained using the same exposure conditions. See ESM Table 2 for antibodies information.
Immunofluorescence intensity of IDE, Vamp2 and α-synuclein were quantified by Image J 1.52p software (NIH, USA) using the following protocol: separated photos of glucagon and Vamp2 were taken using a 40X objective. A specific selection of glucagon-stained area was made by applying a suitable threshold and using the "Create selection" tool within the Image J software on the glucagon image. The selection of glucagon area was transferred into the IDE, Vamp2 or α-synuclein image of the same section, then the intensity of this specific area was quantified using the tool "Integrated Density" within the imaging software, thus yielding a measure of IDE, Vamp2 or α-synuclein intensity/glucagon area. In order to measure the intensity by cell, all cells in images glucagon-stained were quantified with the Image J 1.52p software "cell counter" tool, thus yielding a measure of integrated intensity of Vamp2 staining/alpha-cell number.
Solution was incubated at room temperature for 20 min and added to one well. 6 h after transfection 1.5 ml of complete fresh medium was added, then cells were incubated for 48 h.

Alpha-TC-1.9 proliferation studies
To quantify proliferation rates, cells were seeded on coverslips (at least 100,000 cells/coverslip) and incubated with 10 μmol/l BrdU for 6 h. Cells were fixed with 10% formalin for 5 minutes, washed with PBS, immersed in 70% EtOH at 4°C for 30 min; then, cells were immersed in 1 N HCl for 20 min, followed by a washing step with PBS.
To prevent non-specific binding, the cells were incubated for 1 h in "blocking solution" (1% BSA, 0.2% normal goat serum in PBS) at room temperature. Staining was performed using monoclonal anti-BrdU rat antibody (Abcam, UK) at 4ºC overnight. The samples were then incubated with the anti-rat 594 Alexa-fluor-conjugated secondary antibody (Thermoscientific, USA) for 30 min at room temperature and, finally, were mounted onto glass slides using Fluoroshield with DAPI (Sigma, USA). Images were acquired using the aforementioned Nikon microscope and camera with a 40X objective.
BrdU positive cells were quantified using the free software Image J (NIH, USA).
To detect the presence of primary cilia in proliferating cells, they were seeded and fixed as described above, after "blocking solution" incubation cells were treated with anti-BrdU antibody (Abcam, UK) for 1 hour and subsequently with anti-α-acetylated α-tubulin (Sigma, USA) antibody at 4°C overnight. The next day the samples were washed with PBS to remove excess primary antibody and incubated with the appropriate Alexa-fluor (Thermofisher, USA) secondary antibodies for 30 minutes at room temperature. Finally, they were mounted with Fluoroshield with DAPI mounting medium (Sigma-Aldrich, USA) for photographing and subsequent analysis. Ciliated and non-ciliated proliferating cells were quantified using the free software Image J 1.52p (NIH, USA). See ESM Table 2 for antibodies information.

Ca 2+ signaling experiments
Freshly isolated islets were left to recover in a Krebs-Ringer HEPES-buffered solution (pH = 7.35; 5 mmol/l glucose) for 2 hours in the incubator. After recovery, islets were loaded with 2.5 μM Fluo-4-AM (Invitrogen, USA) for at least 2 hours at room temperature and then, transferred to a perfusion chamber coupled to a thermostatic bath that was mounted on a Zeiss LSM 510 laser confocal microscope. The Ca 2+ probe was excited at 488 nm, while emission was collected with a band-pass filter at 505-530 nm. Individual cells within intact islets were monitored in a thin optical section across the islet (8-10 μm) [53]. Islets were perfused at 35ºC with the above-mentioned solution at either 1 or 16 mmol/l glucose. Beta-cells were identified by their typical Ca 2+ response to glucose characterized by either a sustained Ca 2+ elevation or a transient followed by oscillations as well as their characteristic synchronic behavior due to cell coupling [47,48] oscillations were analyzed in the last 6 minutes of the recording at high glucose concentration to allow complete equilibration of perfused media in the islet chamber. We used a contingency Fisher's exact test to analyze differences in the proportion of blocked and active cells between control and A-IDE-KO groups with a level of significance p<0.05.

IQR method for outlier recognition
We have applied this method for identifying outliers in the qRT-PCRs, setting up a "fence" outside of Q1 (first quartile, 25th percentile) and Q3 (third quartile, 75th percentile). Any values that fall outside of this fence were considered outliers. To build this fence we took 1.5 times the IQR and subtracted this value from Q1; then, we took 1.5 times the IQR and we added this value to Q3 (Q1-1.5*IQR; Q3+1.5*IQR; being IQR=Q3 -Q1). This resulted in the minimum and maximum fence posts that were used to compare each value to. Any values that were less than Q1-1.5*IQR or more than Q3+1.5*IQR were considered outliers.