Original Article BACE2 suppression in mice aggravates the adverse metabolic consequences of an obesogenic diet Daniela Díaz-Catalán 1, 2, Gema Alcarraz-Vizán 1, 2, Carlos Castaño 1, 2, Sara de Pablo 1, Júlia Rodríguez-Comas 1, Antonio Fernández-Pérez 2, 3, Mario Vallejo 2, 3, Sara Ramírez 4, Marc Claret 2, 4, 5, Marcelina Parrizas 1, 2, Anna Novials 1, 2, *, 6, Joan-Marc Servitja 1, 2, **, 6 ABSTRACT Objective: Pancreatic b-cell dysfunction is a central feature in the pathogenesis of type 2 diabetes (T2D). Accumulating evidence indicates that b-site APP-cleaving enzyme 2 (BACE2) inhibition exerts a beneficial effect on b-cells in different models of T2D. Thus, targeting BACE2 may represent a potential therapeutic strategy for the treatment of this disease. Here, we aimed to investigate the effects of BACE2 suppression on glucose homeostasis in a model of diet-induced obesity. Methods: BACE2 knock-out (BKO) and wild-type (WT) mice were fed with a high-fat diet (HFD) for 2 or 16 weeks. Body weight, food intake, respiratory exchange ratio, locomotor activity, and energy expenditure were determined. Glucose homeostasis was evaluated by glucose and insulin tolerance tests. b-cell proliferation was assessed by Ki67-positive nuclei, and b-cell function was determined by measuring glucose- stimulated insulin secretion. Leptin sensitivity was evaluated by quantifying food intake and body weight after an intraperitoneal leptin injec- tion. Neuropeptide gene expression and insulin signaling in the mediobasal hypothalamus were determined by qPCR and Akt phosphorylation, respectively. Results: After 16 weeks of HFD feeding, BKO mice exhibited an exacerbated body weight gain and hyperphagia, in comparison to WT littermates. Glucose tolerance was similar in both groups, whereas HFD-induced hyperinsulinemia, insulin resistance, and b-cell expansion were more pronounced in BKO mice. In turn, leptin-induced food intake inhibition and hypothalamic insulin signaling were impaired in BKO mice, regardless of the diet, in accordance with deregulation of the expression of hypothalamic neuropeptide genes. Importantly, BKO mice already showed increased b-cell proliferation and glucose-stimulated insulin secretion with respect to WT littermates after two weeks of HFD feeding, before the onset of obesity. Conclusions: Collectively, these results reveal that BACE2 suppression in an obesogenic setting leads to exacerbated body weight gain, hyperinsulinemia, and insulin resistance. Thus, we conclude that inhibition of BACE2 may aggravate the adverse metabolic effects associated with obesity. Ó 2021 The Author(s). Published by Elsevier GmbH. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Keywords BACE2; High-fat diet; b-cell proliferation; Insulin secretion; Insulin resistance; Leptin 1. INTRODUCTION T2D is attributed to insulin resistance [2]. Obesity induces insulin resistance affecting peripheral tissues, which results in a compensa- Type 2 diabetes (T2D) is a long-term condition closely linked to the tory increase of pancreatic insulin secretion. Failure of b-cells to upsurge of obesity [1]. The pathophysiology connecting obesity and compensate insulin resistance will ultimately lead to T2D. Therefore, 1 Pathogenesis and Prevention of Diabetes Group, Institut d’Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), Barcelona, Spain 2Centro de Investigación Biomédica en Red de Diabetes y Enfermedades Metabólicas Asociadas (CIBERDEM), Spain 3Instituto de Investigaciones Biomédicas Alberto Sols, Consejo Superior de Investigaciones Científicas (CSIC)/Universidad Autónoma de Madrid, Madrid, Spain 4Neuronal Control of Metabolism (NeuCoMe) Laboratory, Institut d’Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), Barcelona, Spain 5School of Medicine, Universitat de Barcelona, Barcelona, Spain 6 Anna Novials and Joan-Marc Servitja jointly supervised this work. *Corresponding author. Pathogenesis and Prevention of Diabetes Group, Institut d’Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), Rossello 149-153, P5, 08036 Barcelona, Spain. E-mail: anovials@clinic.cat (A. Novials). **Corresponding author. Pathogenesis and Prevention of Diabetes Group, Institut d’Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), Rossello 149-153, P5, 08036 Barcelona, Spain. E-mail: servitja@clinic.cat (J.-M. Servitja). Abbreviations: AD, Alzheimer’s disease; CD, Chow diet; BACE1, b-site APP-cleaving enzyme 1; BACE2, b-site APP-cleaving enzyme 2; BKO, BACE2 knock-out; GSIS, Glucose stimulated insulin secretion; HFD, High-fat diet; hIAPP, human islet amyloid polypeptide; TMEM27, Transmembrane protein 27; T2D, Type 2 diabetes; WT, Wild type Received February 9, 2021  Revision received April 16, 2021  Accepted May 9, 2021  Available online 17 May 2021 https://doi.org/10.1016/j.molmet.2021.101251 MOLECULAR METABOLISM 53 (2021) 101251 Ó 2021 The Author(s). Published by Elsevier GmbH. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 1 www.molecularmetabolism.com Original Article strategies targeting b-cell function and expansion can be essential for mice fed with a HFD presented an exacerbated body weight gain, reestablishing glucose homeostasis in T2D associated with obesity. hyperphagia, hyperinsulinemia, and insulin resistance compared to The pathophysiology of T2D is complex and, in addition to b-cell their wild-type (WT) littermates, indicating that BACE2 suppression function, the control of glucose and energy homeostasis involves the aggravates the adverse metabolic effects induced by obesogenic diets. action of several hormones in peripheral tissues such as the liver, skeletal muscle, and adipose tissue, as well as in regulatory regions of 2. MATERIALS AND METHODS the brain. In this regard, leptin is an adipocyte-derived hormone that communicates fuel availability to the central nervous system by sup- 2.1. Animals pressing appetite and increasing energy expenditure [3]. Together with Bace2DE6 mice were purchased from The Jackson Laboratory and insulin, both hormones exert their effects on body weight and fuel continued in the B6; 129P2 background [25]. Bace2DE6 mice were metabolism through their actions in specific neuronal populations crossed to obtain WT and Bace2DE6/DE6 (BKO) male mice. Mouse within the arcuate nucleus of the hypothalamus [4]. Early after the genotyping was performed by PCR amplification of ear DNA samples. administration of a high-fat diet (HFD), the sustained increase of caloric Mice at 8 weeks of age were individually caged and fed ad libitum intake induces hyperphagia and glucose intolerance [5], which are either with pre-weight chow diet (CD, A04 type, Safe diets, Augy, followed by a compensatory increase in leptin and insulin secretion France) or HFD (45% kcal derived from fat; Research Diets, New prior to the onset of obesity and diabetes [6,7]. Brunswick, NJ, USA) for 2 or 16 weeks. Food intake and animal weight BACE2 (b-site APP-cleaving enzyme 2) is a type I transmembrane were controlled weekly. Epididymal and subcutaneous inguinal white protease [8] that has been involved in the control of glucose meta- adipose depots, gastrocnemius muscle and liver were dissected and bolism [9,10]. Its close homolog BACE1 (b-site APP-cleaving enzyme weighed. Protocols were approved by the Animal Ethics Committee of 1) is highly expressed in the brain and responsible for initiating the the Universitat de Barcelona, and the Principles of Laboratory Animal amyloid cascade in Alzheimer’s disease (AD) [11,12]. BACE2 tran- Care were followed. scripts have been reported to be expressed at very low levels in different brain regions [10,13,14]. Nevertheless, BACE2 is expressed 2.2. Indirect calorimetry in discrete subsets of neurons and glia throughout the adult mouse Indirect calorimetry was carried out using a 16-chamber TSE Pheno- brain and rat brain regions [13,15], although its contribution to the master monitoring system (TSE Systems GmbH, Bad Homburg, Ger- pathogenesis of AD remains controversial [16]. BACE2 is expressed at many). Individually caged mice were placed into the measuring room low levels in most peripheral tissues. Importantly, the highest relative one week before the onset of the experiment for acclimation. Mice expression of BACE2 has been found in pancreatic islets [10,14], were fed with CD or HFD as indicated. In the HFD group, the time of where it is expressed solely in the insulin-producing pancreatic b-cells indirect calorimetry experiments coincided with the sixteenth week of in both humans and mice [10,17]. We previously demonstrated that the diet period. Determination of different parameters was carried out inhibition of the b-secretase enzymatic function of BACE2 in pancreatic over a period of 48e72 h. Oxygen consumption and CO2 production b-cells affects the intracellular trafficking of the insulin receptor as well were directly measured. From these data, respiratory exchange ratio as insulin gene expression and content, suggesting that this protease (RER) and energy expenditure (EE) were calculated as follows: plays an important role maintaining b-cell function [18]. In the same RER ¼ VCO2/VO2; EE ¼ (3.185 þ 1.232  RER) x VO2. line, the genetic inhibition of BACE2 improves the deficient secretory response of insulin in rat pancreatic b-cell line INS1E transfected with 2.3. Locomotor activity human islet amyloid polypeptide (hIAPP) [14]. Moreover, our group Locomotor activity was simultaneously monitored together with calo- demonstrated that in a mouse model of b-cell dysfunction induced by rimetry parameters. Ambulatory movement in their cages was the overexpression of hIAPP in b-cells, BACE2 suppression reverts continuously registered every hour over a period of 48e72 h using an glucose intolerance and improves b-cell survival [9]. Of interest, infrared photocell beam grid and is represented as the averaged of the BACE2 was identified as the sheddase of the pro-proliferative plasma total number of beam breaks in the x-and y-axis during the light and membrane protein TMEM27 in murine and human b-cells [10]. dark periods. Accordingly, insulin-resistant mice treated with a BACE2 inhibitor displayed augmented b-cell mass and improved control of glucose 2.4. Histochemistry homeostasis due to increased insulin levels [10]. Collectively, these The left liver lobe was fixed in 4% paraformaldehyde overnight at 4  C, studies suggest that BACE2 inhibition may have a therapeutic potential then was transferred to 30% sucrose in phosphate-buffered saline for T2D treatment. (PBS) for 24 h at 4  C and embedded in tissue freezing medium (Leica, Pharmacological inhibition of BACE1 is being intensively pursued as a Wetzlar, Germany). We obtained 8-mm-thick sections with a CM1860 therapeutic approach to treat patients with AD [19]. Although signifi- cryostat (Leica) and applied them to poly-lysine coated slides. Liver cant progress has been made in the last years, the majority of current sections were stained with Oil red O as described elsewhere [26]. BACE1 inhibitors lack selectivity and can also target BACE2 [20,21]. Images were taken with Nikon Eclipse E600 fluorescence microscope Interestingly, it has been reported that BACE1 suppression decreases and collected with Olympus Cell^D software v3.4. body weight, protects against diet-induced obesity and enhances in- sulin sensitivity in mice [22,23]. The beneficial effects that BACE2 2.5. Glucose and insulin tolerance tests inhibition exerts on b-cell mass and function in different models of T2D For intraperitoneal glucose tolerance tests (GTTs), mice were fasted for suggests that selective BACE2 inhibitors may be potentially a new path 12 h and were intraperitoneally injected with D-glucose (2 g/kg body for the treatment of this disease [24]. However, the effects of BACE2 weight). For intraperitoneal Insulin Tolerance Tests (ITTs), mice fasted for inhibition on glucose metabolism under an obesogenic regime has not 5 h were intraperitoneally injected with 0.4 IU/kg insulin (Humulin-R, Eli yet been explored. This prompted us to explore the impact of BACE2 Lilly, Indianapolis, IN, USA). Blood glucose levels were measured via tail suppression on the regulation of glucose and energy homeostasis in a vein using an automatic glucometer (Nova Pro) at different time points model of diet-induced obesity. We found that BACE2 knock-out (BKO) after glucose or insulin injection. 2 MOLECULAR METABOLISM 53 (2021) 101251 Ó 2021 The Author(s). Published by Elsevier GmbH. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). www.molecularmetabolism.com 2.6. Insulin and leptin determinations embedded. Three-micrometer-thick pancreatic sections from 3 Blood samples were collected from the tail vein of mice fasted for 12 h. different levels (>150 mm apart) for each pancreas were deparaffi- Blood samples were also collected 15 min after glucose injection in nized, rehydrated, boiled in citrated buffer (10 mM; pH ¼ 6.0) only in GTTs. Plasma insulin and leptin levels were measured with ELISA kits case of Ki67 staining, permeabilized with 1% Triton X-100 (Sigmae according to manufacturer’s guidelines (Crystal Chem, Downers Grove, Aldrich) in Dulbecco’s phosphate buffer saline (PBS; SigmaeAldrich) IL, USA). and blocked with 5% Donkey serum in PBS. Samples were incubated overnight with polyclonal guinea pig anti-insulin (1:500; Dako, 2.7. Glucose-stimulated insulin secretion (GSIS) Glostrup, Denmark) and mouse-anti glucagon (1:1000; Sigmae Pancreatic islets were isolated from WT and BKO males following the Aldrich) or rabbit anti-Ki67 (1:200; Thermo Fisher Scientific, Waltham, dietary intervention by collagenase digestion and handpicked after a MA, USA). As secondary antibodies, Alexa Fluor 488 anti-guinea pig density gradient using Histopaque (SigmaeAldrich, St Louis, MO, USA), (1:250; Jackson I.R., Newmarket, UK), 555 anti-mouse (1:250; Invi- as previously described [27]. The same day of isolation, islets were trogen, Carlsbad, CA, USA) and 455 anti-rabbit (1:400, Dako) were preincubated with KrebseRinger bicarbonate HEPES buffer solution used. Hoechst 33342 (Sigma Aldrich) was used to stain nuclei. Images (115 mmol/L NaCl, 24 mmol/L NaHCO3, 5 mmol/L KCl, 1 mmol/ were taken with Leica TCS SPE confocal microscope and the analysis LMgCl2$6H2O, 1 mmol/L CaCl2$2H2O, 20 mmol/L HEPES, and 0.5% performed using Fiji (ImageJ) version 2.1 [28]. Total b-cell mass was BSA, pH 7.4) containing 2.8 mmol/L glucose for 30 min at 37  C. Eight calculated by multiplying their fractional area per pancreas weight. The islets per assay were then incubated with either 2.8 mmol/L glucose or percentage of proliferating b-cells was calculated as the percentage 16.7 mmol/L glucose. After 1 h, supernatants were collected and (%) of Ki67-positive b-cells with respect to total b-cells. cellular insulin contents were recovered in acid-ethanol solution. Insulin concentration was determined by Insulin Mouse ELISA (Crystal Chem). 2.12. Gene expression analysis Total RNA from adipose tissue and mediobasal hypothalamus (MBH) 2.8. Leptin sensitivity test was isolated using Trizol (SigmaeAldrich) reagent and reverse- Mice were individually housed and acclimatized by handling and sham transcribed using SuperScript (Invitrogen). qPCR was conducted us- injections for a week. Ten-week-old WT and BKO mice on HFD for two ing SYBR Green (Invitrogen) in ABI Prism 7900 HT Sequence Detection weeks were intraperitoneally injected with either vehicle or 5.5 mg/kg of System (Applied Biosystems, Foster City, CA, USA). The primers used recombinant mouse leptin (R&D Systems, Minneapolis, MN, USA) 1 h were as follows: Hprt sense 50 -GGTTAAGCAGTACAGCCCCA-30 , Hprt1 before lights-out. Food intake and body weights were recorded every 12 h. antisense 50 -TCCAACACTTCGAGAGGTCC-30 ; Leptin sense 50 - CATCTGCTGGCCTTCTCCAA-30 , Leptin antisense 50 -ATC- 2.9. Central insulin signaling test CAGGCTCTCTGGCTTCTG-30 ; Npy sense 50 -TACTCCGCTCTGCGA- Mice were fasted for 12 h and anesthetized by an intraperitoneal in- CACTA-30 , Npy antisense 50 -TACTCCGCTCTGCGACACTA-30 ; Agrp jection of ketamine and xylazine (100 mg/kg and 10 mg/kg, respec- sense 50 -AGCTTTGTCCTCTGAAGCTGT-30 , Agrp antisense 50 - tively). We injected 100 mL of a 0.9% saline solution with or without 5 U AGCTTTGTCCTCTGAAGCTGT-30 ; Pomc sense 50 -CCTTTCCGCGACA- of regular human insulin (Humulin R, Eli Lilly) into the inferior cava vein. GAGACTA-30 , Pomc antisense 50 -AGCGAGAGGTCGAGTTTGC-30 . After 10 min, the mediobasal hypothalamus (MBH) was isolated using Expression levels were normalized to the expression of an endogenous a coronal brain matrix. Insulin signaling was analyzed by determining house-keeping gene (Hprt1). phosphorylated and total Akt levels by western blot. 2.13. Statistical analysis 2.10. Western blot analysis All results are expressed as mean  standard error of mean (SEM). MBH was homogenized in RIPA buffer (1% Triton X-100, 1% sodium The D’Agostino & Pearson omnibus normality test was carried out to deoxycolate, 0.1% SDS, 0.15 M NaCl, 0.05 M TriseHCl, pH 7.2) test whether the variables followed a normal distribution supplemented with complete mini protease inhibitor cocktail tablets (alpha ¼ 0.05). Statistical significance was determined by a paired (Roche, Mannheim, Germany). An equal amount of protein (15 mg) was two-tailed Student’s t-test or two-way ANOVA with Tukey’s or Fisher’s heat-denaturized and separated on 12% SDS acrylamide gels and least significant difference (LSD) post hoc analysis as indicated. A transferred onto PVDF Immobilon-P membrane (Millipore, Burlington, value of P < 0.05 was considered statistically significant. All figures MA, USA). Subsequently, membranes were blocked with 5% non-fat and statistical analyses were generated using GraphPad Prism 8 dry milk in Tris-buffered saline including 0.1% Tween-20 (TBS-T; (Graphpad Software, La Jolla, CA, USA). SigmaeAldrich) and incubated overnight at 4  C with the appropriate primary antibody in 5% BSA in TBS-T. Membranes were then incu- 3. RESULTS bated with primary antibodies against Akt (#9272) and p-Akt (#9271) (Cell Signaling, Danvers, MA, USA), followed by incubation with 3.1. BKO mice develop exacerbated obesity and hyperphagia with horseradish peroxidase (HRP)-conjugated anti-rabbit IgG (Cyvita, altered energy metabolism after long-term HFD feeding Marlborough, MA, USA). Immunoreactive bands were visualized using To study the implication of BACE2 in the regulation of energy ho- ECL (Millipore) reagent and image acquisition was performed by using meostasis under obesogenic conditions, we monitored the body weight a chemiluminescent CCD imager ImageQuant LAS 4000 (GE Health- and food intake of WT and BKO mice fed either with chow diet (CD) or care, Velizy-Villacoublay, France). Phosphorylated and total Akt bands HFD for 16 weeks. Both HFD groups gained more weight than the were quantified using ImageJ software version 1.0 (National institutes groups on CD, but remarkably, BKO-HFD mice gained more weight of Health, Bethesda, MD, USA). than WT-HFD mice (Figure 1A). Thus, HFD administration induced a body weight gain of 92.5  4.3% in BKO mice as compared to only 2.11. Immunofluorescence and quantitative image analysis 66.2  6.9% in WT mice (Figure 1A). Both HFD groups displayed a Pancreata from WT and BKO mice fed two weeks on CD and HFD were similar increase in adiposity, measured as the percentage of the dissected, fixed in 10% formalin neutral buffered solution and paraffin weight of epididymal white adipose tissue and subcutaneous adipose MOLECULAR METABOLISM 53 (2021) 101251 Ó 2021 The Author(s). Published by Elsevier GmbH. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 3 www.molecularmetabolism.com Original Article Figure 1: Body weight, energy intake, respiratory exchange ratio, and energy expenditure of BKO mice after long-term HFD feeding. (A) Evolution of body weight of WT and BKO mice on chow diet (WT CD, n ¼ 15; BKO CD, n ¼ 17) or high-fat diet (WT HFD, n ¼ 13; BKO HFD, n ¼ 16) and (B) energy intake (n ¼ 5e7 mice/group) during 16 weeks of dietary treatment. (C) Respiratory exchange ratio (RER) corresponding to a complete 24 h cycle (the gray area represents the dark phase) (n ¼ 3e4 mice/group). (D) Average of RER values represented in C. (E) Locomotor activity and (F) energy expenditure measured during the light and dark phases. Data represent the mean  SEM. Two-way ANOVA multiple comparisons with Tukey’s post hoc test (AeF). *P < 0.05, **P < 0.01, ***P < 0.001; #P < 0.05, ##P < 0.01, ###P < 0.001 (WT-HFD vs. BKO-HFD). tissue with respect to body weight. Liver weight was significantly the higher weight gain observed in BKO-HFD mice. Plasma insulin higher in BKO-HFD mice than in the other groups, which was levels were determined at basal conditions and 15 min after glucose accompanied by enhanced accumulation of hepatic lipid droplets as challenge. Remarkably, basal insulin levels were much higher in BKO- demonstrated by Oil Red O staining of liver sections (Supplementary HFD in comparison to that of the other groups (Figure 2C). Glucose Figure S1). In parallel, we also measured the energy intake (EI) on a administration induced insulin secretion in all groups, although it was weekly basis. Importantly, BKO-HFD mice exhibited increased hyper- more pronounced in WT-HFD mice and especially in BKO-HFD mice. phagia compared to WT-HFD mice, which presented similar values to These results revealed an exacerbated hyperinsulinemia in BKO-HFD mice under CD feeding (Figure 1B). This increase in food intake was mice at fasting conditions and in response to glucose. observed shortly after initiating the dietary treatment and prior to a We subsequently evaluated insulin sensitivity by an ITT. Insulin significant weight gain, suggesting an early deregulation of food intake sensitivity was decreased in both HFD-fed mice groups compared with in BKO mice when challenged with a HFD. their respective CD groups (Figure 2D). Despite similar levels of Subsequently, we performed indirect calorimetry to gain insights into glucose intolerance between the two HFD groups, BKO-HFD mice were the metabolic status of mice. Under CD conditions, both groups pre- more insulin-resistant than the WT-HFD mice, in line with the sented similar oscillations in RER (Figure 1C,D). HFD groups showed a increased insulin levels. Of note, BKO-CD mice showed lower levels of lower RER than CD mice with a reduced oscillatory pattern, consistent fasting glucose than their WT-CD littermates (Figure 2E), with the same with a preference for lipid oxidation as a source of fuel. Remarkably, insulin sensitivity, which could be attributable to the slightly increased RER was significantly lower in BKO-HFD mice compared to WT-HFD basal insulin levels in BKO-CD mice (0.25  0.04 ng/mL) as compared mice during the active dark cycle. These results show a metabolic to WT-CD mice (0.16  0.02 ng/mL). profile that is consistent with the enhanced body weight observed in BKO-HFD mice. Importantly, no significant differences were observed 3.3. Pancreatic islets from BKO mice under HFD feeding display an in locomotor activity (Figure 1E) and energy expenditure normalized to exacerbated insulin secretion and b-cell proliferation body weight (Figure 1F) during the dark and light periods. Taken To assess the impact of BACE2 suppression on the secretory profile of together, these results indicate that the increased body weight the pancreatic b-cells, we isolated pancreatic islets and performed observed in BKO-HFD mice is mainly due to enhanced energy intake glucose-stimulated insulin secretion (GSIS) tests. Islets from all groups rather than a decrease in activity or energy expenditure. had a similar secretory response to high glucose except those from BKO-HFD mice, which displayed a higher basal insulin secretion and a 3.2. BKO mice exhibit enhanced insulin secretion under long-term relatively reduced response to high glucose, even though the total HFD feeding amount of insulin released was higher than that of any other group To determine the effects of BACE2 suppression on whole-body glucose (Figure 3A). homeostasis, mice were challenged with an intraperitoneal glucose Morphometric analysis of pancreas sections revealed that the b-cell tolerance test (GTT). As expected, long-term HFD administration mass and the frequency of large islets were increased in BKO-HFD resulted in a decrease in glucose tolerance when compared with CD mice as compared to BKO-CD and WT-HFD mice (Figure 3B,C). In groups (Figure 2A,B). Interestingly, BKO-HFD mice displayed a similar line with these results, the number of proliferating b-cells, identified by impairment in glucose tolerance compared to WT-HFD mice, despite double staining of insulin and Ki67, was also increased in the BKO-HFD 4 MOLECULAR METABOLISM 53 (2021) 101251 Ó 2021 The Author(s). Published by Elsevier GmbH. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). www.molecularmetabolism.com Figure 2: Glucose homeostasis and insulin secretion in BKO mice after long-term HFD feeding. (A) Intraperitoneal glucose tolerance test (GTT) (n ¼ 11e17 mice/group), (B) Area under the curve (AUC) from the GTT, (C) plasma insulin levels before and 15 min after glucose administration (n ¼ 11e18 mice/group), (D) intraperitoneal insulin tolerance test (ITT) (n ¼ 12e15 mice/group) and (E) fasting blood glucose levels (n ¼ 15e19 mice/group) in WT and BKO mice fed either CD or HFD for 16 weeks. Data represent the mean  SEM. Indicated comparisons were made using two-way ANOVA followed by Tukey’s (A, B, C and E) or Fisher’s LSD (D) post hoc tests. &P < 0.05, &&P < 0.01, &&& P < 0.001 (WT-CD vs. WT-HFD); #P < 0.05, ##P < 0.01, ###P < 0.001 (BKO-CD vs. BKO-HFD); $P < 0.05, $$P < 0.01 (WT-HFD vs. BKO-HFD); *P < 0.05, **P < 0.01, ***P < 0.001. yP < 0.05, yyyP < 0.001, using Student’s paired t-test (C). mice (Figure 3D,E). Thus, the increased plasma insulin levels in BKO bolus of insulin or saline was administered to mice in order to analyze mice following 16 weeks of HFD feeding were associated with an the activation of Akt, a proxy for insulin signaling, in the mediobasal enhanced expansion of the b-cell mass. hypothalamus (MBH). As expected, insulin administration resulted in Akt activation in WT-CD mice (Figure 4E). Conversely, the basal 3.4. BKO mice show impaired leptin response and hypothalamic phosphorylation of Akt was already increased in the WT-HFD mice and insulin signaling with altered neuropeptide expression in both BKO groups and hence insulin was unable to further activate Having observed that BKO-HFD mice display increased body weight, Akt. Collectively, these results revealed an impaired anorexigenic hyperphagia, and hyperinsulinemia in comparison to WT-HFD after 16 response to leptin and insulin signaling in the MBH of BKO mice, even weeks of HFD feeding, we next sought to study these mice following a when fed on CD, which may trigger the increase in food intake and short HFD regime. In fact, BKO-HFD presented enhanced energy intake body weight observed in BKO mice fed on HFD. shortly after the start of the dietary treatment (Figure 1B), suggesting Next, we assessed the expression of appetite-regulating neuropeptide that the mechanisms underlying this phenotype would be early events. genes in the MBH. For this, we determined the expression of the genes First, we evaluated circulating levels of leptin after two weeks on HFD. encoding the orexigenic peptides, neuropeptide Y (Npy) and agouti- In fasting conditions, the levels of leptin were similarly increased in related protein (Agrp), and the anorexigenic peptide proopiomelano- both HFD-fed mice groups (Figure 4A). Consistently, the mRNA cortin (Pomc) after an overnight fasting. The Npy and Agrp expression expression of leptin in the epididymal white adipose tissue was also was significantly increased in the BKO mice, whereas Pomc levels increased in both groups (Figure 4B). These results ruled out the were increased in WT-HFD mice in comparison to WT-CD mice; possibility that differences in energy intake may be explained by dif- however, this transition between diets was not observed in BKO mice ferences in circulating leptin levels. (Supplementary Figure S2). In order to estimate the anorexigenic/ Given that energy intake is regulated in part by leptin and insulin at the orexigenic tone, the ratios between Pomc mRNA, and Npy or Agrp hypothalamus, we next sought to assess how the action of these mRNAs were calculated, revealing that the anorexigenic/orexigenic hormones was affected in the absence of BACE2. First, a leptin balance was potently enhanced in WT mice after two weeks on HFD sensitivity test was performed following two weeks of HFD feeding, (Figure 4F,G). Conversely, the ratios between anorexigenic and corresponding to 10 weeks of mice age. As expected, leptin admin- orexigenic peptides did not show this transition when BACE2-deficient istration led to an anorexigenic effect in WT-CD mice, as shown by a mice were fed on HFD, suggesting imbalanced orexigenic and reduction in food intake (Figure 4C) and body weight (Figure 4D), and anorexigenic outputs in the hypothalamus, in agreement with the this response was attenuated in WT-HFD mice. Remarkably, BKO mice enhanced food intake observed in BKO-HFD mice. were totally unresponsive to leptin, regardless of the dietary treatment (Figure 4C,D). Because leptin synergistically acts with insulin to induce 3.5. BKO mice show early hyperinsulinemia and insulin resistance its anorexigenic actions, and considering that hypothalamic insulin following a short-term HFD regime resistance has been show to underlie deregulation in food intake [4], Due to the impairment in leptin and insulin responses in the hypo- we wondered whether insulin signaling in the hypothalamus may be thalamus of BKO mice and the deregulation of appetite regulatory affected after a short-term HFD feeding. To test this, an intravenous genes shortly after the administration of HFD, we next assessed MOLECULAR METABOLISM 53 (2021) 101251 Ó 2021 The Author(s). Published by Elsevier GmbH. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 5 www.molecularmetabolism.com Original Article Figure 3: Enhanced insulin secretion and b-cell mass and proliferation in BKO mice after long-term HFD feeding. (A) Ex vivo glucose-stimulated insulin secretion at 2.8 mM and 16.7 mM of glucose, represented as % of insulin secretion respect to insulin content of islets from WT and BKO mice fed either CD or HFD for 16 weeks (n ¼ 3 mice/ group). (B) Quantification of b-cell mass and (C) profile of islet size distribution in WT and BKO mice fed either CD or HFD for 16 weeks (n ¼ 5e7 replicates/group). (D) Representative images of islets stained for insulin (green), Ki67 (red) and nuclei with Hoechst (blue). The yellow arrows point to Ki67 positive nuclei in insulin-positive cells. Scale bar, 100 mm. (E) Quantification of b-cell proliferation in pancreas from WT and BKO mice fed either CD or HFD for 16 weeks (n ¼ 3 mice/group). Three sections per animal were analyzed. Data represent the mean  SEM. *P < 0.05, **P < 0.01, ***P < 0.001, using two-way ANOVA followed by Tukey’s post hoc test (A, B, C and E). ##P < 0.01, ### P < 0.001, using Student’s paired t-test (A). glucose homeostasis at the same time point. After two weeks of di- a trend toward a higher b-cell mass in BKO-HFD mice (Figure 6B,C). etary treatment, glucose tolerance was already decreased in both HFD- Indeed, the frequency of islets with a higher percentage of Ki67- fed mice groups compared to CD groups (Figure 5A,B). As observed positive b-cells was increased in BKO-HFD mice (Figure 6D,E). All after 16 weeks of HFD feeding, both WT-HFD and BKO-HFD mice together, these results revealed that suppression of BACE2 function exhibited the same level of glucose intolerance. Fasting insulinemia in leads to an enhanced b-cell secretory capacity and proliferation early BKO-HFD mice (0.34  0.05 ng/mL) was slightly increased relative to after HFD administration. WT-HFD mice (0.24  0.04 ng/mL) and significantly higher compared with BKO-CD (0.14  0.02 ng/mL) mice (Figure 5C). After glucose 4. DISCUSSION administration, plasma insulin levels in the BKO-HFD mice were significantly increased compared with WT-HFD mice and the BKO-CD We previously demonstrated that the absence of BACE2 improves mice, revealing an early higher secretory response in BKO-HFD mice. glucose tolerance and insulin secretion in BACE2 KO mice crossed with The ITT revealed that the HFD groups showed less sensitivity to insulin, Tg-hIAPP mice, which display insulin secretory defects due to the implying that BKO-HFD mice are more insulin resistant than their WT aggregation of hIAPP [9]. Here, we aimed to explore whether the HFD littermates (Figure 5D). HFD groups presented higher fasting absence of BACE2 could affect glucose and energy homeostasis in an glycemia than their respective CD groups (Figure 5E). These data obesogenic context. To achieve this objective, WT and BKO mice were indicate that early at two weeks of HFD, the BKO mice display an challenged with HFD, in order to induce metabolic changes similar to exacerbated insulin secretion both at fasting and after glucose load and early stages of T2D in humans. This strategy provides a useful tool to insulin resistance. study potential therapies for T2D associated with obesity [29,30]. Unexpectedly, we found that, after 16 weeks of HFD feeding, BKO mice 3.6. Pancreatic islets from BKO mice present a more pronounced gained more weight than the WT mice. The exacerbated increase in b-cell proliferation and insulin secretion following a short HFD body weight was due to a higher energy intake rather than a decrease feeding in activity or energy expenditure. Furthermore, the BKO-HFD mice Once characterized the metabolic state of mice after two weeks on exhibited a lower RER than the WT-HFD mice, reflecting an increased HFD, we performed GSIS assays in isolated pancreatic islets to further preference for lipid oxidation as a source of fuel energy. determine the secretory capacity of b-cells. Islets from all groups In parallel, we analyzed the influence of the obesogenic diet on glucose similarly responded to high glucose, except the WT-HFD mice, whose homeostasis in BKO mice. Importantly, glucose intolerance was similar insulin secretory response was attenuated (Figure 6A). Remarkably, in both HFD groups despite the more accentuated obese phenotype of islets from BKO-HFD mice showed an increased insulin secretory BKO mice in relation to WT littermates. An explanation for this pro- response to high glucose when compared with WT-HFD and BKO-CD tection against an expected aggravation of glucose tolerance is the fact islets (Figure 6A). Morphometric analyses of pancreas sections that BKO-HFD mice exhibited increased circulating insulin levels at revealed no significant differences in b-cell mass and islet size fre- fasting conditions and after a glucose challenge. In addition, the insulin quency between the experimental groups, although there was already secretory response to glucose in isolated pancreatic islets from BKO- 6 MOLECULAR METABOLISM 53 (2021) 101251 Ó 2021 The Author(s). Published by Elsevier GmbH. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). www.molecularmetabolism.com Figure 4: BKO mice exhibit impaired leptin sensitivity and central insulin signaling. (A) Circulating leptin plasma levels (n ¼ 5e7/group) and (B) leptin mRNA expression in epididymal white adipose tissue (n ¼ 5e6/group) in WT and BKO mice fed either CD or HFD for 2 weeks. (C) Cumulative energy intake and (D) change in body weight 24 h after an intraperitoneal injection of leptin in WT and BKO mice fed CD or HFD for 2 weeks (n ¼ 5e6 mice/group). (E) Representative western blots for phospho-Akt (p-Akt S473) and Akt in the mediobasal hypothalamus from WT and BKO mice fed CD or HFD for 2 weeks after an intraperitoneal injection of insulin. The graphs on the right represent the ratios of p-Akt S473 to total Akt quantifications normalized to the basal values of WT mice on CD (n ¼ 6e9/group). (F,G) Ratios between Pomc, and Npy (F) or Agrp (G) mRNA levels in the mediobasal hypothalamus of WT and BKO mice fed CD or HFD for 2 weeks (n ¼ 3e6 mice/group). Data represent the mean  SEM. *P < 0.05, **P < 0.01, ***P < 0.001, using two-way ANOVA followed by Tukey’s post hoc test (AeG). HFD mice was increased compared to those from WT-HFD mice. These Leptin and insulin mainly exert their anorectic effect in the arcuate results are in agreement with our previous studies showing that BACE2 nucleus via inhibition of orexigenic NPY/AgRP neurons and activation of inhibition in b cells results in enhanced glucose-induced insulin anorexigenic POMC neurons resulting in reduced food intake and secretion [14]. The more pronounced hyperinsulinemia detected in increased energy expenditure [4]. In this regard, HFD attenuated the BKO-HFD mice was associated with increased b-cell mass and pro- expression of the orexigenic Npy and Agrp genes and increased the liferation. In this regard, previous studies showed that BACE2 inhibition expression of the anorexigenic gene Pomc, most likely reflecting an in genetically obese ob/ob mice leads to increased insulin secretion attempt to moderate energy intake. Conversely, these transitions fa- and improved glucose tolerance [10]. In the present study, we show voring an anorexigenic tone were not observed when BKO mice were that challenging BKO mice with an obesogenic diet resulted in fed a HFD. Importantly, these effects were already observed after two increased plasma insulin levels but also in a more pronounced weeks of HFD treatment, demonstrating an early impairment of the impairment of insulin resistance already after two weeks of HFD regulation of the expression of hypothalamic neuropeptides controlling feeding, before the onset of obesity. food intake, which is in line with the hyperphagia observed in BKO-HFD We also focused our attention on understanding the increase in energy mice shortly after initiating the diet intervention. Collectively, these intake that leads BKO-HFD mice to gain more weight than control results revealed impaired central insulin and leptin responses in the littermates. It is well known that insulin acts as a satiety signal in the BKO mice under chow diet, suggesting that BACE2 may play a role in brain and reduces food intake, similar to the adipose-derived hormone hypothalamic areas controlling food intake. It should be noted that leptin [4]. Indeed, insulin and leptin synergistically act in the arcuate although BACE2 expression is highest in pancreatic islets, and at much nucleus via PI3K/Akt [31]. Therefore, we wondered whether the lower levels in some peripheral tissues, this secretase is expressed in disturbed food intake and glucose homeostasis in BKO mice in the discrete subsets of neurons and glia throughout the adult brain, long-term HFD regime may be caused by an early impairment in the including the ventromedial hypothalamic nucleus [13,15], and there- central nervous system. For this purpose, we analyzed the anorexi- fore a potential role of BACE2 in the brain requires further investigation. genic response to leptin and the hypothalamic insulin signaling after In this regard, it has been shown that proinflammatory TNF induces an two weeks of HFD, before the onset of obesity. Surprisingly, our results increase in BACE2-mediated shedding of vascular cell adhesion showed that food intake and body weight reduction in response to molecule 1 in the cerebrospinal fluid from adult mice, illustrating that leptin administration were attenuated in BKO mice as compared to WT BACE2 inhibition might become apparent under physiopathological animals, regardless of the diet. Furthermore, we observed that insulin conditions [15]. Furthermore, the effects of BACE2 suppression on administration was not able to activate Akt in the mediobasal hypo- hypothalamic signaling and gene regulation also open the possibility thalamus (MBH) in both HFD groups but also in BKO mice fed with CD. that this may impact other regions regulated by hypothalamic outputs. MOLECULAR METABOLISM 53 (2021) 101251 Ó 2021 The Author(s). Published by Elsevier GmbH. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 7 www.molecularmetabolism.com Original Article Figure 5: Glucose homeostasis and insulin secretion in BKO mice after short-term HFD feeding. (A) Intraperitoneal glucose tolerance test (GTT) (n ¼ 9e11 mice/group), (B) area under the curve (AUC) from the GTT, (C) plasma insulin levels before and 15 min after glucose administration (n ¼ 8e12 mice/group), (D) intraperitoneal insulin tolerance test (ITT) (n ¼ 6e7 mice/group) and (E) fasting blood glucose levels (n ¼ 20 mice/group) in WT and BKO mice fed either CD or HFD for 2 weeks. Data represent the mean  SEM. &&& P < 0.001 (WT-CD vs. WT-HFD); ##P < 0.01, ###P < 0.001 (BKO-CD vs. BKO-HFD); $P < 0.05 (WT-HFD vs. BKO-HFD); *P < 0.05, **P < 0.01, ***P < 0.001, using two-way ANOVA followed by Tukey’s (A, B, C and E) and Fisher’s LSD (D) post hoc tests. yyP < 0.01, yyyP < 0.001, using Student’s paired t-test (C). Figure 6: Enhanced insulin secretion and proliferation in pancreatic islets from BKO mice after short-term HFD feeding. (A) Ex vivo glucose-stimulated insulin secretion at 2.8 mM and 16.7 mM of glucose, represented as % of insulin secretion respect to insulin content of islets from WT and BKO mice fed either CD or HFD for 2 weeks (n ¼ 3 mice/ group). (B) Quantification of b-cell mass and (C) profile of islet size distribution in WT and BKO mice fed either CD or HFD for 2 weeks (n ¼ 4e6 replicates/group). (D) Representative images of islets stained for insulin (green), Ki67 (red) and nuclei with Hoechst (blue). The yellow arrows point to Ki-67 positive nuclei in insulin-positive cells. Scale bar, 100 mm. (E) Quantification of b-cell proliferation in pancreas from WT and BKO mice fed either CD or HFD for 2 weeks (n ¼ 3 mice/group). Three sections per animal were analyzed. Data represent the mean  SEM. *P < 0.05, using two-way ANOVA followed by Tukey’s post hoc test (A, B, C and E). #P < 0.05, ##P < 0.01, using Student’s paired t- test (A). Indeed, the autonomic nervous system modulates the secretory pancreatic islets ex vivo showed a pattern of exacerbated glucose- response of pancreatic islets through sympathetic and para- stimulated insulin secretion similar to that observed in vivo. This, sympathetic fibers [33,34], and different hypothalamic areas have together with the increased b-cell mass, suggests that the enhanced been shown to be connected to efferent autonomic fibers innervating insulin secretion in BKO-HFD mice is mainly attributed to a direct effect pancreatic islets in mice [35]. In the current study, BKO-HFD of BACE2 suppression on pancreatic islets. 8 MOLECULAR METABOLISM 53 (2021) 101251 Ó 2021 The Author(s). Published by Elsevier GmbH. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). www.molecularmetabolism.com Previous studies have demonstrated beneficial effects of BACE1 inhi- CONFLICT OF INTEREST bition on whole-body metabolic homeostasis and neuronal leptin sensitivity under a chronic HFD consumption [22,23], pointing towards None declared. distinct roles of BACE1 and BACE2 on metabolism homeostasis. These differences may be explained by the different substrates and tissue APPENDIX A. SUPPLEMENTARY DATA expression pattern of each of these secretases. One of the substrates of BACE2 is TMEM27, which is also highly expressed in the pancreatic Supplementary data to this article can be found online at https://doi.org/10.1016/j. islets [10]. Nevertheless, other b-cell enriched substrates of BACE2 in molmet.2021.101251. b-cells have been identified [32,36]. The molecular mechanisms by which BACE2 deficiency and inhibition promote b-cell function and REFERENCES proliferation are still unresolved, but likely involve the stabilization of TMEM27 as well as other BACE2 substrates yet to be identified that may synergistically contribute to these effects. [1] Ogurtsova, K., da Rocha Fernandes, J.D., Huang, Y., Linnenkamp, U., Hyperinsulinemia is one of the hallmarks of T2D associated with Guariguata, L., Cho, N.H., et al., 2017. IDF Diabetes Atlas: global estimates obesity. In fact, the physiological response to insulin resistance in the for the prevalence of diabetes for 2015 and 2040. Diabetes Research and early stages of obesity is mainly attributed to an increase in pancreatic Clinical Practice 128:40e50. b-cell mass and insulin secretion. Nevertheless, increased insulin [2] Nolan, C.J., Prentki, M., 2019. Insulin resistance and insulin hypersecretion in levels may result in adverse effects on glucose metabolism. On the one the metabolic syndrome and type 2 diabetes: time for a conceptual framework hand, insulin induces fat storage in adipose tissue [37], and therefore shift. Diabetes and Vascular Disease Research 16(2):118e127. an excess of insulin may exacerbate fat accumulation in obesogenic [3] Friedman, J.M., 2019. Leptin and the endocrine control of energy balance. contexts. Indeed, mild suppression of insulin secretion has been Nature Metabolism 1(8):754e764. proven as a strategy for preventing and treating obesity [38,39]. On the [4] Timper, K., Brüning, J.C., 2017. Hypothalamic circuits regulating appetite and other hand, hyperinsulinemia may cause insulin resistance even prior energy homeostasis: pathways to obesity. Disease Models & Mechanisms to the onset of obesity [2,40,41]. Indeed, our results show that 10(6):679e689. following a short HFD feeding of two weeks, BKO mice already display [5] Turner, N., Kowalski, G.M., Leslie, S.J., Risis, S., Yang, C., Lee-Young, R.S., enhanced hyperinsulinemia, which may underlie the accentuated in- et al., 2013. Distinct patterns of tissue-specific lipid accumulation during the sulin resistance in these mice. Moreover, it is also important to stress induction of insulin resistance in mice by high-fat feeding. Diabetologia 56(7): that hyperinsulinemia in obesity impairs insulin sensitivity in the brain 1638e1648. [42,43], leading to deregulation of food intake and eventually body [6] Olofsson, L.E., Unger, E.K., Cheung, C.C., Xu, A.W., 2013. Modulation of AgRP- weight gain that may also contribute to the aggravation of the obese neuronal function by SOCS3 as an initiating event in diet-induced hypothalamic phenotype in BKO mice challenged with HFD. leptin resistance. Proceedings of the National Academy of Sciences of the In summary, the present study shows that BACE2 inhibition in mice United States of America 110(8):697e706. under a HFD feeding exacerbates body weight gain, hyperphagia, [7] Yang, Y., Smith, D.L., Keating, K.D., Allison, D.B., Nagy, T.R., 2014. Variations hyperinsulinemia, and insulin resistance. Thus, we conclude that in body weight, Food Intake and body composition after long-term high-fat diet BACE2 suppression may aggravate the adverse metabolic effects feeding in C57BL/6J mice. Obesity 22(10):2147e2155. associated with an obesogenic diet. These findings should be [8] Lin, X., Koelsch, G., Wu, S., Downs, D., Dashti, A., Tang, J., 2000. Human considered when considering a potential clinical use of non-selective aspartic protease memapsin 2 cleaves the b-secretase site of b-amyloid BACE1/BACE2 or selective BACE2 inhibitors for the treatment of AD precursor protein. Proceedings of the National Academy of Sciences of the or T2D, respectively. United States of America 97(4):1456e1460. [9] Alcarraz-Vizán, G., Castaño, C., Visa, M., Montane, J., Servitja, J.-M., AUTHORS’ CONTRIBUTIONS Novials, A., 2017. BACE2 suppression promotes b-cell survival and function in a model of type 2 diabetes induced by human islet amyloid D. D.-C., G. A.-V., C. C., S. P., J. R.-C., A. F.-P., and S. R. contributed to polypeptide overexpression. Cellular and Molecular Life Sciences 74(15): the performance of experiments and data analysis. M. V., M.C., and M. 2827e2838. P. contributed to data analysis. D. D-C., A. N., and J.-M. S. wrote the [10] Esterházy, D., Stützer, I., Wang, H., Rechsteiner, M.P., Beauchamp, J., manuscript. A. N. and J.-M. S. designed the study. All authors Döbeli, H., et al., 2011. Bace2 is a b cell-enriched protease that regulates contributed to the discussion of results, reviewed the manuscript, and pancreatic b cell function and mass. Cell Metabolism 14(3):365e377. approved the final version of this manuscript. [11] Zhao, J., Fu, Y., Yasvoina, M., Shao, P., Hitt, B., O’Connor, T., et al., 2007. b- site amyloid precursor protein cleaving enzyme 1 levels become elevated in ACKNOWLEDGMENTS neurons around amyloid plaques: implications for Alzheimer’s disease path- ogenesis. Journal of Neuroscience 27(14):3639e3649. This work was supported by projects from the Instituto de Salud Carlos III (PI17/00879 to [12] Vassar, R., Bennett, B.D., Babu-Khan, S., Kahn, S., Mendiaz, E.A., Denis, P., AN and JMS) and by the Spanish Ministry of Economy and Competitiveness (BFU2017- et al., 1999. Beta-secretase cleavage of Alzheimer’s amyloid precursor protein 89336-R to MV), co-funded by the Fondo Europeo de Desarrollo Regional (FEDER; A way by the transmembrane aspartic protease BACE. Science (New York, N.Y.) to build Europe), and by the CERCA Programme and Generalitat de Catalunya (grant 286(5440):735e741. 2014_SGR_520). DDC was supported by Conicyt’s fellowship from the Government of [13] Bennett, B.D., Babu-Khan, S., Loeloff, R., Louis, J., Curran, E., Citron, M., Chile (72170321-6357/2016). SR is a recipient of a Juan de la Cierva Incorporación et al., 2000. Expression analysis of BACE2 in brain and peripheral tissues. (IJC2018-037341-I) program from the Spanish Ministry of Science and Innovation. Journal of Biological Chemistry 275(27):20647e20651. This work was developed at the Center Esther Koplowitz (Barcelona). CIBERDEM (Centro [14] Alcarraz-Vizán, G., Casini, P., Cadavez, L., Visa, M., Montane, J., Servitja, J.M., de Investigación Biomédica en Red de Diabetes y Enfermedades Metabólicas Asociadas) et al., 2015. Inhibition of BACE2 counteracts hIAPP-induced insulin secretory is an initiative of the Instituto de Salud Carlos III. defects in pancreatic b-cells. FASEB Journal 29(1):95e104. MOLECULAR METABOLISM 53 (2021) 101251 Ó 2021 The Author(s). Published by Elsevier GmbH. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 9 www.molecularmetabolism.com Original Article [15] Voytyuk, I., Mueller, S.A., Herber, J., Snellinx, A., Moechars, D., van Loo, G., [29] Winzell, M.S., Magnusson, C., Ahrén, B., 2007. Temporal and dietary fat et al., 2018. BACE2 distribution in major brain cell types and identification of content-dependent islet adaptation to high-fat feeding-induced glucose intol- novel substrates. Life Science Alliance 1(1):e201800026. erance in mice. Metabolism: Clinical and Experimental 56(1):122e128. [16] Wang, Z., Xu, Q., Cai, F., Liu, X., Wu, Y., Song, W., 2019. BACE2, a conditional [30] Mosser, R.E., Maulis, M.F., Moullé, V.S., Dunn, J.C., Carboneau, B.A., b-secretase, contributes to Alzheimer’s disease pathogenesis. JCI Insight 4(1). Arasi, K., et al., 2015. High-fat diet-induced b-cell proliferation occurs prior to [17] Finzi, G., Franzi, F., Placidi, C., Acquati, F., Palumbo, E., Russo, A., et al., insulin resistance in C57Bl/6J male mice. American Journal of Physiology- 2008. BACE2 is stored in secretory granules of mouse and rat pancreatic b Endocrinology and Metabolism 308(7):E573eE582. cells. Ultrastructural Pathology 32(6):246e251. [31] Kwon, O., Kim, K.W., Kim, M.S., 2016. Leptin signalling pathways in hypo- [18] Casas, S., Casini, P., Piquer, S., Altirriba, J., Soty, M., Cadavez, L., et al., thalamic neurons. Cellular and Molecular Life Sciences 73(7):1457e1477. 2010. BACE2 plays a role in the insulin receptor trafficking in pancreatic b- [32] Stützer, I., Selevsek, N., Esterházy, D., Schmidt, A., Aebersold, R., Stoffel, M., cells. American Journal of Physiology - Endocrinology and Metabolism 299(6): 2013. Systematic proteomic analysis identifies b-site amyloid precursor 1087e1095. protein cleaving enzyme 2 and 1 (BACE2 and BACE1) substrates in pancreatic [19] Kennedy, M.E., Stamford, A.W., Chen, X., Cox, K., Cumming, J.N., b-cells. Journal of Biological Chemistry 288(15):10536e10547. Dockendorf, M.F., et al., 2016. The BACE1 inhibitor verubecestat (MK-8931) [33] Ramírez, S., Gómez-Valadés, A.G., Schneeberger, M., Varela, L., Haddad- reduces CNS b-amyloid in animal models and in Alzheimer’s disease patients. Tóvolli, R., Altirriba, J., et al., 2017. Mitochondrial dynamics mediated by Science Translational Medicine 8(363):363e750. mitofusin 1 is required for POMC neuron glucose-sensing and insulin release [20] Hampel, H., Vassar, R., De Strooper, B., Hardy, J., Willem, M., Singh, N., et al., control. Cell Metabolism 25(6):1390e1399 e6. 2020. The b-secretase BACE1 in Alzheimer’s disease. Biological Psychiatry [34] Alonge, K.M., D’Alessio, D.A., Schwartz, M.W., 2021. Brain control of blood 11:1e12. glucose levels: implications for the pathogenesis of type 2 diabetes. Dia- [21] Probst, G., Xu, Y., 2012. Small-molecule BACE1 inhibitors: a patent literature betologia 64(1):5e14. review (2006 - 2011). Expert Opinion on Therapeutic Patents 22(5):511e540. [35] Rosario, W., Singh, I., Wautlet, A., Patterson, C., Flak, J., Becker, T.C., et al., [22] Meakin, P.J., Harper, A.J., Hamilton, D.L., Gallagher, J., McNeilly, A.D., 2016. The brain-to-pancreatic islet neuronal map reveals differential glucose Burgess, L.A., et al., 2012. Reduction in BACE1 decreases body weight, regulation from distinct hypothalamic regions. Diabetes 65(9):2711e2723. protects against diet-induced obesity and enhances insulin sensitivity in mice. [36] Rulifson, I.C., Cao, P., Miao, L., Kopecky, D., Huang, L., White, R.D., et al., Biochemical Journal 441(1):285e296. 2016. Identification of human islet amyloid polypeptide as a BACE2 substrate. [23] Meakin, P.J., Jalicy, S.M., Montagut, G., Allsop, D.J.P., Cavellini, D.L., PloS One 11(2). Irvine, S.W., et al., 2018. Bace1-dependent amyloid processing regulates [37] Blüher, M., Kahn, B.B., Kahn, R.C., 2003. Extended longevity in mice lacking hypothalamic leptin sensitivity in obese mice. Scientific Reports 8(1):55. the insulin receptor in adipose tissue. Science 299(5606):572e574. [24] Ghosh, A.K., Brindisi, M., Yen, Y.C., Lendy, E.K., Kovela, S., [38] Page, M.M., Johnson, J.D., 2018. Mild suppression of hyperinsulinemia to Cárdenas, E.L., et al., 2019. Highly selective and potent human b-sec- treat obesity and insulin resistance. Trends in Endocrinology and Metabolism retase 2 (BACE2) inhibitors against type 2 diabetes: design, synthesis, X- 29(6):389e399. ray structure and structureeactivity relationship studies. ChemMedChem [39] Vetterli, L., Carobbio, S., Frigerio, F., Karaca, M., Maechler, P., 2016. The 14(5):545e560. amplifying pathway of the b-cell contributes to diet-induced obesity. Journal of [25] Dominguez, D., Tournoy, J., Hartmann, D., Huth, T., Cryns, K., Deforce, S., Biological Chemistry 291(25):13063e13075. et al., 2005. Phenotypic and biochemical analyses of BACE1- and BACE2- [40] Lee, K.T.Y., Karunakaran, S., Ho, M.M., Clee, S.M., 2011. PWD/PhJ and WSB/ deficient mice. Journal of Biological Chemistry 280(35):30797e30806. EiJ mice are resistant to diet-induced obesity but have abnormal insulin [26] Cui, A., Hu, Z., Han, Y., Yang, Y., Li, Y., 2017. Optimized analysis of In vivo and secretion. Endocrinology 152(8):3005e3017. In vitro hepatic steatosis. Journal of Visualized Experiments(121):e55178. [41] Shanik, M.H., Xu, Y., Skrha, J., Dankner, R., Zick, Y., Roth, J., 2008. Insulin [27] Moreno-Asso, A., Castaño, C., Grilli, A., Novials, A., Servitja, J.M., 2013. resistance and hyperinsulinemia: is hyperinsulinemia the cart or the horse? Glucose regulation of a cell cycle gene module is selectively lost in mouse Diabetes Care 31(Suppl 2):S262eS268. pancreatic islets during ageing. Diabetologia 56(8):1761e1772. [42] Chen, W., Balland, E., Cowley, M.A., 2017. Hypothalamic insulin resistance in [28] Rueden, C.T., Schindelin, J., Hiner, M.C., DeZonia, B.E., Walter, A.E., obesity: effects on glucose homeostasis. Neuroendocrinology 104(4):364e381. Arena, E.T., et al., 2017. ImageJ2: ImageJ for the next generation of scientific [43] Ono, H., 2019. Molecular mechanisms of hypothalamic insulin resistance. image data. BMC Bioinformatics 18(1):529. International Journal of Molecular Sciences 20(6):1317. 10 MOLECULAR METABOLISM 53 (2021) 101251 Ó 2021 The Author(s). Published by Elsevier GmbH. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). www.molecularmetabolism.com