Raptor levels are critical for β-cell adaptation to a high-fat diet in male mice

Objective The essential role of raptor/mTORC1 signaling in β-cell survival and insulin processing has been recently demonstrated using raptor knock-out models. Our aim was to evaluate the role of mTORC1 function in adaptation of β-cells to insulin resistant state. Method Here, we use mice with heterozygous deletion of raptor in β-cells (βraHet) to assess whether reduced mTORC1 function is critical for β-cell function in normal conditions or during β-cell adaptation to high-fat diet (HFD). Results Deletion of a raptor allele in β-cells showed no differences at the metabolic level, islets morphology, or β-cell function in mice fed regular chow. Surprisingly, deletion of only one allele of raptor increases apoptosis without altering proliferation rate and is sufficient to impair insulin secretion when fed a HFD. This is accompanied by reduced levels of critical β-cell genes like Ins1, MafA, Ucn3, Glut2, Glp1r, and specially PDX1 suggesting an improper β-cell adaptation to HFD. Conclusion This study identifies that raptor levels play a key role in maintaining PDX1 levels and β-cell function during the adaptation of β-cell to HFD. Finally, we identified that Raptor levels regulate PDX1 levels and β-cell function during β-cell adaptation to HFD by reduction of the mTORC1-mediated negative feedback and activation of the AKT/FOXA2/PDX1 axis. We suggest that Raptor levels are critical to maintaining PDX1 levels and β-cell function in conditions of insulin resistance in male mice.


INTRODUCTION
Type 2 diabetes (T2D) is characterized by defective insulin secretion and b-cell expansion in conditions of insulin resistance. b-Cell adaptation to insulin resistance is regulated by the interplay of signals from growth factors, cytokines, insulin, and nutrients such as glucose, fatty acids, and amino acids (AAs). The mTOR complex 1 (mTORC1) is activated by nutrients (AAs, glucose, lipids) and growth factors, thereby linking growth factor, diet, and nutrient excess to b-cell responses.
Abnormalities in mTORC1 signaling have been implicated in human diseases including diabetes. Published observations underscore the importance of mTORC1 signaling on modulation of b-cell mass, insulin secretion and adaptation to insulin resistance. mTORC1 is a protein complex that functions as a nutrient sensor and it is composed of mTOR itself, the regulatory-associated protein of mTOR (Raptor), mammalian lethal with mLST8, PRAS40, and DEPTOR [1]. mTORC1 controls growth, proliferation and metabolism by directly modulating 4 E-BPs and S6 kinases (S6K). One consequence of chronic mTORC1 hyperactivation is the induction of an S6K1dependent negative feedback loop leading to attenuation of AKT signaling in multiple tissues and insulin resistance [2e5]. Activation of mTORC1 by conditional deletion of TSC2 in b-cells (bTSC2 À/À ) induces improved glucose tolerance as a result of increased b-cell mass, proliferation and cell size [6]. The importance of endogenous mTORC1 signaling in b-cells has been recently demonstrated using raptor knock-out models [7e9]. These studies have shown a key role of Raptor in proliferation, size, survival, and maturation of the b-cell, and both in function and insulin processing. In vivo and in vitro studies have shown that inhibition of mTORC1 has a protective effect in conditions of excessive proinsulin misfolding by stimulating autophagy and alleviating ER stress [10,11]. On the other hand, previous studies have suggested the importance of mTORC1 in the adaptation to states of insulin resistance [10e14] including in T2D patients [15] and increased mTORC1 activity has been reported in prediabetic db/db mice compared to nondiabetic littermates [14]. Leibowitz et al. demonstrated that blocking mTORC1 by rapamycin in P. obesus caused severe impairment of b-cells function, increased b-cells apoptosis, and progression of diabetes [11]. These studies are consistent with other studies showing that the use of rapamycin has been also linked with antiproliferative effects, and alteration in the cell cycle what could negatively impact the adaptation of b-cells to insulin resistance [12]. However, several questions remain unanswered as whether reduced in endogenous mTORC1 function is critical for b-cell function in normal conditions or during b-cell adaptation to insulin resistance. Finally, the role of mTORC1 function in adaptation of b-cells to insulin resistant states has not been directly evaluated. In the present study, we use mice with heterozygous deletion of Raptor in b-cells (bra Het ) to assess whether reduced mTORC1 function is critical for b-cell function in normal conditions or during b-cell adaptation to high-fat diet (HFD). bra Het mice showed no differences at the metabolic level, islets morphology, or b-cell function when fed regular chow. Surprisingly, deletion of only one allele of Raptor is sufficient to impair insulin secretion when fed a HFD, increases apoptosis without altering proliferation rate, accompanied by a reduced levels of critical bcell genes like Ins1 and 2, MafA, Ucn3, Glut2, Glp1r, and especially Pdx1, suggesting an improper b-cell adaptation to HFD. Our data demonstrate that Raptor protein levels are also of main importance in the maintenance of the PDX1 levels and b-cell function in the adaptation of b-cell to HFD. Finally, our data showing that increasing FOXA2 levels in islets from bra Het mice exposed to HFD rescue PDX1 levels and insulin secretion highlights the importance of this pathway in HFD adaptation. These set of studies underscore a key role of Raptor levels in maintaining PDX1 levels and b-cell function in conditions of insulin resistance by controlling the mTORC1-dependent negative feedback loop.

Animal generation
RIP-Cre and raptor fl/fl mice have been previously described [8,16,17]. Mice with transgenic overexpression of a rapamycin resistant constitutively active form of S6K in b-cells (caS6K) and braKO;caS6K have been previously described [8,18]. Studies were performed on mice on C57BL6J background. Results of the experiments are shown for male mice at ages shown in figure legends. All animals were maintained on a 12 h lightedark cycle. All procedures were performed in accordance with the University of Miami-approved protocols.

Metabolic studies
Adult mice were given a normal chow diet or a high-fat diet of 60% kcal fat (D12492, Research Diets). Blood glucose levels were determined from blood obtained from the tail vein using Contour glucometer (Bayer). Fasting glucose and insulin were measured after overnight fasting. Glucose tolerance tests and GSIS were performed on overnight-fasted animals by injecting glucose intraperitoneally (2 and 3 mg kg À1 , respectively). Plasma insulin and proinsulin levels were determined using a Mouse Ultrasensitive Insulin ELISA kit and Mouse Proinsulin ELISA kit, respectively (ALPCO Immunoassays).

Immunofluorescence staining and morphometric analysis
Formalin-fixed pancreatic tissues were embedded in paraffin. Immunofluorescence staining was performed using primary antibodies described on Supplementary Table 1. Fluorescent images were acquired using a microscope (Leica DM5500B) with a motorized stage using a camera (Leica Microsystems, DFC360FX), interfaced with the OASIS-blue PCI controller, and controlled by the Leica Application Suite X (LAS X). b-Cell ratio assessment was calculated by measuring insulin and acinar areas using Adobe Photoshop 2021 in five insulin-stained sections (5 mm) that were 200 mm apart. To calculate b-cell mass, the b-cell to acinar ratio was then multiplied by the pancreas weight.
Islets number was examined using standard histological methods on stained pancreas sections with an insulin antibody. Assessment of proliferation was performed in insulin-and Ki67-stained sections. Apoptosis was determined using TUNEL assay (ApopTag Red in Situ Apoptosis Detection Kit, Chemicon) in insulin-stained sections. At least 3,000 b-cells were counted for each animal. For dispersed cell staining, islets were gently dispersed after 5 min incubation with trypsineEDTA (0.25% trypsin and 1 mM EDTA) in Hanks' balanced salt solution without Ca 2þ and Mg 2þ (Gibco Invitrogen) at 37 C followed by fixation in 4% methanol-free formaldehyde onto poly-L-lysinecoated slides. All the morphologic measurements were performed in blinded manner.

Islets studies
After islet isolation, islets were maintained at 37 C in an atmosphere containing 20% oxygen and 5% CO 2 . Insulin secretion from isolated islets was assessed by static incubation. Briefly, after overnight culture in RPMI containing 5 mM glucose and 10% FBS, islets were precultured for 1 h in KrebseRinger medium containing 2 mM glucose. Groups of 10 islets in triplicates were then incubated in KrebseRinger medium containing 2 or 16 mM glucose for 1 h. Secreted insulin in the supernatant and insulin content was then measured using Mouse Ultrasensitive Insulin ELISA kit (ALPCO Immunoassays) and normalized to DNA content. Isolated islets were treated in vitro with proinflammatory cytokines (IL-1b (50U/ml), IFN-g and TNF-a (1000U/ml) (Peprotech, Thermo Fisher Scientific), thapsigargin (1 mM) or palmitate (0.4 mM) (Millipore, Bedford, MA). After 24 h treatment, islets were dispersed into a single cell suspension and fixed for flow cytometry analysis.

Western blotting
Islets from an individual mouse (120e150 islets) were lysed in lysis buffer (125 mM Tris, pH 7, 2% SDS and 1 mM dithiothreitol) containing a protease inhibitor cocktail (Roche Diagnostics). Protein quantity was measured by a bicinchoninic acid assay method, and 40 mg of protein were loaded on SDSePAGE gels and separated by electrophoresis. Separated proteins were transferred onto polyvinylidene difluoride membranes (Millipore, Bedford, MA) overnight. After blocking for 1 h in Li-Cor Blocking buffer, membranes were incubated overnight at 4 C with a primary antibody diluted in 1 Â Tris-buffered salinee1% Tween 20e5% milk followed by 1 h incubation at room temperature with horseradish peroxidase-conjugated secondary antibodies. Antibodies used for immunoblotting are included in Supplementary Table 1, and membranes were developed using Western Bright Sirius Kit (Bio-Express). Band densitometry was determined by measuring pixel intensity using NIH Image J software (v1.49 d [19] freely available at http://rsb.info.nih.gov/ij/index.html) and normalized to tubulin, actin or total protein in the same membrane. Images have been cropped for presentation. Full-size images for the most important western blots are presented in Supplementary Figures.

Flow cytometry
After overnight culture in RPMI containing 5 mM glucose, islets were dispersed into a single-cell suspension and fixed with BD Pharmingen Transcription Factor Phospho Buffer Set (BD Biosciences). Dispersed cells were incubated overnight with conjugated antibodies at 4 C. Dead cells were excluded by Ghost Dye Red 780 (Tonbo), and signal intensity from single stained cells and GFP was analyzed by mean fluorescent intensity in insulin-positive cells using BD LSR II (BD Biosciences). Antibodies used are included in Supplementary Table 1. 2.7. Quantitative real-time PCR Total RNA was isolated using RNeasy (Qiagen) followed by cDNA synthesis using High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems) according to the manufacturer's protocol. Real-time PCR was performed on an ABI 7000 sequence detection system using POWER SYBR-Green PCR Master MIX (Applied Biosystems). Primers were purchased from IDT Technologies. Primer pair for: pdx1 was as follows: 5 0 -CCC CAG TTT ACA AGC TCG CT-3 0 (forward); 5 0 -CTC GGT TCC ATT CGG GAA AGG-3 0 (reverse), ins1: 5 0 -CAC CCC ACC TGG AGA CCT TA-3 0 (forward); 5 0 -TGA AAC AAT GAC CTG CTT GCT G-3 0 (reverse) and ins2: 5 0 -GCA AGC AGG AAG GTT ATT GTT TCA-3 0 (forward); 5 0 -GCT TGA CAA AAG CCT GGG TG-3 0 (reverse).
2.8. RNA-Seq library preparation, sequencing, and data analysis RNA Quality Control and DNase Treatment: A total of 12 mice islets RNA samples were submitted to Ocean Ridge Biosciences (Deerfield Beach, FL) for mRNA-Sequencing. Total RNA was quantified by O.D. measurement and assessed for quality on a 1% agarose e 2% formaldehyde RNA Quality Control (QC) gel. The RNA was then digested with RNase free DNase I (Epicentre; Part #D9905K) and re-purified using Agencourt RNAClean XP beads (Beckman Coulter; Part # A63987). The newly digested RNA samples were then quantified by O.D. measurement. The newly digested RNA samples were then quantified by O.D. measurement and checked for quality. Library Preparation: Amplified cDNA libraries suitable for sequencing were prepared from 250 ng (ng) of DNA-free total RNA using the TruSeq Stranded mRNA Library Prep (Illumina Inc.; Part # 20,020,595). The quality and size distribution of the amplified libraries were determined by chip-based capillary electrophoresis (Bioanalyzer 2100, Agilent Technologies). Libraries were quantified using the KAPA Library Quantification Kit (Kapa Biosystems, Boston, MA).

Sequencing
The 8 libraries were pooled at equimolar concentrations and sequenced in a total of 3 runs on the Illumina NextSeq 500 sequencer using two Mid Output v2 150 cycle kits (part# FC-404-2001) and one High Output v2.5 150 cycle kit (part# 20,024,907). In each case the libraries were sequenced with 76 nt paired-end reads plus 8 nt dual-index reads on the instrument running Next-Seq Control Software version 2.2.0.4. Real time image analysis and base calling were performed on the instrument using the Real-Time Analysis (RTA) software version 2.4.11. Generation of FASTQ files: Base calls from the NextSeq 500 RTA were converted to sequencing reads in FASTQ format using Illumina's bcl2fastq program v2.17.1.14 with default settings. Sequencing adapters were not trimmed in this step.

Adenoviral infection
After overnight culture in RPMI containing 5 mM glucose, islets were infected with adenoviruses carrying the cytomegalovirus promoter (Ad. CMV) or FOXA2 and GFP under the control of the CMV promoter (Ad. FOXA2-GFP (ADV-209226), Vector BioLabs). The particle:plaqueforming unit ratio of the stock virus used in the experiments was 300.
2.11. Statistical analysis Data are presented as mean AE s.e.m. and were considered statistically significant when the P value was <0.05. Student's unpaired t test was used to assess statistical difference between 2 groups using Prism version 9 (GraphPad Software, San Diego, CA). Comparison between more than 2 groups was performed using 2-way ANOVA with repeated measures followed by post hoc 2-tailed Student's t tests. The results were considered statistically significant when the p value was equal than 0.05.

Data availability
All relevant data are available from the authors on request.

Heterozygous raptor deletion in b-cells exhibits normal glucose homeostasis
To decrease endogenous mTORC1 function, we generated mice with heterozygous deletion of Raptor in b-cells by crossing raptor f/f with Rip-Cre mice (bra Het ) ( Figure 1A), both previously described [8,16,17]. To further test if decreased Raptor levels would affect mTORC1 activity, we performed starvation/refeeding with AAs to assess the phosphorylation of S6. Islets from bra Het exhibit decreased phosphorylation of S6 after stimulation with amino acids, indicating that b-cells from bra Het are not able to fully activate mTORC1 (Supp Figure. 1a). However, weight, random fed and fasting blood glucose and insulin levels were normal in the bra Het mice at 3 months of age ( Figure 1BeF). Examination of glucose tolerance and glucose-stimulated insulin secretion in bra Het mice showed no differences when compared to the control mice (raptor f/ f and Rip-Cre) at 3 months of age ( Figure 1GeH). These studies suggest that, in contrast to braKO mice (homozygous deletion of raptor in b-cells) [8], deletion of one raptor allele (bra Het ) displayed normal glucose levels and glucose tolerance in regular chow (RC). Consistent with the results in glucose homeostasis, bra Het mice exhibited normal b-cell mass with similar levels in proliferation, survival, and cell size ( Figure 1I-L). While we previously demonstrated that Raptor/mTORC1 is necessary for maintaining postnatal b-cell mass by controlling apoptosis, size, and proliferation [8], the heterozygous deletion of Raptor does not appear to affect the maintenance of postnatal b-cell mass.

Heterozygous raptor deletion in b-cells exhibits normal insulin secretion and intracellular calcium responses ex vivo
The results obtained with the bra Het mice indicate that decreased mTORC1 is not critical for b-cell mass maintenance. While glucose tolerance was normal in vivo, assessment of insulin content demonstrated a significant decrease in islets from bra Het mice ( Figure 2A); however, the decrease in insulin content was not due to a decrease in mRNA levels ( Figure 2B). Moreover, and in contrast to braKO mice [8], proinsulin levels were not increased in b-cells from bra Het mice compared to controls ( Figure 2C). To further characterize the normal glucose tolerance, we measured b-cell function by static incubation and glucose-mediated calcium imaging in islets from bra Het and control mice. Glucose-dependent insulin release by static incubation exhibited no differences between islets from bra Het and control ( Figure 2D). To complement the insulin secretory responses, Ca 2þ imaging using the calcium indicator Fluo-4AM was determined in islets from bra Het and control mice. Intracellular Ca 2þ responses to secretagogues such as 16 mM glucose, tolbutamide, or 30 mM KCl were not significantly different between bra Het and control islets (Figure 2EeG). These studies demonstrate that although b-cells from bra Het mice exhibit a slight decrease in insulin content, this was not sufficient to alter glucose tolerance at 3 months old (Figure 1CeH) and the responses to glucose and other secretagogues in regular conditions were conserved.
3.3. bra Het mice exhibit impaired glucose homeostasis in HFD Given that bra Het mice fed-RC showed no significant differences in glucose homeostasis, islet morphology and b-cell function, we decided to assess the adaptation to insulin resistance by administering HFD for these results, glucose-stimulated insulin secretion (GSIS) was also impaired in bra Het mice at 10 weeks of HFD ( Figure 3G). In agreement with these results, although random-fed insulin values in bra Het mice were increased after 12 weeks of HFD, this increase was significantly less than in controls ( Figure 3H).  (Figure 4CeE). We next investigated the mechanisms responsible for b-cell loss in bra Het mice in HFD by assessing apoptosis in isolated islets exposed for 24 h to ER stress, oxidative stress inducers such as proinflammatory cytokines, and lipotoxicity. Assessment of apoptosis in insulin positive cells by cleaved-caspase 3 levels in dispersed islets using FACs showed no differences between bra Het and control b-cells after treatment with proinflammatory cytokines (IL1-b (50U/ml), TNF-a and IFN-g (1000U/ml) ( Figure 4F). Unexpectedly, b-cells from bra Het mice were more resistant to thapsigargin ( Figure 4G). In contrast, b-cells from bra Het mice were more susceptible to apoptosis induced by lipotoxicity ( Figure 4H), indicating that b-cells from bra Het islets are particularly sensitive to conditions of excess of lipids. To further dissect the effect of glucose and palmitate on mTORC1 activation, we designed experiments to assess mTORC1 activity by S6 phosphorylation in high glucose and glucolipotoxicity conditions. No difference in mTORC1 activation measured by S6 phosphorylation was found when the islets were cultured only with high glucose (16 mM) for 24 h ( Figure 4I). In contrast, a 30% decreased in S6 phosphorylation was observed when the islets were cultured with high glucose þ palmitate ( Figure 4J). Together, these data demonstrate that deletion of one allele of Raptor has no effect in activation of mTORC1 signaling by high levels of glucose, but it is involved in the response to high glucose and palmitate.
3.5. b-cell function is affected in bra Het mice fed HFD We next designed experiments to assess the mechanisms responsible for the abnormalities in insulin secretion in bra Het and control mice fed HFD. Similar to the findings in RC (Figure 2A,C), islets from bra Het mice fed HFD also exhibit a decrease in insulin content ( Figure 5A vs 2a) and conserved insulin processing measured by proinsulin levels ( Figure 5B vs 2c). Dynamic glucose-responsive insulin secretion by islet perifusion studies using isolated islets from bra Het mice fed HFD showed a decreased response to 16 mM glucose and KCl when compared to control islets (Figure 5CeD). All together, these data suggest that exposure to HFD uncovered a defect in b-cell function in bra Het mice.
3.6. bra Het mice fed HFD exhibit decreased levels of b-cell identity markers The previous results indicate that b-cell function and mass in bra Het islets are reduced after HFD. To obtain a mechanistic insight for these abnormalities, we performed RNAseq in islets from control mice fed regular chow (RC), control mice fed HFD (HFD), and bra Het fed HFD (bra Het þ HFD) for 8 weeks. Analyzes of tissue-cell-specific genes related and acinar genes and using the meta-analysis tools Metascape [20] and TRRUST [21] to evaluate acinar cell contamination in the RNAseq datasets (Supp Figures. 2aeb) shows no significant differences among the three groups. Next, we determined the effects of heterozygous deletion of Raptor in HFD by comparing the RC, HFD and bra Het þ HFD data sets. The unbiased RNAseq analysis revealed 962 affected genes (RC vs HFD) ( Figure 6 and Supp Figure. 2c). This analysis of key b-cell identity genes showed that mRNA for Ins1, MafA, Ucn3, Ppp1r1a, Pdx1 and Ero1b were significantly increased in HFD compared to RC ( Figure 6A). The increase in these identity genes by HFD was significantly reduced in bra Het þ HFD ( Figure 6A). More importantly, Pdx1 mRNA was also lower in bra Het þ HFD mice when compared to HFD littermates ( Figure 6A). To identify the biological processes induced by HFD and the role of Raptor heterozygous deletion, we first performed Gene Ontology (GO) analysis on differentially expressed genes (DEGs) between RC and HFD groups ( Figure 6B). This analysis showed that HFD regulates critical genes for b-cell function such as hormone/protein secretion, insulin secretion, secretory pathways, and vesicle exocytosis ( Figure 6B). Then, we performed GO analysis on DEGs between HFD and bra Het -HFD mice ( Figure 6C). This study showed that b-cells with heterozygous Raptor deletion displayed a decreased in insulin/hormone/protein secretions, and proteolysis defects among others ( Figure 6C). The abnormalities in insulin secretion and increase in apoptosis are reminiscent of the phenotype observed in PDX1 heterozygous mice [22,23] and led us to hypothesize that PDX1 reduction could be responsible, at least in part, for the bra Het -HFD phenotype. In addition, previous studies by Hagman et al.
showed that palmitate inhibits insulin gene expression by reducing the binding of PDX1 and MafA to the insulin promoter [24]. To test this hypothesis, we first assessed PDX1 levels in HFD and bra Het -HFD. A decrease in PDX1 was observed at the protein level by immunostaining and immunoblotting in bra Het -HFD mice for 12 weeks (Figure 6DeE).
Evaluation of PDX1 protein levels by immunoblotting showed no differences in islets from bra Het compared to control littermates in RC (Supp Figure. 3). The reduction in PDX1 levels exclusively in bra Het -HFD could explain in part the differences in b-cell area and function in HFD but not in RC (Figures 1-2). Analysis of PDX1 target genes in the RNAseq data shows that in addition to MafA, other important downstream targets of PDX1 such as Glp1r, and Slc2a2 (Glut2) [25,26] were also decreased in bra Het -HFD mice compared to HFD mice ( Figure 6F).
3.7. bra Het mice fed HFD exhibit decreased levels of PDX1 and nuclear FOXA2 Previous work has shown that PDX1 is regulated by different transcription factors, including HNF-3b/FOXA2, HNF6, HNF-1a, HNF-1b, SP1/3, USF1/2, and PDX-1 itself [27,28]. Analysis of these transcription factors in the RNAseq data showed no differences in these Pdx1 transcriptional regulators among RC, HFD and bra Het -HFD suggesting that some of these transcription factors could be regulated at the protein level or cellular localization (Supp Figure. 4a). Interestingly, analysis by TRRUST of PDX1 regulators and transcription factors that share targets with PDX1 confirm the synergism of HNF6, MafA, FOXA2 and PDX1 in regulation of b-cell identity (Supp Figures. 4bed) [25].
FOXA2 has been previously implicated in pancreatic development and b-cell maturation acting upstream of PDX1 [27,29], and its cooperative function with PDX1 is critical for proper b-cell function [25]. In addition, our group recently identified a novel link between mTORC1 and FOXA2 in transcriptional regulation in a-cells [30]. Based on these studies, we hypothesized that FOXA2 and PDX1 could be involved in the adaptation defect observed in bra Het -HFD. FOXA2 expression levels were similar in control and bra Het islets from mice fed HFD for 12 weeks ( Figure 7A).
To assess if FOXA2 activation and nuclear/cytosolic localization plays a role in this process we assessed the phosphorylation of FOXA2.
Phosphorylation of FOXA2 in T156 was significantly elevated in bra Het -HFD when compared to control-HFD but not in RC ( Figure 7B and Supp Figure. 5). This suggests that AKT activity could be differentially induced in bra Het -HFD as T156 phosphorylation of FOXA2 is mediated by AKT [31]. Since AKT has been previously shown to regulate PDX1 transcription by inducing phosphorylation and exclusion of FOXA2 to the cytosol [31], we assessed FOXA2 cellular localization. Immunostaining and quantification of pancreas sections demonstrated a 70% decreased in nuclear FOXA2 staining in b-cells from bra Het mice fed HFD ( Figure 7C). Indeed, AKT (T308) phosphorylation was significantly elevated in bra Het -HFD islets compared to control-HFD and it was increased by almost 3-fold when compared to islets from bra Het RCfed mice ( Figure 7D). Consistent with reduction in mTORC1 activity, phosphorylation of S6 (Ser240) was totally blunted in bra Het -HFD ( Figure 7E). Therefore, we hypothesized that the decreased levels of S6 phosphorylation in response to palmitate ( Figure 4J) and S6 phosphorylation in bra Het -HFD was associated with a decreased in the previously described mTORC1-mediated feedback inhibition on IRS1/2 [3,4,32] and AKT phosphorylation, ultimately affecting PDX1 levels.  Figure. 6). As previously shown, PDX1 levels were reduced in islets from bra Het compared to controls mice ( Figure 8A). PDX1 levels are reduced by 20% in islets from control mice fed HFD cultured with the control adenovirus ( Figure 8A). Interestingly, an increase in FOXA2 levels is sufficient to rescue PDX1 levels significantly in bra Het islets similar to the control þ adenovirus level ( Figure 8A). Moreover, to test whether these increases in FOXA2 and PDX1 levels are also sufficient to rescue b-cell function in bra Het islets, a similar experiment was performed to evaluate insulin secretion. Confirming our hypothesis and previous results, insulin secretion was impaired in bra Het islets and rescued to control levels when FOXA2 was overexpressed similarly to PDX1 levels in Figure 8A ( Figure 8B). To strength our conclusions, we assessed PDX1 levels by immunostaining in control, braKO, and braKO mice overexpressing a constitutively active form of S6K (braKO;caS6K) previously described [8]. As previously published, PDX1 levels were decreased in b-cells with Raptor deletion [7,9]. Notably, genetic reconstitution of S6K activity increases PDX1 levels in b-cells with Raptor deletion ( Figure 8C).

DISCUSSION
The current studies provide novel insights into how mTORC1 regulates the adaptation of b-cell to HFD. We showed that deletion of one raptor allele is sufficient to impair glucose homeostasis by a defect in insulin secretion. The studies herein also identified that 1) heterozygous deletion of Raptor in b-cells renders these cells more susceptible to lipotoxicity and HFD conditions, 2) mTORC1 regulates b-cell genes that are critical for the adaptation to HFD such as Ins1, MafA, Ucn3, Glut2 and Glp1r, 3) a novel AKT/FOXA2/PDX1 axis activation regulating b-cell adaptation to HFD in conditions of decrease in mTORC1 signaling ( Figure 8D). The increase in FOXA2 levels rescuing PDX1 levels and insulin secretion in islets with heterozygous deletion of raptor suggest that this pathway could play a role in humans treated with mTORC1 inhibitors [33,34]. In addition, it has been hypothesized that in conditions that require adaptive cell proliferation, such as weight increase or metabolic syndrome, mTORC1 inhibitors could contribute to the development of new-onset diabetes after transplantation (NODAT) development [4,35e38]. Therefore, studying the molecular interactions between mTORC1 signaling, FOXA2, and PDX1 and the downstream effects of this pathway in human islets from donors treated with mTORC1 inhibitors, would provide clinical insight, and guide the development of personalized approaches to (braKO), decreased Raptor levels were not sufficient to alter proinsulin content [8]. Although insulin content per cell was significantly decreased in bra Het mice, static glucose-stimulated insulin secretion and intracellular calcium levels were similar to control mice ( Figure. 2). In addition, islet morphology was also comparable to control mice ( Figures. 1I-L), demonstrating that a decrease in raptor levels is not critical to maintaining adult b-cell mass and function. Together, our data demonstrate that partial Raptor deletion has no deleterious effects on b-cell mass and function in normal diet conditions. Furthermore, the decrease in raptor levels appears to have no impact on maintaining basal levels of pS6 ( Figures 1A and 4I). This finding is consistent with the absence of a phenotype in normal diet conditions. However, notable differences in pS6 levels emerge when mTORC1 activity is challenged with palmitate or amino acids after starvation and this explains in part the phenotype of these mice when exposed to HFD ( Figure 4J and Supp Figure. 1). mTORC1 activity is highly upregulated in the liver, fat, muscle and pancreatic islets of obese and high-fat-fed rodents and in islets of T2D humans [2,5,44e47]. One consequence of chronic mTORC1 hyperactivation is the induction of an S6K1-dependent negative feedback loop leading to attenuation of AKT signaling in multiple tissues and insulin resistance [2e5]. The current published evidence supports the concept that mTORC1 has a biphasic regulatory pattern that is consistent with the widely accepted model of b-cell deterioration 'compensation/decompensation switch' during the progression of T2D [15,47e49]. Metabolic stressors such as insulin resistance and nutrient excess increase b-cell mTORC1 in the initial functional compensatory phase [50]. This correlates with hyperinsulinemia and compensatory b-cell hypertrophy and hyperplasia, suggesting that mTORC1 is a key positive regulator of b-cell function and mass [50].
Consistent with this concept, previous studies have shown an improvement of the b-cell mass and function in states of insulin resistance with mTORC1 inhibitors [10,11]. Thus, it would have been anticipated that glucose homeostasis would have been improved in the bra Het mice when fed HFD. Surprisingly, bra Het mice exhibit defective adaptation to HFD with b-cell area reduction by increase in apoptosis and Ucn3 among others [7,9]. Further, published and current studies are consistent with a model in which complete Raptor deletion results in low Ucn3 and Glut2 followed by a defect in b-cell identity [7].
However, the conditions of complete mTORC1 inhibition are not observed in physiology or during disease states. Therefore, the current work extends previous published studies by showing that mTORC1 activity is required for b-cell adaptation in a model of T2D by maintaining the activity of the mTORC1-mediated feedback inhibitory loop ( Figure. 6). This is consistent with the concept that mTORC1 activation could initially play a physiological role in adaptation nutrient excess and obesity. However, chronic mTORC1 hyperactivation caused by sustained nutrient overload induces an mTORC1/S6K1-dependent negative feedback loop causing b-cell exhaustion, functional collapse and ultimate cell death (decompensatory phase) [56e59]. Finally, the resistance of bra Het b-cells to thapsigargin induced apoptosis is intriguing and likely consistent with previously published experiments showing that inhibition of mTORC1 has a protective effect alleviating ER stress possibly by decrease in protein synthesis [10,11,60]. Taken together, these studies suggest that decrease mTORC1 activity can regulate b-cells survival to specific stressors.
Previous work by our group and others have shown that PDX1 levels are regulated by mTOR signaling [7,9,40,42,61]. However, how mTOR regulates PDX1 has not been directly explored. Mice with b-cellspecific deletion of Raptor exhibit a reduction in Pdx1 mRNA expression [9]. Using heterozygous deletion of raptor in b-cells, we show normal PDX1 levels when fed regular chow (Sup. Figure. 3). In HFD, PDX1 mRNA levels increase in controls but this increase is limited in bra Het mice ( Figure. 6A). This limited increase in PDX1 levels was confirmed at the protein levels by WB and staining (Figure 6DeE). Recently, our group identified a novel mTORC1/FOXA2 axis as a link between mTORC1 and transcriptional regulation of key genes responsible for a-cell function and survival [30]. In b-cells, FOXA2 has been previously proposed to regulate PDX1 levels [27] and low PDX1 levels have been associated with dedifferentiation and impaired insulin secretion [25,27,62]. Given the key role of FOXA2 in b-cell regulating maturation, PDX1 expression and insulin secretion, we considered FOXA2 as a potential candidate to regulate PDX1 in HFD conditions in bra Het mice [25,27,63e65]. We found that although there were no differences in the total levels of FOXA2, the levels of phosphorylated FOXA2 were significantly higher and this resulted in retention of FOXA2 in the cytoplasm of b-cells from the bra Het mice in HFD compared to controls (Figure 7BeC). The mechanistic role of FOXA2 in the defects observed in bra Het mice in HFD was further confirmed by rescuing PDX1 levels and insulin secretion. Together, these data suggest that in conditions of reduced mTORC1, the AKT/FOXA2 axis activation plays a critical role in regulating PDX1 levels, a key transcription factor necessary for b-cell adaptation to HFD.

CONCLUSION
In summary, our results uncover that Raptor levels are critical in bcells adaptation to insulin resistance but not in normal conditions and demonstrate the importance of the negative feedback inhibition of mTORC1/S6K on IRS/AKT signaling. Deletion of only one allele of raptor is sufficient to impair insulin secretion when fed a HFD, increases apoptosis without altering proliferation rate, and this is accompanied by abnormalities in transcription of critical b-cell genes including Pdx1.
These findings also support the concept that a decrease in the mTORC1/S6K negative feedback loop leads to an increase in AKT activity and FOXA2 phosphorylation/cytoplasmic retention leading to reduced levels of PDX1. The importance of this AKT/FOXA2/PDX1 axis in adaptation of b-cells to insulin resistance is only observed in conditions of reduced mTORC1 activity. The regulation of this axis could have implications in humans treated with mTOR inhibitors. Finally, these studies also reveal that increasing FOXA2 levels in islets with reduced mTORC1 activity rescue PDX1 levels and insulin secretion, highlighting the importance of the AKT/FOXA2/PDX1 axis as a possible therapeutic tool to improve b-cells in conditions of insulin resistance.