SIRT6 Is a Key Regulator of Pancreatic β-Cell Survival and Function During Aging

Abstract

Pancreatic β-cells undergo senescence and loss during aging; however, the underlying mechanisms remain incompletely understood. This study aimed to investigate what sirtuin 6 (SIRT6) does during β-cell aging. Pancreatic β-cell–specific Sirt6 transgenic (TgSIRT6) mice were generated for this study. DNA damage, cell death, and cell proliferation were analyzed in cell and mouse models. SIRT6 protein levels were decreased in pancreatic β-cells during aging. TgSIRT6 mice exhibited less DNA damage and cell death, including apoptosis, necroptosis, and pyroptosis, in β-cells than control mice. TgSIRT6 mice had increased total islet area and mass in pancreas compared with control mice. As a result, TgSIRT6 mice showed better glucose tolerance and glucose-stimulated insulin secretion than control mice. RRAD and GEM-like GTPase 2 (REM2), an endogenouse inhibitor of high-voltage–activated calcium channels, was negatively regulated by SIRT6. Knockdown of Rem2 in INS-1 cells partially rescued the SIRT6 deficiency– and palmitic acid–induced DNA damage, lipid peroxidation, and cell death. Rem2 β-cell–specific knockout mice had less DNA damage and cell death in β-cells than control mice. Our data suggest that SIRT6 is a critical antiaging factor in pancreatic β-cells and is a potential therapeutic target.

Article Highlights

• Pancreatic β-cell function declines with age, but the underlying mechanism is poorly understood.

• In this study, we attempted to address how to reverse β-cell aging.

• Our data showed that sirtuin 6 (SIRT6) overexpression can reduce age-associated DNA damage, cell death, and functional decline in β-cells.

• Our findings suggest that improving Sirt6 gene expression and function may slow down β-cell decline in older patients.

Graphical Abstract

Introduction

Aging is a major biological driver of metabolic dysfunction, including impaired pancreatic β-cell function (1). As individuals age, the prevalence of metabolic disorders, such as type 2 diabetes (T2D), rises significantly, reaching greater than 20% among those aged 65 or older (1). Despite the clear link between aging and T2D, the molecular mechanisms underlying β-cell dysfunction during aging remain poorly understood.

Pancreatic β-cells undergo a decline in both mass and function with age (2–4). This decline is exacerbated by oxidative stress, DNA damage, and cell death, which contribute to reduced insulin secretion and impaired glucose metabolism (1,5,6). Understanding the molecular mechanisms of β-cell aging is therefore crucial, not only for addressing diabetes but also for mitigating broader metabolic dysfunction associated with aging.

Sirtuin 6 (SIRT6), a member of the NAD⁺-dependent deacetylase family, has emerged as a key regulator of aging and metabolic homeostasis (7,8). SIRT6 is involved in maintaining genomic stability, promoting DNA repair, and regulating metabolic homeostasis (7,9–11). Sirt6 gene expression and enzymatic activity decline with age in pancreatic islets, which parallels with the progression of β-cell dysfunction (6). SIRT6 deficiency in β-cells leads to mitochondrial dysfunction, disrupted calcium homeostasis, and impaired glucose-stimulated insulin secretion (GSIS) in mice, whereas SIRT6 overexpression has been shown to enhance insulin secretion in mice and reduce palmitic acid–induced β-cell apoptosis in mouse MIN6 insulinoma cells (6,12–15). To further investigate the role of SIRT6 in aging-related β-cell survival and function, we selected RRAD and GEM-like GTPase 2 (REM2) as a potential downstream candidate, because both are involved in calcium dynamics and insulin secretion. REM2 has been reported to be a potent endogenous inhibitor of high-voltage–activated calcium channels, such as Ca_v1.2 (encoded by the CACNA1C gene) and Ca_v2.1 (encoded by the CACNA1A gene), which are critical for calcium influx during insulin secretion (16). REM2 also inhibits Ca²⁺/calmodulin-dependent protein kinase II (CaMKII) in neurons (17). The Rem2 gene can be induced by 25 mmol/L glucose in mouse MIN6 insulinoma cells. Overexpression of Rem2 in MIN6 cells potently inhibits high GSIS (18).

In this study, we attempted to address the role of SIRT6 and REM2 in pancreatic β-cell aging using both cell and animal models.

Research Design and Methods

Animals

We have developed a conditional Sirt6 transgenic mouse line by placing a floxed STOP cassette in front of the human Sirt6 gene coding sequence fused with a 3× HA tag sequence as previously described (19). Sirt6 conditional transgenic mice were crossed with Ins1-Cre/ERT2 mice obtained from The Jackson Laboratory (strain no. 026802) to generate Sirt6 β-cell–specific transgenic (TgSIRT6) mice. Tamoxifen was administered to both control (including floxed mice and Cre mice) and TgSIRT6 mice via oral gavage at a dosage of 4 mg per mouse in corn oil for 4 consecutive days. For the 10- and 16-month age groups, tamoxifen was administered when the mice reached 5 and 11 months of age, respectively.

Rem2 β-cell–specific knockout (Rem2 βKO) mice were generated by crossing Rem2 floxed mice (provided by Dr. Daniel J. Liput, Laboratories of Molecular Physiology, National Institute on Alcohol Abuse and Alcoholism, National Institutes of Health [NIH]) with Ins1-Cre/ERT2 mice. Rem2 gene knockout was initiated by injecting tamoxifen at 5 months of age, and phenotyping was performed at 10 and 16 months of age.

TgSIRT6 and control wild-type (WT) mice were maintained on a normal chow diet containing 18% of kilocalories from fat (cat. no. 2018SX; Envigo Bioproducts, Inc.). To add environmental stress to pancreatic β-cells, Rem2 βKO and control mice (including floxed mice and Cre mice) were treated at 5 months of age with a high-fat diet containing 60% of kilocalories from fat (cat. no. D12492; Research Diets, Inc.) for 11 weeks and then switched to normal chow until euthanasia. TgSIRT6 mice and their age-matched control littermates were subjected to the same experimental procedures at the same time. All animal procedures followed the NIH Guide for the Care and Use of Laboratory Animals and were approved by the Indiana University School of Medicine Institutional Animal Care and Use Committee. Data were collected mostly from male mice, and some experiments were performed using both male and female mouse samples. Because of the limited number of female mice, sex differences were not statistically analyzed.

Human Pancreatic Specimens

Formalin-fixed paraffin-embedded human pancreatic specimens were provided by the National Institute of Diabetes and Digestive and Kidney Diseases–funded Integrated Islet Distribution Program (Research Resource Identifier SCR_014387) at City of Hope (NIH grant U24DK098085). Human participant information is described in Supplementary Table 1.

RNA Sequencing and Data Analysis

Primary pancreatic islets were isolated from 10- and 16-month-old control and TgSIRT6 mice at the Islet and Physiology Core of the Indiana University Center for Diabetes and Metabolic Diseases as previously described (20). Total RNAs were extracted from the pancreatic islets using the RNeasy Plus Mini Kit (Qiagen). RNA integrity was verified using the Agilent Bioanalyzer, and only samples with RNA integrity numbers greater than 8 were used. Libraries were prepared using the Stranded mRNA Prep Kit (Illumina) and sequenced on the Illumina NovaSeq 6000 platform. Gene Ontology terms, Kyoto Encyclopedia of Genes and Genomes pathways, and Hallmark gene sets with an adjusted P value less than 0.05 were considered significantly enriched.

Cell Culture and DNA Constructs

INS-1 cells (originally from Dr. Christopher Newgard) were cultured in RPMI 1640 medium; supplemented with 2 mmol/L l-glutamine, 1 mmol/L sodium pyruvate, 10 mmol/L HEPES, 0.05 mmol/L β-mercaptoethanol, 100 units/mL penicillin, 100 μg/mL streptomycin, and 10% FBS; and maintained at 37°C in 5% CO₂ as previously reported (21). The cells had no mycoplasma contamination. Palmitic acids (Millipore Sigma) were conjugated with fatty acid–free BSA and introduced into the culture medium at a final concentration of 0.2 mmol/L as previously described (6). For DNA constructs, human SIRT6 and the catalytically inactive SIRT6 H133Y were cloned into a pcDNA3.1 vector, and single guide RNAs (sgRNAs) targeting rat Sirt6 and Rem2 were designed using the GPP sgRNA designer tool on the Broad Institute website and cloned into a lentiCRISPR v2 vector (gift from Dr. Feng Zhang; plasmid no. 52961; Addgene).

Quantitative PCR

Total RNAs were isolated using TRI Reagent (Millipore Sigma), and cDNA was synthesized using a high-capacity cDNA reverse transcription kit (Thermo Fisher Scientific). Quantitative PCR reactions were performed with the gene primers described in Supplementary Table 2. Relative gene expression changes were analyzed using the 2^(−ΔΔCt) method (22).

Western Blot Analysis

Protein extracts were prepared in lysis buffer (50 mmol/L HEPES, pH 7.5, 150 mmol/L NaCl, 10% glycerol, 1% Triton X-100, 1.5 mmol/L MgCl₂, 1 mmol/L EGTA) supplemented with a protease inhibitor cocktail (Millipore Sigma). Protein samples were resolved on an SDS-PAGE gel and transferred to nitrocellulose membranes (Santa Cruz Biotechnology). Membranes were incubated with primary antibodies as described in Supplementary Table 2. Proteins were detected using horseradish peroxidase–conjugated secondary antibodies, followed by enhanced chemiluminescence detection reagents (Thermo Fisher Scientific). Protein signals were scanned using the ChemiDoc MP System (Bio-Rad Laboratories). Quantitative analysis was performed using ImageJ software.

Glucose Tolerance Tests and Serum Insulin Assays

Glucose tolerance testing and serum insulin analysis were performed 1 week before animal euthanasia. After a 6-h fast, mice were injected intraperitoneally with 2 g/kg d-glucose. Blood glucose levels were measured from tail blood using a Contour glucometer (Ascensia Diabetes Care) 0, 15, 30, 60, and 120 min after glucose injection. Area under the curve analysis was performed using GraphPad Prism 10 (La Jolla, CA). For blood insulin analysis, blood samples were collected from the tail vein at 0 and 15 min after glucose injection. Blood samples were centrifuged at 5,000g for 5 min at 4°C, and sera were collected. Serum insulin levels were determined using the Ultra Mouse Insulin ELISA Kit (cat. no. 90080; Crystal Chem, Inc.) following the manufacturer’s instructions.

Histological Analysis

Mouse pancreatic tissue specimens were fixed in formalin and processed for paraffin embedding and sectioning at the Indiana University Histology Core. Tissue sections (5 μm thick) were deparaffinized in xylene and rehydrated in a graded ethanol series. Tissue sections were stained with hematoxylin-eosin. Images were captured using a Leica DM750 microscope with an EC3 digital camera and processed using ImageJ. Islet areas were quantified using ImageJ. Islet boundaries were manually traced for calculation of the total islet area in pixels. A total of 100 islets were randomly selected and measured for each group. For β-cell mass analysis, mouse pancreata were weighed, laid flat in cassettes, fixed for 4 h in 4% paraformaldehyde, dehydrated, and embedded in paraffin. Longitudinal pancreatic sections were cut at a thickness of 5 μm, collected at 250-μm intervals, and plated on glass slides. This resulted in the collection of six sections per pancreas. Sections were immunolabeled with guinea pig anti-insulin and horseradish peroxidase–conjugated donkey anti–guinea pig antibodies (Vector Laboratories). Insulin immunolabeling was visualized with an immunohistochemistry kit (Vector Laboratories), and sections were counterstained with Eosin Y (Thermo Fisher Scientific). All images were acquired on a slide scanner at ×20 magnification.

Immunofluorescence Microscopy

Antigen retrieval was performed in 10 mmol/L sodium citrate buffer for 20 min. Tissue sections were blocked with 10% donkey serum for 1 h at room temperature, followed by three washes with PBS. Immunofluorescence was performed using the primary antibodies described in Supplementary Table 2 and detected using fluorophore-conjugated secondary antibodies. INS-1 cells grown in glass-bottom dishes were fixed with 4% paraformaldehyde for 10 min at room temperature, then washed three times with PBS and incubated overnight with primary antibodies followed by incubation with fluorophore-conjugated secondary antibodies. Images were captured using a Zeiss fluorescence microscope and analyzed using ImageJ.

TUNEL Assays

Cell death was analyzed using a TUNEL assay kit (Thermo Fisher Scientific) according to the kit manual. For pancreatic sections, at least 10 fields at ×630 magnification were counted for each specimen. For INS-1 cells, at least 50 cells per dish were counted at ×200 magnification.

CUT&RUN Analysis

To analyze chromatin association events of SIRT6 in INS-1 cells, we prepared cells under normal culture conditions and processed cells for CUT&RUN analysis using a kit (cat. no. 14-1048) from EpiCypher, Inc., using either control immunoglobulin G or antibodies for SIRT6, H3K9ac, and H3K4me3. DNA libraries were prepared using a kit (cat. no. 14-1001) from EpiCypher, Inc. Sequencing was performed on the Illumina NextSeq 2000 system. High-quality reads were aligned to the rat reference genome rn6 using bowtie2. Peaks were called using the SEACR algorithm.

Statistical Analysis

All statistical data are expressed as mean ± SEM. Statistical analysis was performed using Prism 10 software from GraphPad. Comparisons between two groups were performed using the two-tailed unpaired Student t test, and comparisons for more than two groups were performed using one- or two-way ANOVA followed by the Tukey post hoc test.

Data and Resource Availability

The RNA sequencing (RNA-seq) data set was deposited at the National Center for Biotechnology Information Gene Expression Omnibus and is accessible through GSE287696. CUT&RUN data are available on request. Requests for resources and reagents used in this study should be directed to the corresponding author (X.C.D.).

Results

Sirt6 Gene Expression Is Decreased in Pancreatic β-Cells During Aging

To investigate the effect of aging on Sirt6 gene expression in pancreatic β-cells, we performed immunofluorescence microscopy of pancreatic sections from 6-, 10-, and 16-month-old WT mice. First, we confirmed pancreatic β-cell aging using the common senescence marker P16 (cyclin-dependent kinase inhibitor 2A [CDKN2A]). Because the antibody detected both P16 and P15 (CDKN2B), P15/16⁺ β-cell percentages significantly increased from age 6 months to ages 10 and 16 months (Fig. 1A and B). Immunofluorescence staining of SIRT6 revealed a significant age-dependent decrease in older pancreatic β-cells (Fig. 1C and D). Similar changes were observed in human β-cells in older human participants without diabetes (Fig. 1E–J). These data suggest that SIRT6 may play a role in the regulation of β-cell aging.

Figure 1.

SIRT6 is decreased and senescence is increased in older β-cells. A and B: Immunofluorescence staining and quantification of insulin (INS) and P15/16 in pancreatic sections from male mice at 6 (6M), 10 (10M), and 16 months of age (16M; n = 5). C and D: Immunofluorescence staining and quantification of INS and SIRT6 staining in pancreatic sections from male mice at 6M, 10M, and 16M (n = 5). E–G: Immunofluorescence staining and quantification of INS and P15/16 in younger and older male human pancreatic sections (n = 3). H–J: Immunofluorescence staining and quantification of INS and SIRT6 in younger and older male human pancreatic sections (n = 3). Scale bar = 15 μm. Data are presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. AU, arbitrary unit.

SIRT6 Overexpression Improves β-Cell Mass and Function

To further explore the role of SIRT6 in β-cell aging, we generated TgSIRT6 mice. Immunofluorescence staining using an HA tag antibody confirmed that the SIRT6 transgene was specifically expressed in β-cells of TgSIRT6 mice (Fig. 2A and Supplementary Fig. 1A and B). It was noted that SIRT6 protein levels were also decreased in β-cells of 16-month-old TgSIRT6 mice compared with 10-month-old TgSIRT6 mice (Supplementary Fig. 2A–C). Next, we analyzed pancreatic islets in control and TgSIRT6 mice at ages 10 and 16 months. Total islet areas were significantly increased in TgSIRT6 mice compared with control mice at both 10 and 16 months (Fig. 2B and C). β-cell mass in TgSIRT6 mice trended up at 10 months of age and significantly increased at 16 months of age compared with control mice (Fig. 2D). Glucose tolerance testing was performed in control and TgSIRT6 mice at 10 and 16 months of age. TgSIRT6 mice showed better glucose tolerance than control mice, with a significant decrease in area under the curve (Fig. 2E and F). To further investigate the effect of SIRT6 overexpression on insulin secretion, we performed GSIS in 16-month-old mice. Serum insulin levels were significantly higher in TgSIRT6 mice than in control mice 15 min after glucose load (Fig. 2G). These data suggest that SIRT6 overexpression can reverse the decline in β-cell mass, insulin secretion, and glucose tolerance.

Figure 2.

Pancreatic β-cell–specific SIRT6 overexpression improves β-cell mass, insulin (INS) secretion, and glucose tolerance in mice during aging. A: Validation of SIRT6 transgene (with HA tag) expression by immunofluorescence staining of male mouse pancreatic sections. Scale bar = 15 μm. B: Representative histological images of pancreatic sections from control and TgSIRT6 male mice at 10 (10M) and 16 months of age (16M; n = 4–6). Scale bar = 400 μm. C: Quantification of pancreatic islet areas in control and TgSIRT6 male mice at 10M and 16M. D: β-Cell mass analysis in control and TgSIRT6 male and female mice at 10M and 16M (10M control: n = 2 males and 2 females; 10M TgSIRT6: n = 3 males and 1 female; 16M control: n = 3 males and 1 female; 16M TgSIRT6: n = 3 males and 1 female). E and F: Glucose tolerance tests (GTTs) were performed in control and TgSIRT6 male mice at 10M and 16M (n = 9–14). Blood glucose levels are shown as a function of time (left panels), and area under the curve (AUC) values are shown on the right. G: GSIS was performed in 16M male mice (n = 4–6). Serum INS levels were measured at baseline and 15 min after injection of a bolus of glucose. Data are presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. ns, not significant.

SIRT6 Overexpression Reduces DNA Damage and Cell Death in β-Cells During Aging

To investigate how SIRT6 overexpression protects β-cells during aging, we performed immunofluorescence staining of markers of senescence, DNA damage, cell proliferation, and cell death in control and TgSIRT6 mouse pancreatic sections. P15/16 trended lower at 10 months of age but significantly decreased in β-cells of TgSIRT6 mice at 16 months of age (Supplementary Fig. 3A and B). Similar changes in another senescence marker, P21 (CDKN1A), were also observed in β-cells of TgSIRT6 mice (Supplementary Fig. 4A and B). 53BP1, a DNA damage marker, was significantly elevated in control β-cells in 16-month-old mice compared with in 10-month-old mice. SIRT6 overexpression reduced 53BP1 in β-cells by ∼50% (Fig. 3A and B). Ki67, a proliferation marker, was remarkably increased in β-cells of TgSIRT6 mice compared with control mice at both 10 and 16 months of age (Fig. 3C and D). Next, we analyzed cell death markers. TUNEL staining showed an increase in cell death in β-cells of control mice but a significant decrease in TgSIRT6 mice at both 10 and 16 months of age (Fig. 4A and E). Cleaved caspase 3, an apoptosis marker, was also significantly decreased in β-cells of TgSIRT6 mice compared with control mice at both 10 and 16 months of age (Fig. 4B and F). Gasdermin D (GSDMD), a marker of pyroptosis, was significantly decreased in β-cells of TgSIRT6 mice compared with control mice at both 10 and 16 months of age (Fig. 4C and G). Costaining of GSDMD and SIRT6 also revealed a general inverse correlation between GSDMD and SIRT6 protein levels but not necessarily in individual β-cells, suggesting both cell-autonomous and nonautonomous effects (Supplementary Fig. 5A–D). Phosphorylated mixed-lineage kinase domain-like pseudokinase, a marker of necroptosis, was also remarkably decreased in β-cells of TgSIRT6 mice compared with control mice at both 10 and 16 months of age (Fig. 4D and H). These data suggest that SIRT6 overexpression can reduce DNA damage, senescence, and multiple forms of cell death and increase proliferation in β-cells.

Figure 3.

SIRT6 overexpression in pancreatic β-cells reduces DNA damage and increases cell proliferation. A: Immunofluorescence staining of insulin (INS; green) and 53BP1 (red) in pancreatic sections from control and TgSIRT6 male mice at 10 (10M) and 16 months of age (16M; n = 5). B: Quantification of the percentages of 53BP1⁺ β-cells in panel A. C: Immunofluorescence staining of INS (green) and Ki67 (red) in pancreatic sections from control and TgSIRT6 male mice at 10M and 16M (n = 5). D: Quantification of the percentages of Ki67⁺ β-cells in panel C. Scale bar = 15 μm. Data are presented as mean ± SEM. **P < 0.01, ***P < 0.001 ****P < 0.0001. ns, not significant.

Figure 4.

SIRT6 overexpression reduces apoptosis, necroptosis, and pyroptosis in mouse pancreatic β-cells during aging. A: Immunofluorescence staining of insulin (INS; red) and TUNEL (green) in pancreatic sections from control and TgSIRT6 male mice at 10 (10M) and 16 months of age (16M; n = 5). B: Immunofluorescence staining of INS (green) and cleaved caspase 3 (CASP3; red), an apoptosis marker, in male mouse pancreatic sections (n = 5). C: Immunofluorescence staining of INS (green) and GSDMD (red), a pyroptosis marker, in male mouse pancreatic sections (n = 5). D: Immunofluorescence staining of INS (green) and phosphorylated mixed-lineage kinase domain-like pseudokinase (p-MLKL; red), a necroptosis marker, in male mouse pancreatic sections (n = 5). E–H: Quantification of TUNEL⁺, CASP3⁺, GSDMD⁺, and p-MLKL⁺ β-cells in control and TgSIRT6 mouse pancreatic sections, respectively. Scale bar = 15 μm. Data are presented as mean ± SEM. *P < 0.05, **P < 0.01, ****P < 0.0001.

To further investigate transcript changes at transcriptomic levels, we performed RNA-seq analysis of primary islets from 10- and 16-month-old control and TgSIRT6 mice. Among differentially expressed genes (DEGs) between 10- and 16-month-old control mouse islets, 1,979 and 825 genes were upregulated and downregulated, respectively (Supplementary Fig. 6A). In the upregulated DEGs, inflammation-, T1D-, and cell death–related biological processes were highly overrepresented (Supplementary Fig. 5B). In the downregulated DEGs, pancreatic β-cell proliferation–, hormone secretion–, and energy homeostasis–related processes were significantly enriched (Supplementary Fig. 5C). At 10 months of age, 642 genes were upregulated and 599 genes were downregulated in TgSIRT6 mouse islets compared with control islets (Supplementary Fig. 7A). The pathways of pancreatic secretion, oxidative phosphorylation, and biological processes such as mitochondrial complexes were enriched in upregulated genes in the 10-month-old TgSIRT6 islets (Supplementary Fig. 7B and C). At the age of 16 months, more genes tended to be downregulated in TgSIRT6 mouse islets compared with age-matched control islets (Supplementary Fig. 8A), including genes involved in oxidative stress, calcium homeostasis, and cell death, whereas genes associated with vesicle processing and mitochondrial complex I assembly tended to be upregulated (Supplementary Fig. 8B). Kyoto Encyclopedia of Genes and Genomes and Hallmark pathway analyses also revealed that protein export and autophagy pathways were positively regulated in addition to pancreatic β-cell hallmark genes, whereas inflammation, cellular senescence, and apoptosis hallmark genes were negatively regulated in the 16-month-old TgSIRT6 mouse islets (Supplementary Fig. 9A and B). Intriguingly, the number of overlapping genes between TgSIRT6 and control mouse islets at both 10 and 16 months of age was relatively small, at only 18 upregulated and 41 downregulated genes, respectively (Supplementary Fig. 10A and B and Supplementary Table 3).

Interestingly, principal component analysis revealed that 16-month-old TgSIRT6 islet transcriptomes shifted toward 10-month-old control islet transcriptomes (Supplementary Fig. 11A). We next incorporated the comparisons between control and TgSIRT6 islets at both time points (Supplementary Fig. 11B–E). Most DEGs (1,883) in the 16-month-old TgSIRT6 islets behaved more like their 10-month-old control islet counterparts (Fig. 5A and Supplementary Fig. 11B). A majority (1,265) of upregulated DEGs in the 16-month-old control islets, relative to 10-month-old control islets, were downregulated in the 16-month-old TgSIRT6 islets, relative to 16-month-old control islets (Fig. 5B). Among those common DEGs, biological processes and pathways involved in inflammation, T1D, oxidative stress, cell senescence, and cell death were highly represented (Fig. 5C). For example, genes involved in apoptosis, pyroptosis, calcium regulation, necroptosis, ferroptosis, senescence, and DNA damage and repair in the 16-month-old TgSIRT6 islets had similar expression profiles in the 10-month-old control islets but not in the 16-month-old control islets (Fig. 5D–F, Supplementary Fig. 12A–C, and Supplementary Fig. 13). These data suggest that SIRT6 overexpression slows down β-cell aging.

Figure 5.

Transcriptomic analysis of differential genes and pathways in control and TgSIRT6 mouse islets. A: Transcriptomic shifts in male mouse islets by age and genotype from 10 (10M) to 16 months of age (16M). B: Venn diagram representation of shared genes between upregulated DEGs in 16M control (n = 3) vs. 10M control (n = 2) mouse islets and downregulated DEGs in 16M transgenic (TG; n = 3) vs. 16M control (n = 3) mouse islets. C: Representative biological processes in the common DEGs in panel B. D–F: Heat map illustrations of expression of genes in apoptosis, pyroptosis, and calcium regulation in 10M control, 16M control, and 16M TgSIRT6 mouse islets (n = 2–3). FC, fold change; FDR, false discovery rate; GO, Gene Ontology.

SIRT6 Negatively Affects Rem2 Gene Expression in β-Cells

Because our RNA-seq data revealed a role of SIRT6 in the regulation of calcium homeostasis in β-cells, we next investigated a potential role of the Rem2 gene, a known inhibitor of voltage-activated Ca²⁺ channels, in β-cell aging. Although Rem2 mRNA levels were not significantly different between control and TgSIRT6 mouse islets at 16 months of age (Fig. 5F), REM2 protein levels were significantly reduced in β-cells of TgSIRT6 mice compared with control mice at both 10 and 16 months of age (Fig. 6A and B). Additionally, REM2 protein levels were also significantly elevated in β-cells of older human participants without diabetes compared with their younger counterparts (Supplementary Fig. 14A–C). To further investigate the regulation of the Rem2 gene by SIRT6, we transfected INS-1 cells with either a control or Sirt6 sgRNA with CRISPR/Cas9. Sirt6 knockdown significantly increased Rem2 mRNA and protein levels (Fig. 6C–E). To further confirm whether the catalytic activity of SIRT6 is required, we transfected INS-1 cells with either WT SIRT6 or a catalytically inactive SIRT6-mutated (SIRT6-H133Y) plasmid. Indeed, WT but not mutated SIRT6 suppressed Rem2 gene expression at both mRNA and protein levels (Fig. 6F–H). To analyze whether SIRT6 is associated with the Rem2 gene promoter region at the chromatin level, we conducted CUT&RUN analysis using three different SIRT6 antibodies. Our data showed that at least two SIRT6 antibodies detected an SIRT6 association with the Rem2 gene promoter region as well as histone marker H3K9ac (a substrate of the SIRT6 enzyme) (Supplementary Fig. 15), suggesting a potential role of SIRT6 in the regulation of the Rem2 chromatin. Approximately 30% of the chromatin association events in the 2,000-bp gene promoters were shared by SIRT6 and H3K9ac data sets (Supplementary Fig. 16A). Among the shared genes, enriched pathways included apoptosis, cell division, DNA damage, insulin secretion, and calcium transport (Supplementary Fig. 16B).

Figure 6.

SIRT6 negatively affects Rem2 gene expression in pancreatic β-cells. A: Immunofluorescence staining of insulin (INS) and REM2 in pancreatic sections from 10-(10M) and 16-month-old (16M) control and TgSIRT6 male mice (n = 5). Scale bar = 15 μm. B: Quantification of percentages of REM2⁺ INS⁺ β-cells in both control and TgSIRT6 mice in panel A. C: Rem2 mRNA analysis in INS-1 cells transfected with either control sgGFP or sgSIRT6 (n = 3). D: Immunofluorescence staining of REM2 in INS-1 cells transfected with sgGFP or sgREM2, sgSIRT6, or both. Scale bar = 50 μm. E: Quantification of REM2 fluorescence intensity in panel D. F: Rem2 mRNA levels in INS-1 cells transfected with vector control, WT Sirt6, or SIRT6-H133Y (Sirt6 catalytically inactive mutation) plasmid (n = 3). G: Western blot analysis of REM2 protein levels in INS-1 cells transfected with vector control, WT Sirt6, or SIRT6-H133Y plasmid. β-Actin serves as a loading control. H: Quantification of REM2 protein levels in panel G. Data are presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. ns, not significant.

Rem2 Knockdown Partially Rescues SIRT6-Deficiency Defects in β-Cells

To further explore the functional relationship between SIRT6 and REM2, we performed single- or double-gene knockdowns in INS-1 cells in the absence or presence of palmitic acids. Sirt6 and Rem2 gene knockdowns were confirmed by immunostaining (Supplementary Fig. 17A and B and Fig. 6D). Sirt6 knockdown increased DNA damage, indicated by 53BP1 staining in INS-1 cells, and palmitic acid treatment further exacerbated this increase, whereas Rem2 knockdown partially reversed DNA damage (Supplementary Fig. 18A and C). Sirt6 knockdown markedly reduced Ki67, but Rem2 knockdown partially reversed this reduction (Supplementary Fig. 18B and D). Next, we analyzed lipid peroxidation, a trigger of ferroptosis. Sirt6 knockdown in INS-1 cells significantly elevated lipid peroxidation levels, indicated by an increase in 4-HNE staining and a decrease in glutathione peroxidase 4, a key enzyme in anti-lipid peroxidation, whereas Rem2 knockdown partially reversed lipid peroxidation (Supplementary Fig. 19A, B, H, and I). Sirt6 knockdown increased DNA damage, indicated by p-H2A.X staining, whereas Rem2 knockdown partially reduced this increase (Supplementary Fig. 19C and J). Sirt6 knockdown increased apoptosis, necroptosis, and pyroptosis in INS-1 cells, whereas Rem2 knockdown resulted in a partial rescue from those forms of cell death (Supplementary Fig. 19D–G and K–N). These data suggest that REM2 mediates some of the SIRT6 deficiency–induced defects.

Rem2 βKO Mice Have Improved β-Cell Survival and Function

To investigate the role of REM2 in β-cells in vivo, we generated Rem2 βKO mice. Immunostaining analysis confirmed that REM2 was specifically deficient in pancreatic β-cells (Fig. 7A and Supplementary Fig. 20A and B). Histological analysis showed an increase in islet area in the Rem2 βKO mice compared with control mice at both 10 and 16 months of age (Fig. 7B and C). Rem2 βKO mice had a significant increase in β-cell mass at both 10 and 16 months of age (Fig. 7D). GSIS was significantly increased 15 min post–glucose load in Rem2 βKO mice at both 10 and 16 months of age (Fig. 7E and F). Rem2 βKO mice had better glucose tolerance than control mice at both 10 and 16 months of age (Fig. 7G and H). Immunostaining analysis revealed that Rem2 βKO mice had a significant decrease in DNA damage but an increase in proliferation in β-cells (Fig. 8A–D). Cell senescence shown by P15/16 and P21 was significantly reduced in β-cells of Rem2 βKO mice at 16 months of age (Supplementary Fig. 21A and B and Supplementary Fig. 22A and B). Cell death analysis revealed that Rem2 gene knockout significantly reduced apoptosis, pyroptosis, and necroptosis (Fig. 8E–L). These data suggest that REM2 is a negative regulator of β-cell survival and function.

Figure 7.

Rem2 βKO improves β-cell mass and glucose tolerance. A: Confirmation of Rem2 βKO in male and female mice at 10 months of age (10M) by immunostaining analysis. Scale bar = 15 μm. B: Representative histological images of pancreatic sections from 10M and 16-month-old (16M) control and Rem2 βKO male mice (n = 5). Scale bar = 400 μm. C: Quantification of pancreatic islet areas in control and Rem2 βKO male mice. D: β-Cell mass analysis in control and Rem2 βKO male mice (n = 4). E and F: GSIS in 10M and 16M control and Rem2 βKO male mice (n = 4). G: Glucose tolerance tests (GTTs) and area under the curve (AUC) analysis in 10M control (n = 7) and Rem2 βKO (n = 9) male mice. H: GTTs and AUC analysis in 16M control and Rem2 βKO (n = 4) male mice. Data are presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. INS, insulin; ns, not significant.

Figure 8.

Deletion of Rem2 in pancreatic β-cells reduces age-associated DNA damage and cell death. A and B: Immunofluorescence and quantification analysis of 53BP1 in pancreatic sections from control and Rem2 βKO male and female mice (n = 5 [3 males and 2 females]). C and D: Immunofluorescence and quantification analysis of Ki67 in control and Rem2 βKO male and female mouse pancreatic sections (n = 5 [3 males and 2 females]). E: TUNEL assays in control and Rem2 βKO male and female mouse pancreatic sections (n = 5 [3 males and 2 females]). F: Cleaved caspase 3 (CASP3) staining in control and Rem2 βKO male and female mouse pancreatic sections (n = 5 [3 males and 2 females]). G: GSDMD staining in control and Rem2 βKO male and female mouse pancreatic sections (n = 5 [3 males and 2 females]). H: Phosphorylated mixed-lineage kinase domain-like pseudokinase (p-MLKL) staining in control and Rem2 βKO male and female mouse pancreatic sections (n = 5 [3 males and 2 females]). I–L: Quantification of 53BP1⁺, Ki67⁺, TUNEL⁺, and CASP3⁺ β-cells in control and Rem2 βKO mouse pancreatic sections in panels E–H. Scale bar = 15 μm. Data are presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. INS, insulin; M, month.

Discussion

Pancreatic β-cells are key players in glucose homeostasis and diabetes pathogenesis (23). With aging, both pancreatic β-cell proliferation and insulin secretory function gradually decline (3,24,25). This work demonstrates a significant role of SIRT6 in the protection against age-related β-cell decline.

β-cell SIRT6 deficiency leads to an insulin secretion defect resulting from mitochondrial dysfunction, oxidative stress, and other factors (13–15). SIRT6-deficient β-cells are more susceptible to palmitic acid–induced apoptosis than WT cells in MIN6 insulinoma cells (6). In this study, we demonstrate that SIRT6 overexpression significantly reduces the apoptosis, necroptosis, and pyroptosis associated with aging in vivo or lipotoxicity in vitro. Although our previous work showed that SIRT6 is decreased in older mouse islets and palmitate-treated MIN6 cells, this work firmly demonstrates the antiaging function of SIRT6 in vivo in reducing DNA damage, lipid peroxidation, apoptosis, necroptosis, and pyroptosis. The antiaging effect of SIRT6 in β-cells can be further investigated in Western diet–treated TgSIRT6 mice in future studies. Because SIRT6 plays a critical role in DNA repair and genome integrity (7,26), age-associated SIRT6 decline may contribute to the elevated DNA damage levels in older β-cells.

Calcium flux regulation plays a critical role in insulin secretion from pancreatic β-cells (4). SIRT6 has been implicated in the regulation of calcium homeostasis (14,27). It has been reported that SIRT6 regulates calcium signaling through modulation of intracellular levels of ADP-ribose that can activate transient receptor potential melastatin 2 (TRPM2) in pancreatic cancer cells (27). TRPM2, a nonspecific Ca²⁺ channel, has been shown to mediate glucose- and incretin-induced insulin secretion (28,29). Because SIRT6-deficient β-cells exhibit mitochondrial defects and insufficient ATP production (14), it is likely that the indirect effect of SIRT6 deficiency on Ca²⁺ flux also plays a significant role. In this study, we observed a connection between SIRT6 and REM2, a potent endogenous inhibitor of high-voltage–activated calcium channels. The Rem2 gene is negatively regulated in INS-1 cells by SIRT6 in a catalytic activity–dependent manner. Significantly, knockdown of Rem2 partially rescued SIRT6 deficiency– and palmitic acid–induced DNA damage and multiple forms of cell death in INS-1 cells. Additionally, β-cell–specific Rem2 gene knockout in mice led to a significant increase in β-cell mass and a decrease in β-cell apoptosis, necroptosis, and pyroptosis and improved glucose tolerance. The relationship between SIRT6 and REM2 in β-cells can be further corroborated by the characterization of β-cell survival and function in Sirt6 and Rem2 β-cell–specific double-knockout mice and Sirt6 and Rem2 β-cell–specific double-transgenic mice in future studies. In addition to Ca²⁺ channels, REM2 has been shown to inhibit CaMKII in neurons (17). Because CaMKII positively regulates insulin secretion in pancreatic β-cells (30), it is plausible that SIRT6 also regulates insulin secretion through suppression of Rem2 gene expression and subsequent activation of Ca²⁺ channels and CaMKII.

In summary, our data suggest that SIRT6 is an important factor that helps maintain pancreatic β-cell survival and function during aging. Age-associated SIRT6 insufficiency in β-cells increases the risk of DNA damage and cell death. Therefore, improving SIRT6 function may ameliorate pathological changes, such as glucose intolerance or diabetes, associated with β-cell aging.

This article contains supplementary material online at https://doi.org/10.2337/figshare.29828771.