Lamin C Counteracts Glucose Intolerance in Aging, Obesity, and Diabetes Through β-Cell Adaptation

Lamin A, lamin C, and progerin proteins, produced by alternative RNA processing of the Lmna gene, are intermediate filaments that bridge peripheral heterochromatin to the nuclear envelope (1), thus participating in the maintenance of chromatin architecture (2,3) and in transcriptional regulation (4). Mutations in the Lmna gene result in laminopathies, including progeria, lead to metabolic abnormalities such as lipodystrophy, hypertriglyceridemia, insulin resistance, and metabolic syndrome, thus suggesting an important role for A-type lamins in the regulation of glucose and lipid metabolism (5). Comparing knockin mice expressing all three lamin isoforms (lamin A, lamin C, and progerin) or lamin C only–expressing mice (Lmna^LCS/LCS), we previously reported antagonistic functions of Lmna isoforms in energy expenditure and life span, pointing out a key function for this gene in the control of energy homeostasis (6). Despite showing increased body weight, Lmna^LCS/LCS have an unexpectedly increased life span, suggesting a role for lamin C in improving metabolic health during aging.

Pancreatic β-cells play a critical role in maintaining glucose homeostasis through the secretion of insulin, and their dysfunction results in diabetes. Whereas type 1 diabetes (T1D) is an autoimmune disease that results from the destruction of β-cells, type 2 diabetes (T2D) is due to energy imbalance and is characterized by impaired peripheral insulin sensitivity followed by defects in insulin secretion. The common hallmark of T1D and T2D is an inadequate supply of functional insulin-producing β-cells, leading to the dysregulation of glucose metabolism and associated metabolic complications (7). The restoration of normal β-cell mass and function has therefore become of great interest for the generation of novel antidiabetic drugs. In this context, β-cell regeneration has been extensively explored in mice and humans. However, the mechanisms involved in the functional plasticity required for the regeneration of β-cell mass remain poorly understood (7,8).

Although recent reports highlighted that Lmna isoforms are involved in the structure and function of the exocrine pancreas (9), their contribution to the endocrine pancreas and, more particularly, in insulin-producing β-cells remains to be investigated. We hypothesized that lamin C might contribute to maintaining adequate glucose homeostasis during aging through a role in pancreatic β-cell plasticity or regeneration. Using a comprehensive analysis of the metabolic phenotype of Lmna^LCS/LCS mice using different models related to β-cell failure (i.e., aging, streptozotocin [STZ]–induced diabetes, and obesity-induced diabetes), we demonstrate that exclusive lamin C expression favors β-cell mass, insulin production, and insulin secretion in conditions in which dual expression of lamin A and lamin C from the Lmna gene is accompanied by abnormal glucose homeostasis. This adaptive effect on β-cell mass and function is associated with an increase in mitochondrial activity and of global translation rates, providing a molecular link between lamin C and β-cell fitness during aging. Using a model of T2D (diet-induced obesity [DIO]), we report that Lmna^LCS/LCS mice enhance insulin secretion to adapt to DIO and to maintain normoglycemia. Finally, in STZ-induced diabetic mice, β-cell failure was rescued in Lmna^LCS/LCS through increased regeneration of β-cells. Altogether, our findings demonstrate that the exclusive expression of lamin C improves β-cell function and mass during metabolic stress and suggest that the manipulation of Lmna isoforms can be targeted to counteract diabetes.

Lmna^LCS/LCS knockin mice were bred onto the C57BL/6J background as described (10). All animal procedures were conducted in accordance with the French national animal care guidelines. Mice were maintained in pathogen-free conditions in our animal facility (E34–172–16). Mice were housed under a 12-h light/dark cycle and given an ad libitum regular chow diet (CD) (A04; SAFE). All experiments were conducted by authorized personnel (agreements MDT34–433 and CC-I-34UnivMontp-f1–12) and approved by the Ethics Committee of the Languedoc-Roussillon region Montpellier (CEEA-LR-12003; Occitanie, France). Male mice were used for in vivo experiments, and isolated islets were prepared from male or female mice.

Pancreas sections (4 µm) were deparaffinized, rehydrated, stained with hematoxylin and eosin, and visualized with a NanoZoomer. Islet diameters were measured using NDP.view software.

For immunohistochemistry, rehydrated sections were incubated in citrate buffer at 95°C for 30 min. After blocking, sections were incubated in primary antibodies (Supplementary Table 2) overnight at 4°C and detected using appropriate secondary antibodies. Nuclei were counterstained with Hoechst 33342.

For the intraperitoneal glucose tolerance test (IPGTT), 18-h–fasted mice were injected intraperitoneally with glucose (1.7 g/kg), and glycemia was measured using a blood glucometer (Roche). Insulin levels were quantified using Mouse Ultrasensitive Insulin ELISAs (ALPCO). To induce T2D, 12-week-old mice had unrestricted acces to a high-fat diet (HFD) (Research Diets, Inc.) for 25 weeks. To induce acute β-cell loss, 13-week-old mice were injected intraperitoneally with a single high dose (200 mg/kg) of STZ (Sigma-Aldrich) dissolved in 0.1 mol/L sodium citrate buffer (pH 4.4).

Mouse islets were isolated by type V collagenase digestion (Sigma-Aldrich) of pancreata. After separation in a density-gradient medium (Histopaque-1119; Sigma-Aldrich), islets were handpicked. They were then cultured for 18–20 h at 37°C in a 95% air/5% CO₂ atmosphere in RPMI 1640 (Thermo Fisher Scientific) containing 10% FCS and 100 µg/mL penicillin-streptomycin.

To evaluate insulin secretion, three replicates of five islets were preincubated for 1 h at 37°C in Krebs-Ringer bicarbonate buffer. After low (3 mmol/L) or high (20 mmol/L) glucose stimulation, secreted insulin and total insulin cell content were measured. Data are expressed as a ratio of secreted insulin/total insulin content.

Pancreata were lysed in 2 mL acid/ethanol lysis buffer and incubated overnight at 4°C. After neutralization with 10% of 1 mol/L Na₂CO₃, total insulin and glucagon content were measured by ELISA (Mouse Ultrasensitive Insulin ELISA, ALPCO; Glucagon ELISA, Mercodia) (10 μL).

Total RNA was extracted from isolated islets using TRIzol (Life Technologies). Samples were quantified with a NanoDrop ND-1000; 1.5 μg RNA was then converted to cDNA with the Maxima First Strand cDNA Synthesis Kit (#K1642; Thermo Fisher Scientific). Gene expression was normalized to cyclophilin and TBP.

For mtDNA quantification, islets were digested using proteinase K. Quantitative PCR (qPCR) analysis was performed with primers specific for mitochondrial genes (mtND5 and mt16s) and normalized to nuclear genes (LN11 and HK2).

qPCR was performed with a LightCycler 480 and LightCycler 480 SYBR Green I Master (Roche). Primer sequences are provided in Supplementary Table 1.

After isolation, 150 islets were immersed in a solution of 2.5% glutaraldehyde. After dehydration, samples were embedded in EMbed812. The 70-nm sections were counterstained with uranyl acetate and observed on a Hitachi-7100 microscope.

Oxygen consumption rate (OCR) was measured in 96-well plates using an Seahorse XFe96 (Agilent Technologies). Briefly, 25 islets per well were plated in poly-d-lysine–coated wells in XF Assay Medium containing 3 mmol/L glucose, 1 mmol/L sodium pyruvate, and 1 mmol/L glutamine (pH 7.4) for 1 h at 37°C without CO₂. OCRs were monitored in response to glucose (20 mmol/L), oligomycin (2 μmol/L), FCCP (1 μmol/L), and rotenone (0.5 μmol/L). OCR measurements were normalized by total DNA level using a Qubit dsDNA HS assay kit (Life Technologies).

Seventy isolated islets were incubated in Met/Cys-depleted RPMI medium (Sigma-Aldrich) for 30 min and then supplemented with 50 μCi ³⁵S-Met/Cys EasyTag Express (PerkinElmer) for 30 min at 37°C. Islets were lysed in RIPA buffer, and 10 μg of total protein was size-fractionated on 4–12% gradient SDS-PAGE. The gel was dried and quantified using a Phosphoimager (GE Healthcare).

Islets dissociated with 0.05% trypsin and 0.5 mmol/L EDTA were fixed with 1% paraformaldehyde and permeabilized with 0.05% Triton X-100. After blocking, cells were incubated anti-insulin antibody, washed, reacted with an FITC-conjugated antibody, and analyzed by FACSCanto II (BD Biosciences).

Total RNA was extracted using TRIzol and quantified on Qubit (Life Technologies). RNA quality was verified using RNA Pico Chips on a 2100 Bioanalyzer (Agilent Technologies). Total DNA was purified using MasterPure Complete DNA (Illumina). Libraries for RNA sequencing (RNA-seq) and reduced representation bisulfite sequencing (RRBS) were prepared and sequenced by the MGX sequencing platform (Montpellier, France).

For Assay for Transposable-Accessible Chromatin with high-throughput sequencing (ATAC-seq) analyses, islets were cryopreserved in 50% FCS, 40% RPMI, and 10% DMSO and processed by Active Motif.

All data are deposited in the National Center for Biotechnology Information’s Gene Expression Omnibus (GEO) under accession number GSE122361.

Results were expressed as the mean ± SEM. Statistical comparisons were made using the Student t test. Statistically significant differences are indicated as *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

A detailed description of all procedures and bioinformatic analyses can be found in the Supplementary Data.

The resources generated during and/or analyzed during the current study are available from the corresponding authors on reasonable request. The genome-wide data generated and/or analyzed during the current study are deposited in National Center for Biotechnology Information’s GEO under accession number GSE122361.

To evaluate whether lamin C regulates β-cell function, we evaluated glucose homeostasis in Lmna^LCS/LCS mice (6). Initially designed to generate a mouse model of progeria, the c.1827C>T;p.G609G point mutation was introduced into a floxed neo cassette (Supplementary Fig. 1A). This cassette was inserted in intron-10 of the Lmna gene, downstream of the polyadenylation site of lamin C, thus preventing the formation of prelamin A transcripts by blocking lamin A–specific splicing (10). The conditional mutant allele (Lmna^LCS) resulted in the exclusive expression of lamin C. This was confirmed by comparing its levels in brain, liver, pancreas, and muscle from Lmna^+/+ (wild-type [WT]) and Lmna^LCS/LCS mice (Supplementary Fig. 1B). We observed increased expression of lamin C in Lmna^LCS/LCS mice compared with WT in most tissues except in the brain, where the lamin C/lamin A ratio is already high even in WT mice, as previously shown (11).

With the experimental Lmna^LCS/LCS model established, we then investigated the metabolic phenotype of these mice. The Lmna^LCS/LCS mice gained more weight than control mice as they aged (Fig. 1A). This increase in body weight is mainly due to an increase in white adipose tissue without any further change in body composition, as shown in our previous study (6). Body weight and fasting glucose levels in Lmna^+/+ and Lmna^LCS/LCS mice were similar until 25 weeks of age (Fig. 1A and B). Interestingly, the significant increase in body weight was observed from 30 weeks of age in Lmna^LCS/LCS mice, but was even more pronounced at 75 weeks of age. The following aging experiments are thus conducted by comparing “young” (25 weeks old) to “old” animals (75 weeks old). We observed a significant increase in fasting glycemia in control mice as they aged (Fig. 1B). Interestingly, Lmna^LCS/LCS mice remained normoglycemic, although an increased body weight is typically associated with hyperglycemia. These findings prompted us to analyze glucose homeostasis in aged Lmna^LCS/LCS mice. Glucose clearance was similar in young animals of both genotypes (Fig. 1C, left panel), whereas old Lmna^LCS/LCS mice showed improved glucose tolerance compared with age-matched Lmna^+/+ mice (Fig. 1C, right panel). As expected, no differences in basal and glucose-stimulated insulin secretion (GSIS) were observed in young animals from either genotype (Fig. 1D, left panel). Basal and stimulated insulin secretion was, however, increased in old Lmna^LCS/LCS mice compared with old WT mice (Fig. 1D, right panel). Altogether, these data suggest that although Lmna^LCS/LCS mice become obese as they age, their glucose clearance and insulin secretion are improved, suggesting that lamin C may modulate the function of pancreatic islets of Langerhans during aging.

Our observations in aged Lmna^LCS/LCS mice prompted us to study whether lamin C could contribute to pancreatic islet function by regulating pancreatic mass and/or insulin secretion. Old Lmna^LCS/LCS mice had heavier pancreata, whereas no differences in mass were observed between genotypes in young animals (Fig. 2A). Histological analyses revealed that while young Lmna^+/+ and Lmna^LCS/LCS pancreata were microscopically similar, pancreata from old Lmna^LCS/LCS mice displayed significantly larger islets of Langerhans compared with Lmna^+/+ mice (Fig. 2B). This effect was further amplified in the pancreata of “very old” (141 weeks old) Lmna^LCS/LCS mice, as evidenced by hyperplasic islets (Supplementary Fig. 1C). In young animals, measurements of islet size (Fig. 2C, top panel) and staining of insulin and glucagon-positive cells within the pancreatic islets (Fig. 2D, top panel) revealed no differences, whereas old Lmna^LCS/LCS animals displayed enlarged islets with an increase in insulin-positive cells, but no differences in the number of glucagon-positive cells (Fig. 2C and D, bottom panel). Quantification of the β/α-cell ratio confirmed an increased number of insulin-positive β-cells in aged Lmna^LCS/LCS mice (Fig. 2E), whereas no changes were observed on glucagon-positive α-cells (Supplementary Fig. 2A). In addition, when compared with young mice, old Lmna^LCS/LCS mice exhibited an increased number of islets per pancreas (Fig. 2F). Furthermore, total pancreatic insulin level was increased in LCS mice compared with control mice (Fig. 2G), whereas global glucagon level was not significantly different (Supplementary Fig. 2B). These data demonstrate that the expression of lamin C protects from aging-induced β-cell mass loss by increasing the size and the number of islets of Langerhans and the number of insulin-positive cells.

We then investigated the functionality of these abundant and large Lmna^LCS/LCS islets through GSIS experiments. Although no difference in GSIS among genotypes was observed in pancreatic islets isolated from young animals, islets derived from old Lmna^LCS/LCS mice displayed a 29.5% higher GSIS than age-matched control islets after 20 mmol/L glucose stimulation (Fig. 2H). Interestingly, no difference of insulin secretion in basal, 2.8 mmol/L glucose stimulation was observed, suggesting that the increased insulinemia observed in fasted Lmna^LCS/LCS mice (Fig. 1D) could be due to a cumulative effect of the increased number of islets and their improved insulin secretion capacity.

To confirm the direct effect of lamin C on β-cells, we next generated MIN6 pancreatic β-cell lines in which lamin A expression was abolished using CRISPR/Cas9-mediated genome editing (MIN6^LCO). Targeting exon 11 of the Lmna gene resulted in a complete depletion of lamin A expression. In the MIN6^LCO cells expressing only lamin C (Supplementary Fig. 3A), insulin secretion in response to a 20 mmol/L glucose stimulation was increased compared with the control MIN6^WT cells (Supplementary Fig. 3B), indicating that lamin C can enhance insulin secretion and improve β-cell function.

Taken together, these results showed that old Lmna^LCS/LCS mice have increased β-cell number and improved insulin secretion in response to glucose both in vitro and in vivo. These adaptive mechanisms may account for the normoglycemia observed in old Lmna^LCS/LCS mice despite their obesity.

To counteract obesity, β-cells need to increase their secretory capacity by progressive expansion of their mass (12). To evaluate how Lmna^LCS/LCS β-cell function adapts to metabolic challenges, we investigated the capacity of Lmna^LCS/LCS-derived islets to counteract hyperglycemia observed during obesity-induced T2D in young mice. Challenging 12-week-old Lmna^+/+ and Lmna^LCS/LCS mice with a 60% kcal HFD for 25 weeks resulted in increased body weight (Fig. 3A). This body weight gain was significantly higher in Lmna^LCS/LCS mice and was associated with increased food intake when compared with the control mice (Supplementary Fig. 4). No differences were observed among genotypes in the consumption of a regular CD. Monitoring of fasting glycemia over the course of HFD revealed that during the first 10 weeks, Lmna^LCS/LCS mice were hyperglycemic compared with control mice, probably due to their higher body weight (Fig. 3B). However, at week 15 under HFD, a significant improvement in fasting glycemia was observed. At week 25, Lmna^LCS/LCS mice demonstrated a further improvement in glucose tolerance (Fig. 3C) and showed increased insulin secretion in response to in vivo glucose stimulation (Fig. 3D). GSIS experiments using islets isolated from HFD-fed mice for 25 weeks demonstrated that while Lmna^+/+-derived islets failed to secrete insulin in response to glucose, islets from HFD-fed Lmna^LCS/LCS mice efficiently secreted insulin in response to glucose (Fig. 3E).

These results demonstrate that Lmna^LCS/LCS mice efficiently counteract HFD-induced hyperglycemia by increasing insulin secretion. Taken together with our aging studies, these results suggest that a lamin C–specific adaptive mechanism is engaged by insulin-producing cells to maintain glucose responsiveness in order to counteract the effects of aging and obesity-induced T2D.

To further explore the role of lamin C on β-cell number and function, we evaluated the capacity of Lmna^LCS/LCS mice to regenerate β-cells in an acute β-cell loss context. Thirteen-week-old Lmna^+/+ and Lmna^LCS/LCS animals were treated with a single dose of STZ (200 mg/kg) to mimic diabetes, and the regeneration of β-cells was measured 18 days after STZ injection. Three days after STZ treatment, Lmna^+/+ and Lmna^LCS/LCS mice became diabetic, with no significant difference in fasting glucose level (Fig. 4A). From day 4, while Lmna^+/+ mice remained hyperglycemic, blood glucose levels in Lmna^LCS/LCS mice rapidly decreased, with a complete normalization of glycemia 18 days after STZ injection. Moreover, we observed ∼15% of body weight loss during the first 4 days following STZ treatment for both genotypes. At 18 days after STZ injection, Lmna^+/+ mice remained as lean as they were before the treatment, while Lmna^LCS/LCS mice recovered their initial body weight (Supplementary Fig. 5A). Consistent with our previous observations, Lmna^LCS/LCS showed improved glucose tolerance 18 days after STZ injection (Fig. 4B), suggesting functional insulin-producing cells in STZ-treated Lmna^LCS/LCS mice.

Morphological analysis of the pancreata 18 days post–STZ injection revealed that the number of insulin-positive cells was dramatically reduced in Lmna^+/+ islets (Fig. 4C, left panel). Conversely, Lmna^LCS/LCS insulin-positive β-cells recovered efficiently, with comparable staining of insulin- and glucagon-positive cells before and after STZ treatment. Overall, 18 days after STZ injection, a 2.5-fold increase in insulin-positive cells was observed in the pancreas of Lmna^LCS/LCS mice compared with control mice (Fig. 4C, right panel).

The increased number of islets in Lmna^LCS/LCS mice led us to hypothesize that Lmna^LCS/LCS mice are either protected from STZ-induced apoptosis and/or are able to efficiently regenerate β-cells after STZ-induced damage. At 72 h after STZ treatment, β-cells were strongly depleted equally in Lmna^+/+ and Lmna^LCS/LCS mice, as shown by representative immunostaining as well as β-cell mass quantification (Fig. 4D). The susceptibility of β-cells to STZ-induced apoptosis was confirmed by immunostaining (Supplementary Fig. 5B, left panel), and quantitative assessments 48 h after STZ treatment using TUNEL assays showed no significant difference in the number of apoptotic cells between Lmna^+/+ and Lmna^LCS/LCS mice (Supplementary Fig. 5B, right panel). This finding was further confirmed in the MIN6 β-cell line treated with STZ, in which no differences in the number of apoptotic cells were observed between MIN6^WT and MIN6^LCO cells (Supplementary Fig. 5C).

We next studied β-cell regeneration after STZ treatment in Lmna^LCS/LCS mice. The presence of proliferating cells after STZ treatment was quantified using phospho-histone H3 (phospho-H3) antibody staining on pancreatic sections at day 9 post–STZ injection. Pancreata from Lmna^LCS/LCS mice had a significantly higher number of phospho-H3–positive cells compared with control mice (Fig. 4E), suggesting an increased regenerative capacity in Lmna^LCS/LCS mice. To identify the functional profile of these newly generated cells, we next performed coimmunostaining for phospho-H3 (red) and different pancreatic markers, including insulin (β-cells), lectin (ductal cells), glucagon (α-cells), and somatostatin (δ-cells). Coimmunostaining was observed for phospho-H3 and insulin as well as lectin, but not with glucagon or somatostatin, suggesting that proliferating cells observed in Lmna^LCS/LCS mice may arise from duplication of pre-existing β-cells and from ductal cells (Fig. 4F), two well-known processes for β-cell neogenesis (8).

Altogether, these results suggest that Lmna^LCS/LCS mice are able to counteract STZ-induced diabetes by regenerating β-cells.

Given the important role of lamins in chromatin organization and in the control of gene expression (4), we next evaluated the transcriptional status, DNA methylation, and chromatin accessibility of islets isolated from 75-week-old Lmna^+/+ and Lmna^LCS/LCS mice. RNA-seq–based analysis revealed that 1,507 genes (9.3% of the sequenced genes) were significantly differentially regulated among genotypes (P < 0.05), with 1,024 genes upregulated and 483 genes downregulated in Lmna^LCS/LCS mice relative to controls (Fig. 5A). Gene ontology enrichment analysis showed an enrichment of genes associated with mitochondria (15.92%) and global translation (15.59%). Most genes associated with mitochondrial function (205 out of 240 genes) and translation pathways (169 out of 235 genes) were significantly upregulated in Lmna^LCS/LCS islets compared with Lmna^+/+ islets. RRBS-based methylome analysis revealed that the observed variations in gene expression were not correlated to changes in DNA methylation (Fig. 5B). Indeed, independent of the region of the genome that was analyzed (Supplementary Fig. 6), no differentially methylated cytosines were found in 1,383,868 cytosines tested. Moreover, no differentially methylated regions were found in 872,305 tested regions (GEO accession number GSE122361). In order to evaluate whether the transcriptional dysregulation was due to a global remodelling of the chromatin compaction and/or accessibility, we performed ATAC-seq on pancreatic islets isolated from old Lmna^+/+ versus Lmna^LCS/LCS mice. ATAC-seq results revealed that no difference in chromatin accessibility profiles was observed between Lmna^+/+ and Lmna^LCS/LCS isolated islets (Fig. 5C), suggesting that the transcriptional dysregulations observed in RNA-seq analysis were not due to changes in global chromatin accessibility upon the exclusive expression of lamin C.

To go further into the molecular mechanisms by which lamins could modulate transcription programs, we analyzed the chromatin architecture by investigating the spatial nuclear localization of selected histone marks that are known to modulate gene expression (13). Immunofluorescence analyses of pancreatic sections showed an enrichment of the active enhancer mark H3K27Ac at the nuclear periphery in Lmna^LCS/LCS β-cells compared with control (Fig. 5D, left panel), whereas no difference was observed for the localization of the repressive mark H3K9Me3 (Fig. 5D, right panel). Interestingly, the distribution of the H3K27Ac epigenetic mark to the nuclear periphery is associated with a modification of the nuclear architecture. Indeed, we observed that the chromatin is relocated to the periphery of the nucleus in Lmna^LCS/LCS mice, as indicated by Hoechst staining (Fig. 5D), without changing the global chromatin accessibility, as shown by our ATAC-seq analysis.

Taken together, our transcriptomic- and epigenomic-wide studies demonstrate that the transcriptional reprogramming of genes involved in mitochondrial function and global translation cannot be assigned to either changes in DNA methylation or chromatin accessibility signatures (RRBS and ATAC-seq) but more likely to the relocalization of epigenomic marks and/or a remodeling of the nuclear architecture.

Because mitochondrial homeostasis is critical for both the maintenance of β-cell function and life span promotion (14), we next focused on the regulation of mitochondrial gene expression and function. First, we validated our RNA-seq data by quantitative RT-PCR analysis and confirmed that Lmna^LCS/LCS-derived islets expressed increased levels of key genes involved in the control of mitochondrial function, including the genes encoding electron transport chain proteins (Nduf, Atp5, and Cox protein family) (Fig. 6A). In addition, the amount of mtDNA relative to nuclear DNA was significantly higher in Lmna^LCS/LCS islets (Fig. 6B). Ultrastructural analysis of β-cells by transmission electron microscopy (Fig. 6C) together with the OCR of islets (Fig. 6D) further supported a higher number of functional mitochondria in Lmna^LCS/LCS islets. Taken together, these results suggest that both mitochondrial number and activity are increased in pancreatic islets isolated from Lmna^LCS/LCS mice.

RNA-seq data also highlighted a role for lamin C in the regulation of protein translation-related genes in islets from old Lmna^LCS/LCS mice. In particular, genes involved in ribosome synthesis (Rpl and Mrpl protein family) and in the initiation phase of eukaryotic translation (eukaryotic initiation factors) were overexpressed in Lmna^LCS/LCS islets. qPCR experiments confirmed that the expression of genes directly implicated in different steps of translation was significantly increased in Lmna^LCS/LCS islets compared with Lmna^+/+ islets (Fig. 6E). Because ribosome synthesis is tightly correlated with rRNA synthesis, we also measured 18S and 28S rRNA levels in isolated islets and found that their expression levels were 1.5-fold increased in Lmna^LCS/LCS islets (Fig. 6F). To confirm that translation rates are increased in Lmna^LCS/LCS islets, we measured ³⁵S-Met/Cys incorporation and found that global protein synthesis levels were significantly increased in Lmna^LCS/LCS islets (Fig. 6G). Because insulin is the major protein synthetized in β-cells, we measured insulin content in Lmna^+/+ and Lmna^LCS/LCS isolated islets. Because Lmna^LCS/LCS islets are larger than control islets, classical insulin content measurement may not quantitatively reflect the abundance of insulin per cell. Therefore, we performed FACS analysis to measure intracellular insulin levels in dissociated β-cells. As shown in Fig. 6H, quantification of FACS experiments revealed that insulin production was increased by ∼25% in Lmna^LCS/LCS β-cells.

Altogether, these data suggest that the increased insulin secretion observed in Lmna^LCS/LCS mice is likely due to an increase in global translation rates, which enhances insulin production. The secretion is supported by an increased number of functional mitochondria that play a crucial role in insulin exocytosis upon glucose stimulation.The observed lamin C–mediated increase in insulin production and secretion may explain the improved metabolic phenotype observed in obese, aged Lmna^LCS/LCS mice.

Loss of β-cell function and mass is involved in the development of T1D and T2D and may contribute to the metabolic complications observed during aging. Our results demonstrate for the first time that the exclusive expression of lamin C acts as a key regulator of β-cell function and regeneration, thus protecting against hyperglycemia observed in several forms of metabolic stresses, such as aging, DIO, and STZ-induced diabetes. By combining in vitro, ex vivo, and in vivo studies, we report that lamin C expression increases β-cell number during aging, counteracts β-cell failure during DIO, and promotes β-cell regeneration after STZ treatment. Most importantly, we demonstrate that lamin C modulates a specific transcriptional program that contributes to enhance the translation of insulin and promotes insulin secretion through a mitochondria-coupled mechanism.

It has been shown that β-cell turnover decreases with aging due to the onset of senescence (15). Indeed, the expression of p16^Ink4a, a key marker of senescence, limits islet proliferation and regeneration in an age-dependent manner (16). Conversely, the downregulation of p16^Ink4a in a Pten-null background reduces the age-induced decline of β-cell proliferation and enhances islet mass in aged mice (17), as observed in our aged Lmna^LCS/LCS mice. While one may speculate that a decrease in cellular senescence in our Lmna^LCS/LCS mice may underlie the observed maintenance of β-cell mass during aging, no difference in the expression of the senescence marker p16^Ink4a was observed in our Lmna^LCS/LCS mice versus controls (data not shown), suggesting that delayed senescence does not contribute to the maintenance of functional islets in our model. This hypothesis is further strengthened by the observation that β-cells that exclusively express lamin C adapt to pathological conditions unrelated to aging, such as diabetes induced in young mice.

The specific functions of lamin A versus lamin C in β-cell maintenance and function during aging and/or metabolic stress is a central question. A limitation of the Lmna^LCS/LCS model is that we cannot rule out potential contributing effects of lamin C in other organs, as well as an effect of lamin A deficiency in cells expressing lamin C only. Because the gene targeting strategy used in this study to generate the Lmna^LCS/LCS model precludes specifically expressing lamin C only in β-cells and subsequently performs lineage-tracing experiments, this question remains to be addressed. Interestingly, mice expressing lamin A only (Lmna^LAO) are healthy, with no detectable phenotypical or histopathological abnormalities, despite a complete loss of lamin C expression (18), even in 24-month-old mice. Challenging these Lmna^LAO mice with HFD or STZ treatment may help to accurately characterize their specific functions in glucose homeostasis.

To decipher the molecular mechanism by which lamin C regulates pancreatic islet functions, we analyzed the impact of lamin C exclusive expression using genome-wide DNA methylation, ATAC-seq, and high-throughput RNA-seq.

RNA-seq analysis revealed that the expression of genes involved in mitochondrial function and protein translation was altered in aged Lmna^LCS/LCS islets compared with controls. Unlike the decline of mitochondrial function observed in aging-associated diseases such as diabetes (14), we found increased mitochondrial activity in aged Lmna^LCS/LCS islets that likely contributes to the observed improvement in insulin secretion. It is interesting to note that decreased mitochondrial biogenesis was previously observed in adipose tissue from Lmna^LCS/LCS mice aged 45 weeks (6). Contrary to white adipose tissue in which mitochondria are relatively less abundant and crucial for adipocyte homeostasis in normal diet and temperature conditions, mitochondrial function in the β-cell is directly coupled to insulin biosynthesis and insulin exocytosis. Because both studies were not conducted in animals at the same ages (45 weeks [6] vs. 75 weeks old in the current study), we cannot exclude an age-dependent effect in these tissues. We can, however, propose that the differential expression of lamin isoforms may have a tissue-specific impact on mitochondrial gene transcription.

RNA-seq analysis and translation rate analysis revealed an increase of global protein synthesis in the islets of aged Lmna^LCS/LCS associated with an increased life span, as described in our previous study (6). It is widely accepted that reduced protein synthesis promotes longevity (19). Indeed, protein biosynthesis in aging is accompanied by more frequent production of damaged proteins due to transcriptional or translational errors and cotranslational protein misfolding. The increased life span associated with an upregulation of global protein synthesis in our aged-Lmna^LCS/LCS mice suggests that proteins synthetized within the pancreatic islets are well-folded and biologically active, possibly due to the maintenance of an efficient protein quality control system. This hypothesis is supported by our demonstration that insulin, the major protein synthesized by β-cells, is properly secreted and efficient during aging, because aged Lmna^LCS/LCS mice remain glucose tolerant and normoglycemic. This functional insulin overproduction may contribute to the maintenance of a positive-feedback loop of insulin-regulated protein synthesis via the mammalian target of rapamycin pathway (20).

How Lmna isoforms are able to modulate gene transcription in β-cells remains to be elucidated. Previous studies have described epigenetic dysregulation in pancreatic islets during T2D, showing a differential methylation pattern associated with islet dysfunctions (21) and changes in the open chromatin landscape of human islets (22). Moreover, DNA methylation is a dynamic process in aging β-cells. De novo methylation of proliferation-related genes and demethylation of genes related to β-cell function increase insulin secretion during aging (23). Nevertheless, we did not observe any difference in DNA methylation in Lmna^LCS/LCS versus WT pancreatic islets. This finding shows that the selective expression of lamin C does not influence overall DNA methylation in pancreatic islets and points to another regulatory mechanism.

In metazoan nuclei, specific lamina-associated domains contribute to the overall spatial organization of the genome and are associated with gene repression (24). Cellular defects in cells from patients with progeria include widespread alterations in chromatin organization, such as the loss of heterochromatin domains and changes in epigenetic markers (25). Furthermore, it has been shown that lamin A/C can modulate transcription through the regulation of epigenetic factors (26). Therefore, the expression of only one Lmna isoform may have a crucial impact on the spatial compartmentalization of inactive or active chromosome domains, which in turn may have consequences on transcriptional dysregulation. Surprisingly, we did not observe any global changes in chromatin accessibility through ATAC-seq experiments in our mouse models. Nevertheless, Lmna^LCS/LCS pancreas show an active enhancer mark (H3K27Ac) relocalization to peripheral heterochromatin. The localization of an active histone mark in a heterochromatin environment has already been described showing that acetylation of histone in silent domains drives the mislocalization and derepression (27). Furthermore, there is increasing evidence for active and regulated transcription in the heterochromatin at the nuclear periphery (28). Altogether, these observations lead us to speculate that the phenotype observed in Lmna^LCS/LCS mice could be due to transcriptional reactivation of a subset of mitochondrial and translational genes through histone mark relocalization rather than by reorganizing global chromatin accessibility or modifying DNA methylation. The role of lamin C in the nuclear localization of specific loci and their respective histone marks requires further investigations.

In conclusion, Lmna^LCS/LCS mice present an enhanced adaptive response to the increased metabolic demands during the course of diabetes and aging. This is probably due to efficient glucose sensing and insulin production by their neogenerated islets (29). Our study reveals an important role for lamin C in pancreatic β-cells, namely a protective role against metabolic defects arising from β-cell dysfunction. This effect may promote the healthy aging of the Lmna^LCS/LCS mice and sustain their extended life span (6). The discovery of new drugs targeting Lmna alternative splicing may represent an innovative therapeutic strategy to counteract aging β-cell decline and improve diabetes treatment.

Acknowledgments. The authors thank David Lleres and Cyril Esnault (Institut de Génétique Moléculaire de Montpellier [IGMM], Montpellier, France) for discussions and scientific critique, Montpellier Ressources Imagerie (MRI) and Réseau d’Histologie Expérimentale de Montpellier (RHEM) for assistance in cytometry and experimental histology, and the animal facility platforms Réseau des Animaleries de Montpellier (RAM) and Zone d’Expérimentation et de Formation de l’IGMM (ZEFI). The authors also thank Sylvain de Rossi, Myriam Boyer, and Chantal Cazevieille (MRI, Montpellier, France) for technical assistance, Magali A. Ravier (Institut de Génomique Fonctionnelle [IGF], Montpellier, France) for advice for the islet isolation from mice, and Amal Makrini (IGMM, Montpellier, France) for the bioinformatics support.

Funding. This work was supported by grants from the Swiss National Science Foundation (Ambizione PZ00P3-168077 to I.C.L.-M.), Conseil Régional Hauts de France et Métropole Européenne de Lille (to X.G. and J.-S.A.), European Genomic Institute for Diabetes (ANR-10-LABX-46 to J.-S.A.), Agence Nationale de la Recherche (BETAPLASTICITY, ANR-17-CE14-0034 to J.-S.A.), European Foundation for the Study of Diabetes (to J.-S.A.), I-SITE Université Lille Nord-Europe (EpiRNA-Diab to J.-S.A.), Société Francophone du Diabète (to J.-S.A.), INSERM, CNRS, Lille University, Association pour la Recherche sur le Diabète (to J.-S.A.), and OSEO Innovation Stratégique Industrielle CaReNA (I 13 03 008W to J.T.).

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. M.d.T. and C.C. designed, performed, and interpreted the majority of the experiments. P.C. performed histological analysis. M.P. and C.B. contributed to the RNA-seq and RRBS analysis as well as the bioinformatics analysis. X.G. gave advice for islet isolation from mice. I.C.L.-M. contributed to the aging experiments. M.d.T. and C.C. wrote the manuscript. I.C.L.-M., J.-S.A., and J.T. reviewed and approved the manuscript. M.d.T. and C.C. are the guarantors of this work and, as such, had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.