Lrrc55 is a novel prosurvival factor in pancreatic islets

Abstract

Pancreatic islets adapt to the increase in insulin demand during pregnancy by upregulating β-cell number, insulin synthesis, and secretion. These changes require prolactin receptor (PrlR) signaling, as mice with PrlR deletion are glucose intolerant with a lower β-cell mass. Prolactin also prevents β-cell apoptosis. Many genes participate in these adaptive changes in the islet, and Lrrc55 is one of the most upregulated genes with unknown function in islets. Because Lrrc55 expression increases in parallel to the increase in β-cell number and insulin production during pregnancy, we hypothesize that Lrrc55 might regulate β-cell proliferation/apoptosis (thus β-cell number) and insulin synthesis. Here, we found that Lrrc55 expression was upregulated by >60-fold during pregnancy in a PrlR-dependent manner, and this increase was restricted only to the islets. Overexpression of Lrrc55 in β-cells had minimal effect on β-cell proliferation and glucose-stimulated insulin secretion but protected β-cells from glucolipotoxicity-induced reduction in insulin gene expression. Moreover, Lrrc55 protects β-cells from glucolipotoxicity-induced apoptosis, with upregulation of prosurvival signals and downregulation of proapoptotic signals of the endoplasmic reticulum (ER) stress pathway. Furthermore, Lrrc55 attenuated calcium depletion induced by glucolipotoxicity, which may contribute to its antiapoptotic effect. Hence our findings suggest that Lrrc55 is a novel prosurvival factor that is upregulated specifically in islets during pregnancy, and it prevents conversion of adaptive unfolded protein response to unresolved ER stress and apoptosis in β-cells. Lrrc55 could be a potential therapeutic target in diabetes by reducing ER stress and promoting β-cell survival.

INTRODUCTION

Pancreatic islets adapt to conditions of increased insulin demand, such as obesity and pregnancy, by upregulating β-cell mass and function. Diabetes occurs when pancreatic islets cannot produce enough insulin to meet metabolic requirement because of β-cell functional defect and/or reduced β-cell mass. Compared with nondiabetic individuals, patients with type 2 diabetes have a lower β-cell mass and a higher β-cell apoptosis rate (5). One contributing factor is the high free fatty acid and glucose level, a condition termed glucolipotoxicity, that is often seen in prediabetic, obese, and insulin-resistant individuals (26). Glucolipotoxicity causes β-cell dysfunction and cell death, leading to β-cell loss and deterioration in glycemic control.

Pregnancy is characterized by insulin resistance, and it provides a model for successful β-cell adaption to many of the same stimuli that lead to diabetes; as such it can be used to uncover mechanisms that might protect from diabetes. During pregnancy, a number of beneficial changes occur in pancreatic islets, including 1) increased insulin synthesis and glucose-stimulated insulin secretion, 2) reduction in the threshold of glucose stimulation, 3) improved gap-junction coupling, 4) increased glucose oxidation and cAMP metabolism, 5) increased β-cell proliferation, and 6) protection from apoptosis. These adaptive changes require prolactin receptor (PrlR) signaling (2, 35), as mice with PrlR deletion (3, 14) develop glucose intolerance during pregnancy that has been associated with a blunted β-cell proliferation rate, a lower β-cell mass, and diminished serum insulin levels. Furthermore, activation of PrlR also prevents β-cell apoptosis. In vitro exposure to prolactin or placental lactogens activates the prolactin receptor and protects human islets and β-cells in culture from cell death induced by serum deprivation or cytokines (21, 38). Activation of PrlR also protects rodent islets and INS-1 cells (a rat β-cell line) from glucolipotoxicity-induced apoptosis (21). Of note, studies in humans have linked low prolactin levels during pregnancy with increased risk of prediabetes/diabetes (27, 39).

A large number of genes are regulated in the β-cells during pregnancy in a PrlR-dependent manner, and many of them are linked to upregulation of β-cell mass and function to avert gestational diabetes (2, 3, 15, 18–21, 42, 44). For example, tryptophan hydroxylase-1 (Tph1) is upregulated by >60-fold during pregnancy, and it stimulates β-cell proliferation during pregnancy (19). Osteoprotegrin is another a prolactin-regulated gene that is upregulated by pregnancy, obesity, and during β-cell regeneration; it stimulates β-cell proliferation by inhibiting RANKL (20). Recently, Pax8 was identified as a candidate gene for gestational diabetes, as its expression is induced by pregnancy and prolactin, and overexpression of Pax8 protected islets against apoptosis (24). Hence it appears that many genes are recruited in the islets during pregnancy to increase insulin production and promote β-cell survival, highlighting the importance of identifying additional genes that may participate in this signaling network to prevent gestational diabetes.

One of the most highly upregulated genes during pregnancy is Lrrc55 (leucine-rich repeat containing 55) (19). Lrrc55 is expressed mainly in the brain, and, under nonpregnant conditions, its expression is barely detectable in the pancreatic islets. During pregnancy, however, its expression is upregulated by more than 60-fold, but its function in pancreatic islet is unknown. Because the increase in Lrrc55 expression during pregnancy parallels the increase in prolactin and placental lactogens, and prolactin and placental lactogens stimulate β-cell proliferation and insulin synthesis and prevent β-cell apoptosis, we hypothesized that Lrrc55 may be involved in these processes in β-cells. Here, we demonstrated that, during pregnancy, Lrrc55 is upregulated specifically in the islets. Overexpression of Lrrc55 in β-cells had minimal impact on β-cell proliferation or glucose-stimulated secretion. However, it protected β-cells from apoptosis caused by exposure to a diabetic milieu, exemplified by high levels of the saturated free fatty acid palmitate (PA), and attenuated the activation of the apoptosis pathway.

MATERIALS AND METHODS

Materials.

Chemicals were purchased from Sigma-Aldrich unless otherwise specified. Cell culture reagents were purchased from Life Technologies. Collagenase P was from Roche.

Mice.

C57BL/6, heterozygous PrlR-null mice (PrlR^+/−) on a C57BL/6 background, and obese db/db mice and their lean controls were purchased from Jackson Laboratory. Mice were maintained on a 12-h:12-h light/dark cycle with liberal access to food and water and studied at 3–4 mo of age. Working stock of PrlR^+/− mice was generated by crossing PrlR^+/− mice with wild-type PrlR^+/+ mice. Timed pregnant mice from each group at gestational days 0 (G0), G9, G12, G15, and postpartum day 4 (P4) were studied. Obese db/db mice and lean controls were studied at 3–4 mo of age after at least 2 wk of diabetes (fasting blood glucose >15 mM). All experimental procedures were approved by the Animal Use Review Committee at the University of Calgary in accordance with standards of the Canadian Council on Animal Care.

Cell culture.

INS-1-832/13 cells were obtained from Dr. Chris Newgard (12). Cells were seeded at 2 × 10⁵ cells/well in RPMI 1640 media supplemented with HEPES (0.5 M), l-glutamine (100 mM), sodium pyruvate (50 mM), β-mercaptoethanol (2.5 mM), 10% FBS, and penicillin/streptomycin and grown to 80% confluence. PA was prepared by dissolving 100 mM sodium PA in 50% ethanol at 70°C in a shaking water bath, which was then complexed to 10% BSA (endotoxin free), filtered, and dissolved in RPMI 1640 to reach a final concentration of 0.5–1.0 mM PA. Stock solutions of thapsigargin (1 mM) were prepared in DMSO and dissolved in RPMI 1640. For experiments, INS-1-832/13 cells were treated with 1 mM PA in the presence of 33 mM d-glucose (total concentration) for 6–24 h (1) or 1–10 μM thapsigargin for 6–24 h, as indicated in the figures. MIN6 cells were obtained from Dr. Oka (17). Cells were cultured in DMEM supplemented with l-glutamine (2 mM), 10% FBS, and penicillin/streptomycin and treated similarly as described above for INS-1-832/13 cells. Controls were incubated in culture media containing vehicle and the nonmetabolizable l-glucose to reach a final concentration of 33 mM glucose.

Islet isolation.

Pancreatic islets were isolated from nonfasted adult female mice. The pancreas was first distended using collagenase P (0.66 mg/ml in Hanks’s balanced salt solution, 2.5 mL/pancreas), surgically removed, and then incubated at 37°C for 15 min under constant agitation. Islets were handpicked and cultured overnight in RPMI 1640 supplemented with 10% FBS and penicillin/streptomycin. For some experiments, islets were treated with 0.5 mM PA in the presence of 33 mM d-glucose (total concentration) for 72 h. Control islets were incubated in culture media containing vehicle and the nonmetabolizable l-glucose for a final concentration of 33 mM glucose, as a control for osmolality.

Microarray and real-time qPCR.

Total islet RNA (100–200 islets/mouse) was extracted using the RNeasy Mini Kit (Qiagen), and RT-qPCR was performed as previously described (13). The relative amount of RNA was normalized to the housekeeping gene phosphoglycerate kinase 1 (Pgk1), and expression level was quantified by the ∆∆Ct method. Pgk1 was chosen after testing >10 genes from Qiagen’s reference gene panel, and it showed the least variability of all the genes tested when we compared islets from nonpregnant and pregnant mice (13). Primers used are listed in Supplemental Table S3 (all Supplemental material is available at: https://doi.org/10.6084/m9.figshare.9828374.v1). For microarray experiments, 150 islets each from three PrlR^+/+ and four PrlR^+/− mice on day 15 of pregnancy were used. Total RNA was extracted and genomic DNA removed with RNeasy Plus Micro Kit (Qiagen). RNA quality of RNA integrity number (RIN) was assessed using the Agilent RNA 6000 NanoChip on 2100 Bioanalyzer (Agilent Technologies). RNA was quantified using a NanoDrop 1000 instrument (NanoDrop Technologies), and 250 ng of RNA with RIN >8.5 was labeled using Whole Transcript Kit (Ambion). Each sample of 5.5 μg of ssDNA was biotinylated, fragmented, and hybridized to an Affimetrix GeneChip Mouse Gene 1.0 ST array at 45°C for 16 h, according to the manufacturer’s instruction. GeneChip array data files were generated using GeneChip Command Console Software (AGCC), and statistical analysis was carried out using GeneSpring software (Agilent Technologies) (31). The raw data sets discussed in this study have been deposited in the NCBI Gene Expression Omnibus and are accessible through GEO Series accession number GSE125350 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE125350). Of the 20,000 genes represented on the array, the fold change was calculated compared with control.

Insulin secretion assay.

MIN6 cells were seeded at 2 × 10⁵ cells/well in a 24-well plate and cultured in DMEM (25 mM glucose) supplemented with 10% FBS and penicillin/streptomycin for 48 h until they reached 80% confluence (30). Cells were washed with Krebs-Ringer buffer (KRB) containing no glucose, followed by incubation in 500 μL of KRB with no glucose for 120 min to restore the cells to basal state. Glucose-stimulated insulin secretion was performed by sequential incubation of cells in 500 μL of KRB with 0 mM, then 2.8 mM, 5.5 mM, 11 mM, 16 mM, or 22 mM glucose for 1 h each at 37°C. The supernatant from the secretion assay was collected for insulin analysis, using the STELLUX Chemiluminescence ELISA (ALPCO). Total insulin was determined by the acid-ethanol extraction method (16). The amount of insulin secreted was normalized to the DNA content.

Adenovirus construction and infection.

Adenovirus containing Lrrc55 and mCherry or green fluorescent protein (GFP) (AdLrrc55 or AdGFP) was constructed using the AdEasy system. INS-1 and MIN6 cells were infected with a multiplicity of infection (MOI) of 100:1 (i.e., 100 viral particles per cell) for 24 h, and experiments were carried out 48 h after infection. For infection of islets, we pretreated whole islets with EDTA (2 mM) at 37°C for 5–7 min to gently disrupt the islet capsule, then infected with adenovirus at MOI of 100:1 in RPMI media overnight. Fresh media were added the next day for a total infection time of 48 h.

Cell death assay.

INS-1-832/13 cells (2 × 10⁵ cells/well) or islets (100 islets/well) were seeded on 0.4% gelatin-coated coverslips in a 24-well plate, cultured overnight, and infected with AdGFP or AdLrrc55 for 48 h. The cells were then treated with 0.5–1.0 mM PA in the presence of 33 mM glucose for 24 h (INS-1-832/13 cells) to 72 h (mouse islets) (1). The PA-containing media were then removed; cells were fixed with 4% paraformaldehyde solution and then permeabilized with 0.1% Triton X-100. Apoptotic β-cells were detected by a colorimetric TdT-mediated dUTP nick-end labeling (TUNEL) staining kit per the manufacturer’s instructions (Roche) or by immunostaining for cleaved caspase-3. To detect cleaved caspase-3, fixed and permeabilized cells were incubated with rabbit-cleaved caspase-3 antibody at 1:1,000 dilution in 1% goat serum/PBS at 4°C overnight (catalog no. 9661, Cell Signaling) followed by 1-h incubation with goat anti-rabbit-Cy3 (Jackson ImmunoResearch) or goat anti-rabbit-Alexa Fluor-488 antibodies (Molecular Probes) at 1:300 dilution in 1% goat serum/PBS at room temperature for 1 h. After three washes with PBS, cells were mounted with Dako fluorescent mounting medium (Dako).

Western immunoblotting.

Whole-cell protein extracts were obtained from INS-1-832/13 cells as previously described (15). Protein (20–30 μg) was separated by SDS-PAGE and transferred onto polyvinylidene difluoride filters, blocked in 3% BSA at room temperature for 1 h, then incubated with primary antibodies [rabbit anti-binding immunoglobulin protein (BIP) at 1:1,000; Enzo Life Sciences; rabbit anti-phospho-eukaryotic initiation factor 2 (eIF2), eIF2, inositol-requiring enzyme 1 (IRE1)-α, actin, GAPDH antibodies, or mouse anti-CHOP (C/EBP-homologous protein) antibody, all at 1:1,000; all from Cell Signaling] at 4°C overnight. Blots were then washed with Tris-buffered saline with 0.1% (vol/vol) Tween and then incubated for 1 h with horseradish peroxidase-conjugated secondary antibodies (donkey anti-rabbit or donkey anti-mouse at 1:10,000; Amersham). Protein was visualized by the enhanced chemiluminescence method and scanned within the linear range using ImageJ software.

Calcium measurement.

Changes in intracellular free calcium were measured via ratiometric Ca²⁺ imaging using the Ca²⁺ indicator dye Fura-2 AM (Life Technologies) as previously described (22). Briefly, cells grown on glass coverslips were loaded with 5 μM Fura-2 AM and 0.02% (wt/vol) Pluronic F-127 in HEPES-buffered saline solution for 20 min, then transferred to the recording chamber mounted on a TE2000-U inverted microscope equipped with an S-Fluor ×20/0.5 NA objective lens (Nikon). Cells were excited alternately using 340- and 380-nm light from a monochromator, and cellular fluorescence was acquired at >510 nm once every 8 s using a 4.2-megapixel CMOS camera. Fluorescent signals within individual cells were analyzed with RatioPro software (Horiba Scientific).

Statistical analysis.

Comparisons among three or more groups were performed by ANOVA with Tukey’s multiple-comparison tests, whereas comparisons between two groups were performed by a Student’s t-test or Mann-Whitney U-test where appropriate. A P value <0.05 was considered a statistically significant difference.

RESULTS

Lrrc55 is upregulated in islets during pregnancy in a PrlR-dependent manner.

In search of novel PrlR targets, we isolated islets from heterozygous PrlR-null (PrlR^+/−) and wild-type (PrlR^+/+) mice on day 15 of pregnancy and performed gene expression analysis using Affymetrix GeneChip Mouse Gene 1.0 ST Array. We used heterozygous mice because the homozygous PrlR-null (PrlR^−/−) mice are infertile (14, 25). The microarray identified PrlR as being downregulated by 1.65-fold in the PrlR^+/− mice compared with the PrlR^+/+ mice. Of the >27,000 genes thus identified on the array, 155 genes were differentially regulated by >1.65-fold, and, with the use of the PATHER gene classification system, 21% of these 155 genes are classified as being involved in regulation of metabolic processes (Supplemental Table S1). We also cross referenced our list of differentially regulated genes with published data that compared gene expression pattern in islets from pregnant and nonpregnant wild-type mice (19, 28), and we identified Lrrc55 as one of the most differentially regulated genes by pregnancy and by PrlR status (Supplemental Table S2). We also detected a significant reduction in Tph1 (19), osteoprotegerin (Tnfrsf11b) (20), and Npas4 (30) expression in the PrlR^+/− mice during pregnancy, consistent with published results.

Lrrc55 was identified as a paralog of Lrrc26, which has been previously described as a γ-subunit of BK channels (43). Lrrc55 is highly expressed in neurons, with negligible expression in the pancreatic islets in nonpregnant mice. During pregnancy, however, expression of Lrrc55 increases by >60 fold, peaks on day 15 of gestation, and declines toward the prepregnancy level post partum (Fig. 1A). Of note, Lrrc55 expression in the pregnant islets is significantly higher than in the hypothalamus, the organ with relatively high Lrrc55 expression under both nonpregnant and pregnant conditions (Fig. 1B). This increase is blunted in the PrlR^+/− mice, and it is restricted to the islets because the level of Lrrc55 does not change significantly in other tissues during pregnancy (Fig. 1B). Expression of its paralog, Lrrc26, does not increase in the islets during pregnancy, whereas expression of Lrrc38 increased sixfold (Fig. 1C).

Fig. 1.

Lrrc55 expression. A: islets were isolated from nonpregnant and pregnant mice on days 9 and 15 of pregnancy (gestational day 15, G15) and postpartum day 4 (P4), and mRNA expression of Lrrc55 was determined by RT-qPCR, normalized to the housekeeping gene phosphoglycerate kinase 1 (PGK1) and expressed relative to the levels found in nonpregnant PrlR^+/+ mice. The expression levels were compared among the gestational stages within each genotype and between PrlR^+/+ and PrlR^+/− mice at each gestational stage. n = 4 mice/gestational stage/genotype. **P < 0.01 and ***P < 0.001 compared with nonpregnant mice, and #P < 0.05 and ##P < 0.01 compared with PrlR^+/+ mice at the same gestational stage. B: Lrrc55 mRNA expression levels in various tissues were compared between nonpregnant and pregnant (day 15) PrlR^+/+ mice. Lrrc55 expression was normalized to the housekeeping gene PGK1 and expressed relative to the levels found in islets of nonpregnant PrlR^+/+ mice. n = 3–4 mice/group. C: expression of Lrrc55 paralogs, Lrrc26 and Lrrc38, was determined in islets isolated from nonpregnant and pregnant (day 15) PrlR^+/+ mice. mRNA levels were normalized to the housekeeping gene PGK1 and expressed relative to the Lrrc55 levels found in nonpregnant PrlR^+/+ mice. n = 4 mice/group. *P < 0.05 and **P < 0.01 compared with the nonpregnant mice.

As the increases in Lrrc55 expression and β-cell proliferation in islets during pregnancy both peak on day 15 of gestation, we surmised that Lrrc55 might participate in regulation of β-cell proliferation. Hence we overexpressed Lrrc55 in mouse islets using an adenovirus containing Lrrc55. Interestingly, we found that 0.94% of cells overexpressing Lrrc55 stained positive for Ki-67, an established marker for proliferating β-cells, whereas 0.58% of cells infected with the control virus were Ki-67 positive (Fig. 2). Although this difference is statistically significant, it likely reflects the slight drop in the percentage of Ki-67-positive cells in the GFP controls compared with the untreated cells, and Lrrc55 overexpression per se has little effect on β-cell proliferation.

Fig. 2.

Lrrc55 and β-cell proliferation. A: islets were infected with AdLrrc55 or AdGFP for 48 h and immunostained for Ki-67 and insulin. Treatment with 22 mM glucose for 48 h was used as a positive control for β-cell proliferation. No treatment condition represents islets cultured in media with 11 mM of glucose. Percentages of Ki-67-positive β-cells that were also positive for Lrrc55 or green fluorescent protein (GFP) control are presented as means ± SE from 3 separate experiments, and a total of at least 1,000 cells/condition were counted. Data were analyzed by 1-way ANOVA with Tukey’s multiple-comparison test. ****P < 0.001 and **P < 0.01 compared with the no-treatment condition, and ##P < 0.01 compared with the AdGFP-infected control. B, top: representative image showing that most cells from isolated islets are insulin-positive β-cells. Green arrows point to Ki-67-positive cells. Top: yellow arrows point to insulin and Ki-67-double-positve cells. Bottom: yellow arrows point to Ki-67-positive Lrrc55-overexpressing cells.

Lrrc55 expression in β-cells is upregulated by a diabetogenic milieu.

During pregnancy, β-cells increase insulin production to compensate for maternal insulin resistance. The associated increase in protein folding elicits an unfolded protein response (UPR), indicated by the upregulation of IRE-1α and CHOP on day 12 of pregnancy (Supplemental Fig. S1). However, β-cells are protected from progressing to apoptosis during pregnancy (14). This observation prompted us to determine the role of Lrrc55 in apoptosis. First, we examined Lrrc55 expression in the presence of high levels of fatty acid and glucose, an environment commonly found in diabetes. When we exposed mouse islets to PA in the presence of glucose, there was a significant increase in Lrrc55 expression within 24 h (Fig. 3A) although the magnitude of increase was not nearly as robust as that seen during pregnancy (i.e., 15-fold increase in PA-treated islets vs. 60-fold increase during pregnancy). Expression of other family members, such as Lrrc38, was not significantly increased upon PA exposure (Supplemental Fig. S2). Because PA is known to cause ER stress in β-cells, we also measured Lrrc55 expression after treating mouse islets with thapsigargin, a SERCA inhibitor and ER stress inducer (30). Interestingly, similar to exposure to PA, we also observed a significant increase in Lrrc55 expression within hours of thapsigargin exposure (Fig. 3B). Lastly, we found that Lrrc55 is upregulated by >10-fold in islets of obese db/db mice, another model of metabolic syndrome and diabetes (Fig. 3C). These results suggest that Lrrc55 is upregulated in conditions associated with cellular stress and diabetes.

Fig. 3.

Expression of Lrrc55 in response to stimuli. A: mouse islets were treated with 0.5 mM palmitate (PA) in the presence of 33 mM glucose for 12–48 h. Lrrc55 mRNA expression levels were determined by RT-qPCR, normalized to the housekeeping gene PGK1, and expressed relative to the levels found in untreated controls that were cultured with only vehicle without PA for 12 h. n = 3–5 mice/group. Data were analyzed by ANOVA with Tukey’s multiple-comparison tests. *P < 0.05 and **P < 0.01 compared with the 12-h untreated controls. B: mouse islets were treated with thapsigargin (1 µM) for the indicated times, and Lrrc55 mRNA expression levels were determined by RT-qPCR, normalized to PGK1, and expressed relative to untreated controls that were cultured with media containing DMSO for 12 h. *P < 0.05 compared with DMSO-treated cells by ANOVA with Tukey’s multiple-comparison tests. C: islets were isolated from obese diabetic db/db mice, lean controls, and c57/BL6 mice, and mRNA expression of Lrrc55 was determined by RT-qPCR, normalized to the housekeeping gene PGK1, and expressed relative to the levels found in nonpregnant c57/BL6 female mice. Results are presented as means ± SE from 4–6 mice/genotype.

Lrrc55 attenuated stressor-induced β-cell apoptosis.

Because expression of Lrrc55 in pancreatic islets is upregulated by >60-fold on day 15 of pregnancy in a PrlR-dependent manner (Fig. 1A) and prolactin protects β-cells against cytokine (38) and high-fat-induced cell death (21), we wished to determine whether Lrrc55 plays a protective role in β-cell apoptosis. To address this question, we overexpressed Lrrc55 in INS-1-832/13 cells and mouse islets using an adenovirus vector and then incubated the cells with PA for 24 h (INS-1-832/13) to 72 h (islets) to cause cellular dysfunction and cell death. We found that overexpression of Lrrc55 protected β-cells from PA-induced apoptosis, as measured by TUNEL and caspase-3 activation (Fig. 4, A–C). Interestingly, islets from pregnant mice expressed 60-fold more Lrrc55 than islets from nonpregnant mice, and they are also more resistant to PA-induced apoptosis (Supplemental Fig. S3). Because PA perturbs ER function, causing unresolved ER stress leading to β-cell apoptosis, we treated β-cells with a well-known ER stress inducer, thapsigargin, to determine whether Lrrc55 is also protective under this condition in β-cells. In agreement with results obtained from PA-treated cells, overexpression of Lrrc55 protected β-cells from thapsigargin-induced apoptosis (Fig. 4D).

Fig. 4.

Lrrc55 protects β-cells from apoptosis. A: INS-1-832/13 cells were infected with AdLrrc55 or AdGFP for 48 h followed by treatment with palmitate (1 mM) in the presence of 33 mM glucose for 24 h. Cell death was measured by counting TUNEL-positive cells as a fraction of the total cell population. Uninfected INS-1-832/13 cells incubated in culture media containing 1% BSA were used as controls. At least 5,000 INS-1-832/13 cells were counted per experiment. Results are presented as means ± SE for the number of TUNEL-positive cells measured in 3 independent experiments. Data were analyzed by 1-way ANOVA with Tukey’s multiple-comparison tests. ***P < 0.001 vs. Lrrc55-overexpressing cells treated with palmitate. Representative images of TUNEL staining of the AdGFP- or AdLrrc55-infected cells treated with palmitate are presented. GFP, green fluorescent protein. B: isolated mouse islets were infected with AdLrrc55 or AdGFP for 48 h followed by treatment with palmitate (0.5 mM) in the presence of 33 mM glucose for 72 h. Cell death was measured by counting TUNEL-positive cells as a fraction of the total cell population. Each experiment uses islets pooled from 2–3 mice. At least 3,000 cells were counted per experiment. n = 3 independent experiments. Data were analyzed by Student’s t-test. *P < 0.05 vs. GFP-expressing cells. Arrow points to cells double-positive for GFP and TUNEL in islets. C: using the same treatment conditions as described above, the percentage of Lrrc55-positive or GFP-positive cells also stained positive for cleaved caspase-3 was quantified. Results are presented as means ± SE of 3 independent experiments; each used islets pooled from 3–4 mice, and at least 250 cells were counted for each condition per experiment. **P < 0.01 compared with cells infected with AdGFP by Student’s t-test. White arrows point to cells double-positive for GFP and cleaved caspase-3 in islets. D: INS-1-832/13 cells infected with AdGFP or AdLrrc55 for 48 h were treated with 10 μM thapsigargin (Tg) for 24 h in serum-free media, and cell death was measured by counting TUNEL-positive cells as a fraction of the total cell population. Cells incubated in media containing 0.01% DMSO in the absence of Tg served as the vehicle control for this experiment. Results are presented as means ± SE of apoptotic cells measured in 3 independent experiments. ***P < 0.001 vs. Tg-treated but not adenovirus-infected INS-1-832/13 cells.

To determine whether Lrrc55 protects β-cells from apoptosis through modulation of components of the UPR upstream of ER stress, we examined the activation of the ER stress pathways. Here, we found that the Lrrc55-mediated protective effect was accompanied by attenuation of the intrinsic ER stress pathway, as measured by expression of IRE-1α, CHOP, caspase-9, and Bax/Bcl-2 ratio (Fig. 5). There was also a trend for lower p-eIF2/eIF2 expression in the Lrrc55-overexpressing cells.

Fig. 5.

Lrrc55 expression attenuated the activation of endoplasmic reticulum (ER) stress pathways. Activation of the apoptosis pathway was determined in INS-1-832/13 (A, C, D, E, G, and I) and dissociated mouse islets (B, F, H, J). mRNA expression [inositol-requiring enzyme 1 (IRE1)-α, CHOP, caspase-9, Bax, and Bcl-2] or protein expression [IRE1-α, CHOP, phospho-eukaryotic initiation factor 2 (eIF2), and binding immunoglobulin protein (BIP)] were determined and compared between cells infected with either AdGFP or AdLrrc55 and treated with palmitate (PA) in the presence of 33 mM glucose. Uninfected cells not treated with PA (“no Ad no PA”) were used as control for the effect of culturing cells in vitro on the expression of molecules of the apoptosis pathway. Results are presented as means ± SE from 3–6 separate experiments. Data were analyzed by 1-way ANOVA with Tukey’s multiple-comparison tests. For A–H, **P < 0.01 and *P < 0.05 compared with uninfected cells not treated with PA at 6 or 12 h for INS-1-832/13 cells and at 24 h for islets. ##P < 0.01 compared with uninfected INS-1-832/13 cells not treated with PA at 24 h (A and E). For I and J, **P < 0.01 and *P < 0.05 comparing AdLrrc55- to AdGFP-infected cells at the same time point. Dark shaded bar = no adenovirus infection and no PA treatment; light shaded bar = AdGFP + PA; open bar = AdLrrc55 + PA. GFP, green fluorescent protein.

Lrrc55 attenuated PA-induced calcium depletion.

Two observations here suggest that Lrrc55 may modulate the intracellular Ca²⁺ store. First, Lrrc55 prevented thapsigargin-induced apoptosis in β-cells, and thapsigargin is known to deplete Ca²⁺ from the ER. Second, ER stress can be activated by ER Ca²⁺ depletion (6, 7), and Lrrc55 expression attenuated the ER stress response. To test the effect of Lrrc55 expression on intracellular Ca²⁺ store, we treated INS-1-832/13 cells with PA for 24 h and loaded the cells with the fluorescent calcium indicator Fura-2. Loaded cells were then acutely stimulated with cyclopiazonic acid (CPA), which inhibits the Ca²⁺-ATPase and promotes ER Ca²⁺ release, leading to dissipation of ER calcium store. The ensuing increase in cytosolic fluorescence indirectly reflects the releasable pool of ER Ca²⁺. Consistent with previous reports, treatment of β-cells with PA caused a modest depletion of the ER Ca²⁺ store, which was restored by Lrrc55 expression (Fig. 6). Overexpression of Lrrc55 per se, in the absence of exposure to PA, had no effect on CPA-induced ER Ca²⁺ release or basal intracellular calcium levels (data not shown).

Fig. 6.

Lrrc55 attenuated glucolipotoxicity-induced endoplasmic reticulum (ER) calcium depletion. INS-1 cells infected with AdGFP or AdLrrc55 for 48 h were treated with 1% BSA or 1 mM palmitate (PA) for 24 h before being loaded with 5 μM Fura 2-AM for 40 min, washed, and placed in HEPES-buffered saline solution for imaging. The cytosolic-free Ca⁺² level was measured as described in materials and methods. Where indicated by the horizontal bar, 20 μM cyclopiazonic acid (CPA) was added to release the ER Ca⁺² stores and was maintained in the bath for the duration of the recording. A smooth line was used to connect data points (means ± SD) acquired under each experimental condition, as denoted by the legend. B: summary of the effects of PA on cytosolic-free Ca⁺² during the 3 time periods indicated in A. The fluorescence ratio attained after addition of CPA in calcium-free medium is presented under 2. The data are presented as the means ± SD of n = 25–40 cells under each condition. Statistical comparisons among the 3 treatment conditions at a given time point were performed by ANOVA, followed by a Tukey’s multiple-comparison test; **P < 0.01, ***P < 0.001. AUC, area under the curve; GFP, green fluorescent protein.

Lrrc55 has minimal effect on glucose-stimulated insulin secretion.

Lastly, we investigated the role of Lrrc55 in glucose-stimulated insulin secretion because both Lrrc55 expression and insulin synthesis and secretion from islets are upregulated during pregnancy. Here, we found that β-cells overexpressing Lrrc55 secreted slightly less insulin at the basal state (i.e., 2.8 mM glucose) compared with the control although the difference did not reach statistical significance (Fig. 7A). Furthermore, Lrrc55 overexpression did not have a significant effect on glucose-stimulated insulin secretion (Fig. 7A). Interestingly, overexpression of Lrrc55 in β-cells was associated with an increase in insulin content (Fig. 7B), and PA treatment caused a modest but statistically significant drop in insulin gene expression in control cells, an effect that was reversed by overexpression of Lrrc55 (Fig. 7C).

Fig. 7.

Effect of Lrrc55 on glucose-stimulated insulin secretion and insulin content. A: glucose-stimulated insulin secretion was measured in MIN6 cells infected with AdGFP or AdLrrc55 as outlined in materials and methods. Amount of insulin secreted was normalized to total cellular DNA content. Results are presented as means ± SE from 4–6 separate experiments. Statistical comparisons among the 3 treatment conditions at a given time point were performed by ANOVA, followed by a Tukey’s multiple-comparison test, and *P < 0.05 compared with basal (2.8 mM glucose). GFP, green fluorescent protein. B: insulin content of MIN6 cells infected with AdGFP or AdLrrc55 for 48 h. Results are presented as means ± SE from 3 separate experiments. **P < 0.01 compared with cells infected with AdGFP. C: INS-1-832/13 cells infected with AdGFP or AdLrrc55 were treated with 1 mM palmitate (PA) in the presence of 33.3 mM glucose. INS-1 mRNA expression was analyzed by RT-qPCR, normalized to the expression level of PGK1, and expressed relative to control cells (incubated with control no adenovirus no PA media for 6 h). Results are presented as means ± SE of 5 independent experiments. Statistical comparisons among the 3 treatment conditions were performed by ANOVA, followed by a Tukey’s multiple-comparison test. ***P < 0.001 compared with the no Ad and no PA controls at 6 h. ###P < 0.001 and ##P < 0.01 compared with the GFP + PA-treated cells at the same time point. Dark shaded bar = no adenovirus infection and no PA treatment; light shaded bar = AdGFP + PA; open bar = AdLrrc55 + PA.

DISCUSSION

Pancreatic islets adapt to increased insulin demand by several mechanisms, including reducing the threshold of glucose stimulation, increasing insulin synthesis, glucose-stimulated insulin secretion, and gap-junction coupling, increasing glucose oxidation and cAMP metabolism, and upregulating β-cell proliferation (34, 36, 40, 41). All of these mechanisms peak around days 14–15 of rodent gestation, which is equivalent to the time of second trimester in human pregnancy. One of the most highly upregulated genes in the islets at this time point is Lrrc55. Before this work, nothing about the role of Lrrc55 in β-cells has been reported. Under normal conditions, Lrrc55 is selectively expressed in the brain and essentially undetectable in the pancreatic islets (43). Here, we report the novel observations that 1) Lrrc55 is significantly upregulated in pancreatic islets during pregnancy, and this upregulation is unique to islets and not observed in other tissues; 2) overexpression of Lrrc55 has little effect on β-cell proliferation; 3) Lrrc55 has minimal effect on glucose-stimulated insulin secretion; 4) Lrrc55 protects β-cells from glucolipotoxicity-induced apoptosis, potentially via attenuating activation of the ER stress pathways and maintaining intracellular calcium homeostasis; and 5) this may lead to the improved insulin synthesis and storage necessary to maintain both maternal and fetal glucose homeostasis during pregnancy.

Pancreatic β-cells can adapt to the increased insulin requirement of pregnancy by increasing β-cell proliferation and insulin synthesis (34). Previously, we have shown that PrlR signaling is important for this process, as PrlR^+/− mice developed impaired glucose tolerance because they failed to expand their β-cell mass adequately in response to the insulin resistance of pregnancy (14). Signaling molecules that participate in this process include but are not limited to Jak2, Stat5 (21), menin (18), IRS-2, Akt, p21 (15), FoxM1 (44), Foxd3, survivin (42), Tph-1 (19), Erk1/2 (2, 15), and MafB (3). In search of novel PrlR targets, we compared the gene expression pattern of islets from PrlR^+/+ and PrlR^+/− mice on day 15 of pregnancy by microarray analysis. We then examined published lists (19, 28) that compared gene expression of islets from pregnant vs. nonpregnant mice and identified genes that are differentially regulated by at least 2-fold by the pregnancy status and 1.5-fold by the PrlR status. Interestingly, some of the highest hits identified by this method turned out to be genes that were already known to have an important function in islets, such as Tph1 (19, 32) and Npas4 (30, 37), validating our approach.

We chose to examine Lrrc55 because it is one of the most highly upregulated genes in the islets during pregnancy. Little is known about the function of Lrrc55 in any tissue, and its function in islets has never been reported. Lrrc55 was identified as a paralog of Lrrc26, which is a γ-subunit of BK channels (43). BK channels are expressed in both rodent and human β-cells, where they participate in membrane repolarization and regulate Ca⁺² entry, action potential amplitude, and glucose-stimulated insulin secretion (4, 29). BK channels have been found on plasma membrane, ER, Golgi, and mitochondria (11). Interestingly, there is evidence that BK channels may protect β-cells from cell death, as pancreatic β-cells from transgenic mice with BK channel deletion have increased susceptibility to H₂O₂-induced apoptosis (8). Lrrc55 is a γ-subunit of BK channels, and, in HEK cells, overexpression of Lrrc55 causes a large negative shift in voltage-dependent BK channel activation (i.e., more channel activity at the resting membrane potential); the functional significance of this was not reported (43). In pancreatic islets, its expression is increased significantly during pregnancy, and we found it to be upregulated in islets of diabetic db/db mice. In agreement with this finding, Lrrc55 is one of the most upregulated genes in another model of glucose intolerance and impaired β-cell function, i.e., in transgenic mice with β-cell-specific deletion of TNF receptor 2 (i.e., βTRAF2) or βTRAF3 (23). These observations suggest that Lrrc55 may participate in regulation of β-cell function under conditions of increased insulin demands.

β-Cells are very sensitive to ER stress, as they are continually under pressure to synthesize and release large amounts of insulin. In response to changing nutrient conditions, islets increase insulin synthesis, and the associated increase in protein folding can elicit a UPR. However, in states of prolonged overnutrition, such as diabetes and obesity, the early adaptive responses cannot restore cellular homeostasis, and apoptotic pathways become activated (6). We exposed β-cells to high concentrations of glucose and free fatty acids (i.e., PA) to mimic the glucolipotoxic conditions often observed in diabetes (1) and found that overexpression of Lrrc55 provided protection against glucolipotoxicity-induced apoptosis. When we examined components of UPR, we found that Lrrc55 expression either prevented or delayed the activation of proapoptotic signals, such as CHOP, BAX, caspase-3, and caspase-9, while upregulating the expression of prosurvival signals, including Bcl-2. These findings are consistent with previous studies showing that the Prl-mediated protection against apoptosis also involves upregulation of Bcl2 expression (21). Interestingly, an elevated free fatty acid level is observed during human pregnancy (33); hence Lrrc55 may be part of the gene network upregulated during pregnancy to protect β-cells from the deleterious effects of elevated free fatty acid and insulin resistance.

As chronic treatment with PA has been shown to reduce ER Ca²⁺ store in β-cells (6, 10) and we observed that Lrrc55 can protect β-cells against apoptosis induced by the SERCA inhibitor thapsigargin, we measured the effect of Lrrc55 expression on intracellular Ca²⁺ dynamics. Consistent with published work, we found that β-cell exposure to PA caused a modest depletion of the releasable pool of ER Ca²⁺ as revealed by the ER Ca²⁺-ATPase inhibitor CPA. Expression of Lrrc55 restored the releasable pool of Ca²⁺ to a level comparable to the level seen in cells that were not exposed to PA. Because BK channels have been found on ER (11) and we observed colocalization of Lrrc55 with ER markers in β-cells (data not shown), it is plausible that Lrrc55 regulates ER Ca²⁺ flux and attenuates PA-mediated Ca²⁺ depletion. Alternatively, Lrrc55 may regulate cytosolic Ca²⁺ by altering Ca²⁺ flux across the plasma membrane, which secondarily affects ER Ca²⁺ store. The mechanism will need to be the subject of future studies.

Another interesting observation is that overexpression of Lrrc55 in β-cells led to an increase in insulin content and prevented the PA-induced reduction in insulin expression. This is not due to changes in expression of Pdx-1 or MafA, as we did not detect an increase in Pdx-1 or MafA expression under these conditions (data not shown).

Hence our results suggest the following model: in the ER lumen, optimal Ca²⁺ concentration is required for proper protein folding and chaperone function. Exposure to glucolipotoxicity distorts the intracellular Ca²⁺ store and ER function, which leads to an accumulation of improperly folded proteins and activation of ER stress (9). Lrrc55 expression appears to moderate PA-induced intracellular Ca²⁺ depletion, dampens the activation of ER stress, and protects cells against glucolipotoxicity-induced apoptosis. How Lrrc55 regulates intracellular Ca²⁺ stores is presently being investigated.

It is tempting to speculate that islets upregulate Lrrc55 as part of the UPR response because, during pregnancy, islets increase insulin synthesis and secretion in response to the insulin resistance of pregnancy. By upregulating prosurvival factors, islets can avoid tipping the balance toward a proapoptotic response. In support of this, we observed an increase in IRE-1α expression by day 9 of pregnancy and CHOP by day 12 of pregnancy in mouse islets, suggesting activation of UPR in islets in normal pregnancy. Taken together, it appears that Lrrc55 is a novel prosurvival factor that is upregulated in islets to prevent conversion of adaptive UPR to unresolved ER stress and apoptosis. Identification of Lrrc55 targets will add to our understanding of how β-cells adapt and prevent their own demise in a diabetogenic, stressed environment, which would ultimately contribute to our collective effort to treat and prevent β-cell failure and diabetes.

GRANTS

This work was supported by funds from Natural Sciences and Engineering Research Council of Canada (RGPIN/04937-2015), Diabetes Canada (OG-3-12-3670-CH), and Alberta Children’s Hospital Foundation to C. Huang, and Canadian Institute of Health Research (MOP 142222 and MOP 142467) to F. Lynn and A. Braun.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

G.M., A.P.B., F.C.L., and C.H. conceived and designed research; G.M., V.S., B.H., M.P., B.D.K., and A.P.B. performed experiments; G.M., V.S., B.H., M.P., B.D.K., A.P.B., and C.H. analyzed data; G.M., V.S., B.H., A.P.B., F.C.L., and C.H. interpreted results of experiments; G.M., V.S., B.H., and C.H. prepared figures; G.M. and C.H. drafted manuscript; G.M., V.S., B.H., M.P., B.D.K., A.P.B., F.C.L., and C.H. edited and revised manuscript; G.M., V.S., B.H., M.P., B.D.K., A.P.B., F.C.L., and C.H. approved final version of manuscript.