Copine 3 “CPNE3” is a novel regulator for insulin secretion and glucose uptake in pancreatic β-cells

Copine 3 (CPNE3) is a calcium-dependent phospholipid-binding protein that has been found to play an essential role in cancer progression and stages. However, its role in pancreatic β-cell function has not been investigated. Therefore, we performed a serial of bioinformatics and functional experiments to explore the potential role of Cpne3 on insulin secretion and β-cell function in human islets and INS-1 (832/13) cells. RNA sequencing and microarray data revealed that CPNE3 is highly expressed in human islets compared to other CPNE genes. In addition, expression of CPNE3 was inversely correlated with HbA1c and reduced in human islets from hyperglycemic donors. Silencing of Cpne3 in INS-1 cells impaired glucose-stimulated insulin secretion (GSIS), insulin content and glucose uptake efficiency without affecting cell viability or inducing apoptosis. Moreover, mRNA and protein expression of the key regulators in glucose sensing and insulin secretion (Insulin, GLUT2, NeuroD1, and INSR) were downregulated in Cpne3-silenced cells. Taken together, data from the present study provides a new understanding of the role of CPNE3 in maintaining normal β-cell function, which might contribute to developing a novel target for future management of type 2 diabetes therapy.


Results
Expression profile of copine genes in human islets and INS1 cell line. Using publicly available RNA-seq and microarray gene expression data obtained from many human pancreatic islets, we mapped the expression profile of the copines family. As shown in Fig. 1A, the microarray expression of CPNE3 was the highest expressed gene, followed by CPNE1, compared to other CPNE genes. To corroborate the microarray expression data, we analyzed the expression profile of CPNE genes using RNA-seq from human islets. As illustrated in Fig. 1B, CPNE3 gene showed the highest expression compared with other CPNE genes. Similarly, CPNE1 was ranked as the second-highest expressed gene (Fig. 1B). Based on these expression data in human pancreatic islets, CPNE3 was selected for functional investigations in pancreatic β-cell function. CPNE3 expression was further validated at protein level from fresh nondiabetic human pancreatic islets (n = 1) (Fig. 1C). Although microarray expression of CPNE3 was correlated negatively with HbA1c levels (Fig. 1D), no differential expression of CPNE3 in hyperglycemic donors (HbA1c ≥ 6%) compared to normoglycemic (HbA1c < 6%) donors was observed (not shown). Interestingly, RNA-seq expression for CPNE3 in hyperglycemic donors (HbA1c ≥ 6%) was significantly reduced (p ≤ 0.05) compared to normoglycemic (HbA1c < 6%) donors (Fig. 1E). Additionally, as clonal rat INS-1 cells are the most commonly used tool for functional validation, we profiled the expression of copine genes in INS-1 cells (n = 3). As shown in Fig. 1F, qPCR expression analysis revealed that all of the copine genes were expressed in INS-1 (832/13) cells. Cpne2 (Ct = 14.2) was the highest expressed gene based on the Ct values, whereas Cpne3 (Ct = 22.5) was ranked the fourth.

Silencing of Cpne3 in INS-1 (832/13) cells influences insulin secretion. As Cpne3 was chosen for
functional studies in pancreatic β-cell function, we performed siRNA-silencing of Cpne3 in INS-1 (832/13). Silencing efficiency assessment 24 h post-transfection by qPCR exhibited a significant reduction (75%; p < 0.05) in mRNA expression of Cpne3 as compared to the negative control ( Fig. 2A). This was further confirmed at the protein level as evaluated by western blot analysis (Fig. 2B). To assess the impact of Cpne3 silencing on insulin secretion, transfected cells were incubated for 1 h with 2.8 mM or 16.7 mM glucose. The results revealed a decrease in glucose-stimulated insulin secretion (GSIS) at 2.8 mM glucose (~ 20%; p < 0.05) and 16.7 mM glu- www.nature.com/scientificreports/ cose (~ 40%; p < 0.01) (Fig. 2C). More, a significant reduction (~ 40%; p < 0.05) on insulin secretion was observed in Cpne3-silenced cells stimulated with 35 mM KCl (a depolarizing agent) for 1 h compared to control cells (Fig. 2C). On the other hand, no significant effect was found when Cpne3-silenced cells were stimulated with 10 mM α-KIC (an agent that stimulates mitochondrial metabolism and ATP synthesis) as shown in Fig. 2C. Moreover, measurement of insulin content in Cpne3-silenced cells revealed a significant reduction (∼ 40; p < 0.05) than the negative control cells (Fig. 2D).
Impact of Cpne3 silencing on cell viability, cell proliferation, apoptosis and glucose uptake.. To gain more insights on a possible mechanistic defect of Cpne3 silencing on insulin secretion, we tested the effect of Cpne3 silencing on cell viability, apoptosis and glucose uptake. As shown in Fig. 3A, cell viability assessment by MTT assay showed to be unaffected in Cpne3-silenced cells compared to the negative control. For apoptosis analysis, Annexin-V and PI staining were used to assess apoptosis in transfected and control cells in the presence or absence of a combination of pro-apoptotic cytokines (IL-1β, TNFα and INFγ) for 24 h. As illustrated in Fig. 3B, the percentage of apoptosis (early and late) in transfected cells cultured in the absence or presence of cytokines was comparable to control cells. In addition, we investigated the impact of Cpne3 silencing on cell proliferation using EdU labeling assay. We found that Cpne3 silencing has no effect on cell proliferation when compared to normal cells as shown in Fig. 3C. To this end, we investigated the impact of Cpne3 silencing on glucose uptake in INS-1 (832/13) cells. As shown in Fig. 3D, a marked reduction (~ 40%) of glucose uptake was observed in Cpne3-silenced cells compared to control cells.

Cpne3 silencing influences β-cell functional genes in INS-1 (832/13). The consequences of Cpne3
silencing on the function of β-cell were investigated for several genes at mRNA and protein levels. Analysis of the mRNA expression of genes involved in insulin production revealed a significant (~ 30%; p < 0.05) downregulation of Ins1 and Ins2 (Fig. 4A,B). Expression of Pdx1 was not affected in Cpne3-silenced cells relative to the negative control (Fig. 4C). Likewise, protein expression was markedly reduced of pro/insulin (~ 20%; p < 0.05) and NeuroD1 (~ 30%; p < 0.05) (Fig. 4G,H), whereas PDX1 expression was not affected in CPNE3-silenced cells relative to the negative control ( Fig. 4I). More, analysis of mRNA expression of Glut2, Insr and Gck, which involved in glucose-sensing and insulin signaling were significantly downregulation of Glut2 (~ 30%; p < 0.05) (

Discussion
In this study, we profiled the transcript levels of the 9 members of copine family in human pancreatic islets using microarray and RNA-sequencing data obtained from a large number of donors. CPNE3 was highly expressed in human islets, inversely correlated with HbA1c levels and reduced in hyperglycemic islets. In vitro, functional studies showed that the silencing of Cpne3 in INS-1 cells impaired insulin secretion reduced glucose uptake efficiency and downregulated expression of Ins1, Ins2, Insr and Glut2. Taken together, the provided bioinformatics and functional evidence suggest that CPNE3 is an important regulator for pancreatic β-cell function. This study is the first to link CPNE3 with the pathophysiology of diabetes to the best of our knowledge. Previously, CPNE3 was linked with cancer pathogenesis and metastasis; for example, CPNE3 was shown to up-regulate in breast tumors and glioblastoma 24 and regulate ErbB2-dependent cancer cell motility in breast cancer 19 . Additionally, CPNE3 was reported to be downregulated in patients with acute myocardial infarction 25 . The finding that CPNE3 was highly expressed in human islets among other CPNE genes, while Cpne2 was the highest in rat INS-1 cells is not surprising.
Such expression disparity could be attributed to a homogenous expression pattern within all islet cells or being highly expressed by different cell population in pancreatic islets such as β, α or δ cells as reported on the heterogeneity of islet-cell where certain proteins are expressed in a specific sub-population of β-cells 26 . Also, the variations in the expression pattern of copine genes may ascribed to species-specificity (human vs. rat) or the nature of investigated cells that there are differences between freshly isolated normal cells and immortalized cells 27,28 . Importantly, understanding the difference of species expression specificity is of great importance for selecting the precise validation functional model. www.nature.com/scientificreports/ Expression analysis showed that CPNE3 is reduced in hyperglycemic islets (Fig. 1). However, the finding raises a question on whether the observed reduction is causative for T2D or a consequence of glucotoxicity. A previous study showed no effect of short-term exposure of human islets to high glucose concentrations on CPNE3 expression 29 . Typically, expression change is considered a consequence of glucotoxicity when the change after glucose exposure has a similar direction to diabetic islets. Still, it is early to consider reducing CPNE3 expression in diabetic islets as causative for the pathogenesis of T2D. Moreover, the fact that CPNE3 expression is correlated negatively with HbA1c levels (which is a marker for blood sugar levels) indicates the important role of CPNE3 in glucose metabolism.
As demonstrated, gene silencing of Cpne3 in INS-1 cells dramatically affected insulin secretion without cytotoxic effect or apoptosis (Fig. 3). Although the expression of CPNE3 was associated with cancer pathogenesis and metastasis [13][14][15][16]19,24 , this could indicate that CPNE3 has a differential physiological function/role that depends on the cell or microenvironment. Cpne3-silenced cells exhibited a down-regulation at mRNA and protein levels of genes involved in insulin biosynthesis, Ins1 and Ins2 but not the transcription factors that activate the insulin gene promoter (Pdx1) [18][19][20] . Additionally, expression of Glut2, Gck and Insr were also downregulated in Cpne3silenced cells. Glut2 and Gck are key players in glucose uptake and glucose-sensing machinery in pancreatic β-cells to respond to physiological blood glucose changes 30,31 . The low affinity of GLUT2 permits glucose sensing and controls the circulation's glucose uptake rate. Several reports have shown that defects in glucose-sensing machinery impair insulin secretion, leading to severe hyperglycemia. The present study provides evidence that the decreased expression of GLUT2 was accompanied by reduced glucose uptake efficiency, which is a crucial driving factor for insulin release 32 .
Moreover, insulin signaling in β-cells is essential to maintain proper function 33 . The significant decrease in Insr after Cpne3 silencing could lead to a defect in insulin secretion. Silencing of Insr in pancreatic β-cell has been shown previously to impair insulin secretion 33 . Collectively, these results indicate that CPNE3 has a possible role in insulin sensing by reducing the expression of regulators of glucose sensing and insulin signaling.
Although the mechanism behind the role of CPNE3 in insulin secretion is not clear. A speculated mechanism arises from the fact that CPNE3 contains two Ca 2+ -and phospholipid-binding domains, 'C2 domains' which are found in protein such as protein kinase C (PKC), phospholipase C (PLC) and synaptotagmin 34 . PKC and PLC are well-known for their crucial role in regulating insulin secretion β-cells 27,35 . Synaptotagmins are thought to play roles in exocytosis and membrane trafficking phenomena 36 . Nevertheless, more functional investigations are warranted to dissect the exact mechanism through which CPNE3 affects β-cell function.
In conclusion, our data suggest that CPNE3 is a novel regulator in pancreatic β-cell function. Our data might pave the way to consider CPNE3 as a drug target to develop a new molecular therapy for T2D.

Materials and methods
Microarray and RNA-sequencing from human pancreatic islets. Microarray expression data (GEO, accession number: GSE41762) from human pancreatic islets were retrieved from a publicly available database. Robust Multi-array Analysis methods were used to normalized the raw data as described previously 37 . RNA-seq data with accession number; GSE50398 were obtained from the GEO database. Data normalization was processed as previously described 38 .

Culturing of INS-1 cell line and siRNA transfection.
INS-1 (832/13) cells, a gift from Dr. C. Newgard from Duke University, were maintained in a complete RPMI 1640 medium as previously described 39 . Cells were seeded in a 24-well plate (200,000 cells/well) in Antibiotics free medium overnight. At confluency ~ 60%, cells were transfected with 40 nM of two siRNA sequences against CPNE3 (IDs: S162185 and S162183) (Thermo Fisher Scientific, USA) using 1.0 µl of lipofectamine 3000 transfection reagent in Opti-MEM media (Thermo Fisher) as previously described 39 . A previously defined negative control sequence was used at a final concentration of 40 nM 40 . After 24 h of transfection, the media was replaced with complete RPMI 1640 media with antibiotics and qPCR was performed to assess silencing efficiency. For measurements of insulin content, the total protein was extracted from transfected cells using an M-PER reagent (Thermo Fisher Scientific) and quantified by Pierce BCA protein assay kit (Thermo Fisher Scientific). Diluted protein (1:250) was then used to evaluated insulin content using a rat insulin ELISA kit (Mercodia, Sweden). Finally, insulin content was normalized to the total amount of protein.