Insulin produced in and secreted from pancreatic β cells is the most crucial hormone for the control of glucose homeostasis in the blood circulatory system. Impaired biosynthesis and secretion of insulin leads to somatic damage via hyperglycemia. Therefore, the amount of insulin is tightly regulated in order to maintain a very narrow blood glucose range. The elevation of glucose levels induces calcium-mediated secretion of insulin, which, in turn, promotes the absorption of glucose from the blood into peripheral tissues, thereby reducing blood glucose levels. Intriguingly, insulin also regulates not only its own biosynthesis and secretion, but also β-cell mass, via intracellular signaling cascades, in which Akt is the prime mediator (Kulkarni et al., 1999; Otani et al., 2004; Paris et al., 2003; Withers et al., 1998). In addition to stimulating insulin secretion, glucose also stimulates the biosynthesis of insulin at the transcriptional and post-transcriptional levels (German and Wang, 1994; Giddings et al., 1982; Permutt and Kipnis, 1972). The acute production of insulin after short-term (< 2 h) glucose stimulation mostly occurs by the enhancement of pre-existing insulin mRNA stability and translation, rather than de novo synthesis of preproinsulin mRNA, whereas the subsequent production during long-term (> 2 h) glucose stimulation is mediated by both the former and latter processes (Brunstedt and Chan, 1982; Itoh and Okamoto, 1980; Welsh et al., 1985), thereby ensuring that sufficient insulin mRNA is present to serve as the substrate for rapid insulin biosynthesis under conditions of greater insulin demand. Indeed, insulin mRNA comprises a large proportion (> 30%) of the total mRNA pool in pancreatic β cells and has an exceptionally long half-life (> 24 h; Goodge and Hutton, 2000; Itoh and Okamoto, 1980). This increase in insulin mRNA stability and translation depends on its 5′ and 3′ untranslated regions (UTRs; Wicksteed et al., 2001).

Polypyrimidine tract binding protein 1 (PTBP1), which is also known as heterogeneous nuclear ribonucleoprotein I (hnRNP I), is a ubiquitous RNA-binding protein (RBP) that binds to the pyrimidine-rich region in the 3′ UTR of target mRNAs through four RNA recognition motifs (RRM) and contributes to their stability (Sawicka et al., 2008; Tillmar and Welsh, 2002; Tillmar et al., 2002). It is also known to function in diverse cellular processes, including splicing, polyadenylation, mRNA localization, and translation initiation (Sawicka et al., 2008). In pancreatic β cells, PTBP1 stabilizes insulin mRNA by binding to the pyrimidine-rich region in its 3’ UTR (Tillmar and Welsh, 2002; Tillmar et al., 2002), as also shown for iNOS and PGK2 mRNAs (Pautz et al., 2006; Xu and Hecht, 2007). This binding is increased by glucose stimulation (Tillmar and Welsh, 2002; Tillmar et al., 2002). Although PTBP1 mRNA levels have been reported to increase after glucose stimulation in mouse insulinoma MIN6 cells (Webb et al., 2000), the molecular mechanisms by which glucose regulates PTBP1 expression have not been clearly elucidated. Here, we provide evidence that glucose-stimulated PTBP1 expression is mediated by the insulin receptor (IR) signaling pathway via Akt.

Animals and immunostaining

Male C57BL/6 mice were kept in an environmentally controlled room under a 12-h light-dark cycle and provided with chow and water ad libitum. All animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC) at Ajou University. Pancreata from eight-week-old male C57BL/6 mice and twenty-four-year-old adult rhesus monkeys (Macaca mulatta) were fixed in 4% paraformaldehyde solution overnight, embedded in paraffin, and sectioned onto glass slides. Paraffin sections were deparaffinized in OTTIX PLUS & OTTIX SHAPER (DiaPath, Martinengo BG, Italy) and after an antigen retrieval process, stained with primary antibodies against PTBP1 and insulin. Each bound antibody was detected using immunofluorescent staining with secondary antibodies or using the appropriate Elite ABC horseradish peroxidase (HRP)-diaminobenzidine (DAB) system (Vector Labs, USA). Nuclei were stained with DAPI. Slides were viewed using a Leica DMi8 microscope (Leica, USA). Details of primary antibodies are listed in Supplementary Table S1.

Plasmidd and site-directed mutagenesis

Myc-tagged PTBP1 plasmid was obtained from Addgene (USA). Human IR cDNA was amplified by RT-PCR from human pancreas RNA (Stratagene, USA) using an oligo-dT (18 bp) primer for reverse transcription. IR cDNA was incorporated into a 3 × Flag vector. An IR mutant (IR-3YA), in which Tyr1158/1162/1163 residues are substituted with Ala, was generated from wild-type IR (IR-WT) using a QuikChange II XL site-directed mutagenesis kit (Stratagene).

Cell culture, transfection, treatment, and insulin secretion

Mouse insulinoma βTC6 and human hepatocellular carcinoma Hep3B cells were cultured in high-glucose DMEM (Invitrogen, USA) supplemented with 10% FBS (HyClone, USA). βIRWT and βIRKO cells were established from wild-type (βIRWT) or β-cell-specific insulin receptor-deficient (βIRKO) mice (Assmann et al., 2009; Kim et al., 2011; 2012; Kulkarni et al., 1999; Lee et al., 2012). βIRWT and βIRKO cells were cultured in high-glucose DMEM (Invitrogen) supplemented with 10% FBS (HyClone). Transfection of siRNAs and plasmids was carried out using Lipofectamine RNAiMAX or 2000 (Invitrogen), according to the manufacturer’s instructions. Scrambled siRNA (Silencer Negative Control #1; Ambion, USA) or empty vector were transfected as negative controls. Details of the siRNAs are listed in Supplementary Table S2. For glucose, insulin, and IGF-1 treatment, cells were starved for 12 h in DMEM containing 5 mM glucose and 0.1% FBS and further incubated for 12 h with the indicated concentrations of glucose, insulin, or IGF-1. For insulin secretion assay, cells starved for 12 h in DMEM containing 5 mM glucose without FBS were washed three times in PBS and subsequently stimulated with the indicated concentration of glucose for 20 min. At the end of the experiment, the buffer was collected, centrifuged to remove cellular debris and saved for quantification of insulin. The insulin levels were measured using a Mouse Insulin ELISA Kit (ALPCO, USA) according to the manufacturer’s instructions. Transfections of the expression vectors and siRNA (Santa Cruz) for PTBP1 were carried out 24 and 48 h before treating with glucose using Lipofectamine 2000 and RNAiMAX (Invitrogen), respectively.

Western blot analysis

Whole-cell lysates prepared using RIPA buffer (10 mM Tris-HCl [pH 7.4], 150 mM NaCl, 1% NP-40, 1 mM EDTA, and 0.1% SDS) were separated by SDS-PAGE, transferred onto polyvinylidene difluoride (PVDF) membranes (Millipore, USA), and incubated with primary antibodies to detect PTBP1, proinsulin, IRβ, p-IR (Tyr1162/1163), Akt, or β-actin. Membranes were then incubated with the corresponding HRP-conjugated secondary antibodies (Santa Cruz Biotechnology, USA), which were detected using enhanced luminescence (GE Healthcare, UK). Sources and working dilutions of primary antibodies are listed in Supplementary Table S1.

Quantitative real-time PCR analysis (RT-qPCR)

Total RNAs were isolated from βTC6, βIRWT, and βIRKO cells using the TRIzol reagent (Thermo Fisher Scientific, USA), according to the manufacturer’s instructions. Quantitative real-time PCR (RT-qPCR) analysis was performed as previously described (Lee et al., 2012). A list of primer sequences is shown in Supplementary Table S3.

Cell synchronization and flow cytometry

Cells were synchronized by double thymidine block for 24 h and then released from growth arrest by adding fresh medium. For cell cycle analysis, cells were collected by trypsinization, washed with phosphate-buffered saline (PBS), and fixed with 70% ethanol. Fixed cells were washed with PBS and stained with propidium iodide solution. DNA content was measured by flow cytometry and cell cycle profiles were analyzed using Cell Quest and MOD Fit software (BD Bioscience, USA).

Statistical analysis

All values are expressed as the mean ± SEM. Differences between mean values were compared statistically by Student’s t-test. All statistical analyses were performed using GraphPad Prism (GraphPad Software). A P value < 0.05 was considered statistically significant.

Treatment with glucose or insulin increases PTBP1 levels in pancreatic β-cell lines

Although the presence of PTBP1 in insulin-producing mouse β-cell lines and isolated islets is well known (Knoch et al., 2004; 2006; Tillmar and Welsh, 2002; Tillmar et al., 2002; Webb et al., 2000), its expression pattern in vivo has not been firmly evaluated. We performed immunostaining analysis for PTBP1 in pancreatic sections from mice and monkeys after overnight fasting. We found that PTBP1 appeared to be present throughout the pancreas, including endocrine (islets) and exocrine tissues, but its subcellular localization differed between cell types. It was predominantly found in the nucleus of β cells, but in the cytoplasm of exocrine cells (Figs. 1A and 1B). This is in agreement with the notion that glucose induces the translocation of PTBP1 from the nucleus to the cytoplasm in insulinoma INS-1 cells and isolated rat islets and cytosolic PTBP1, in turn, prevents the degradation of insulin mRNA through binding to the pyrimidine-rich region in its 3′ UTR (Knoch et al., 2004). These results also suggest that the nucleocytoplasmic translocation of PTBP1 in β cells is more glucose-inducible than it is in the exocrine cells. Consistent with previous reports (Knoch et al., 2004; Tillmar and Welsh, 2002; Tillmar et al., 2002), silencing PTBP1 in immortalized β cells isolated from the pancreas of wild-type (βIRWT) mice (Assmann et al., 2009; Kim et al., 2011; 2012; Lee et al., 2012) lowered both mRNA and protein levels of proinsulin (Fig. 1C and Supplementary Fig. S1) and consistently reduced glucose-stimulated insulin secretion (GSIS) from mouse insulinoma βTC6 cells (Fig. 1D). Conversely, ectopic expression of PTBP1 increased proinsulin levels (Fig. 1E) and enhanced GSIS from βTC6 cells (Fig. 1F).

As glucose-induced up-regulation of PTBP1 mRNA levels in MIN6 cells has been reported (Webb et al., 2000), we evaluated mRNA and protein levels of PTBP1 after glucose stimulation and observed that glucose dose-dependently increased mRNA and protein levels of PTBP1 in βTC6 cells (Figs. 2A and 2B). Moreover, mRNA and protein levels of PTBP1 were also increased by insulin in a dose-dependent manner (Figs. 2C and 2D), but not by insulin-like growth factor 1 (IGF-1, Supplementary Fig. S2). Given that glucose stimulates insulin secretion, these results indicate that glucose-induced expression of PTBP1 is mediated, at least in part, by insulin released into the extracellular space.

Glucose-induced expression of PTBP1 in pancreatic β cells is mediated by the IR signaling pathway

To investigate the potential role of insulin in mediating glucose-induced PTBP1 expression, we employed βIRWT and βIRKO cells established from WT and β-cell-specific insulin receptor (IR)-knockout mice, respectively (Assmann et al., 2009; Kim et al., 2011; 2012; Kulkarni, et al., 1999; Lee et al., 2012). Similar to the results in βTC6 cells (Fig. 2B), PTBP1 levels were dose dependently increased by glucose in βIRWT cells (Fig. 3A). By contrast, glucose-induced PTBP1 expression was not observed in βIRKO cells, where PTBP1 levels were significantly lower than that in βIRWT cells (Fig. 3A). Likewise, treatment with insulin elevated PTBP1 levels in βIRWT cells, but not in βIRKO cells (Fig. 3B). Neither insulin nor IGF-1 induced an increase in PTBP1 levels in human hepatocellular carcinoma Hep3B cells (Supplementary Figs. S3A and S3B). These results suggest that insulin has a role in mediating glucose-induced PTBP1 expression in β cells.

Next, we further confirmed the involvement of IR and its downstream molecules in insulin-induced PTBP1 expression. Consistent with the lower mRNA and protein levels of PTBP1 in βIRKO cells compared with βIRWT cells (Figs. 4A and 4B), silencing IR in βTC6 cells significantly lowered PTBP1 levels (Fig. 4C). Insulin binding to IR induces autophosphorylation at Tyr1158, Tyr1162, and Tyr1163 in the activation loop of the β subunits (IRβ), which increases the catalytic activity of the receptor and, in turn, transduces the signal to downstream signaling proteins (Hubbard, 1997; Tang et al., 2017; Taniguchi et al., 2006), including Akt, which is a key mediator of insulin responses such as gene expression, protein synthesis, proliferation, and glucose metabolism (Taniguchi et al., 2006). Mutation of these autophosphorylation sites reduces insulin-stimulated kinase activity and causes a parallel loss of biological function (Kim et al., 2011; 2012). Thus, we directly confirmed the effects of IR activity on PTBP1 expression by ectopic re-expression of Flag-tagged IR-WT and IR mutant (IR-3YA), in which the Tyr1158/1162/1163 residues are substituted with Ala (Fig. 4D), in βIRKO cells. Ectopic IR-WT re-expression in βIRKO cells restored PTBP1 levels, but in IR-3YA-transfected βIRKO cells, PTBP1 levels were not altered (Fig. 4E). IR autophosphorylation at Tyr1158, Tyr1162, and Tyr1163 was only detected in IR-WT-transfected cells, not in IR-3YA-transfected cells, likely due to endogenous insulin secretion (Fig. 4E). In addition, silencing Akt in βTC6 cells significantly lowered PTBP1 levels (Fig. 4F). Together, these results indicate that PTBP1 is expressed in β cells under the control of the IR → Akt → PTBP1 pathway (Fig. 4G).

PTBP1 expression correlates with insulin levels during cell cycle progression

Because insulin is a key regulator of β-cell proliferation, which is a major mechanism that contributes to maintaining adult β-cell mass (Assmann et al., 2009; Dor et al., 2004; Folli et al., 2011; Georgia and Bhushan, 2004; Kim et al., 2012; Kulkarni et al., 1999; Otani et al., 2004; Saltiel and Kahn, 2001; Withers et al., 1998), we investigated whether insulin and PTBP1 levels are correlated during cell cycle progression. To address this question, βIRWT and Hep3B cells were initially synchronized at G1/S phase by double thymidine block, after which they were released from growth arrest by adding fresh medium. During the cell cycle progression of βIRWT cells, PTBP1 expression decreased and was positively correlated with proinsulin levels (Supplementary Fig. S4). However, this change in PTBP1 levels was absent in Hep3B cells (Supplementary Fig. S3C).

Here, we provide evidence that glucose induces PTBP1 expression and this is mediated by the IR signaling pathway via Akt (Fig. 4G). In addition to stimulating insulin secretion, glucose induces the rapid biosynthesis of insulin at the transcriptional and post-transcriptional levels (German and Wang, 1994; Giddings et al., 1982; Permutt and Kipnis, 1972) and the latter includes the enhancement of pre-existing insulin mRNA stability and translation (Brunstedt and Chan, 1982; Itoh and Okamoto, 1980; Welsh et al., 1985). In response to glucose, PTBP1 binds to the pyrimidine-rich region in the 3′ UTR of insulin mRNA, leading to its stabilization. It is likely that these posttranscriptional mechanisms mediated by PTBP1 are responsible for the acute increase in insulin biosynthesis in the 2 h following glucose stimulation (Itoh and Okamoto, 1980), while the subsequent production of insulin is mediated by an increase in PTBP1 transcription due to glucose stimulation. Taken together, our findings provide an important insight into the mechanism underlying glucose-induced PTBP1 expression in β cells. It will be important in future experiments to determine downstream molecules that is activated by the IR signaling pathway via Akt and directly participates in PTBP1 transcription.