cMyc is a principal upstream driver of beta-cell proliferation in rat insulinoma cell lines and is an effective mediator of human beta-cell replication

Abstract

Adult human β-cells replicate slowly. Also, despite the abundance of rodent β-cell lines, there are no human β-cell lines for diabetes research or therapy. Prior studies in four commonly studied rodent β-cell lines revealed that all four lines displayed an unusual, but strongly reproducible, cell cycle signature: an increase in seven G(1)/S molecules, i.e. cyclins A, D3, and E, and cdk1, -2, -4, and -6. Here, we explore the upstream mechanism(s) that drive these cell cycle changes. Using biochemical, pharmacological and molecular approaches, we surveyed potential upstream mitogenic signaling pathways in Ins 1 and RIN cells. We used both underexpression and overexpression to assess effects on rat and human β-cell proliferation, survival and cell cycle control. Our results indicate that cMyc is: 1) uniquely up-regulated among other candidates; 2) principally responsible for the increase in the seven G(1)/S molecules; and, 3) largely responsible for proliferation in rat β-cell lines. Importantly, cMyc expression in β-cell lines, although some 5- to 7-fold higher than normal rat β-cells, is far below the levels (75- to 150-fold) previously associated with β-cell death and dedifferentiation. Notably, modest overexpression of cMyc is able to drive proliferation without cell death in normal rat and human β-cells. We conclude that cMyc is an important driver of replication in the two most commonly employed rat β-cell lines. These studies reverse the current paradigm in which cMyc overexpression is inevitably associated with β-cell death and dedifferentiation. The cMyc pathway provides potential approaches, targets, and tools for driving and sustaining human β-cell replication.

Fig. 1.

Expression of the three early G₁/S molecules induces the four downstream G₁/S molecules. The goal of these studies was to determine whether the three early G₁/S molecules can induce the expression of, i.e. are upstream of, the four late G₁/S molecules. Top panel, Isolated rat islets were transduced with Ad.lacZ (750 moi) or the combination of Ad.cdk4, Ad.cdk6, and Ad.cyclin D3 (Combo) in a total of 750 moi, 250 moi of each, such that they were expressed at levels, assessed by immunoblot, comparable to those in Ins1 cells, as required for interpreting the three panels below representative of three experiments. Bottom three rows, In response to expression of cdk6, cdk4, and cyclin D3 at levels observed in Ins1 cells shown in the top panel, cdk1, cdk2, and cyclins A and E also were induced in rat islets in levels approximating those in Ins1 cells, as assessed by immunoblot (top row), and quantified by densitometry (middle row). qRT-PCR (bottom row) shows that mRNAs encoding cdk1, cdk2, and cyclins A and E are also induced by the combination of cdk4, cdk6, and cyclin D3, again qualitatively mimicking events in Ins1 cells (n = 3–4 for each panel). Error bars indicate sem, and asterisks indicate P < 0.05. These results indicate that the three early G₁/S molecules are capable of augmenting expression of the four late G₁/S molecules, as is observed in Ins1 cells.

Fig. 2.

The three early G1/S molecules are increased at the protein, but not the mRNA Level. Top row, To confirm a previous report (9), Ins1 and RIN cells were examined for expression of cdk4, cdk6, and cyclin D3 proteins. As can be seen, these three early G₁/S molecules are elevated in Ins1 and RIN cells, as compared with normal rat islets (Rat), confirming earlier results. Bottom row, To determine whether the three early G₁/S molecules are increased at the mRNA or post-mRNA level, mRNA encoding these three molecules were examined by qRT-PCR. As seen in the lower row, although cdk 4, cdk6, and cyclin D3 proteins are all increased in RIN and Ins1 cells, they are not increased at the mRNA level. This suggests that their up-regulation in Ins1 and RIN cells results from enhanced stability or enhanced translation (n = 4 for each panel). Error bars indicate sem. P values are shown.

Fig. 3.

cMyc is moderately and reproducibly increased in Ins1 and RIN cells at the protein and mRNA levels, with no evidence of gene amplification. A, cMyc protein is detectable but of low abundance in rat islets. In contrast to rat islets and to the 18 molecules shown in Supplemental Fig. 1, it is consistently 5- to 7-fold more abundant in Ins1 and RIN cells. B, cMyc mRNA is also increased by a factor of approximately 14 in Ins1 and RIN cells, as compared with isolated rat islets, determined by qRT-PCR using intron-spanning primers. C, cMyc gene abundance as assessed using quantitative PCR (qPCR) of genomic DNA from rat islets, RIN, and Ins1 cells, using single-exon primers reveals no evidence of gene amplification. In all experiments n = 5–6, and bars indicate sem. P values are shown within the bars.

Fig. 4.

cMyc is required for Ins1 and RIN proliferation, and also for increases in the Seven Signature G₁/S Molecules. A, Silencing of cMyc in Ins1 cells, as demonstrated by dose-related reductions in cMyc immunoblot. “Con siRNA” indicates a nonsense siRNA. “myc siRNA” indicates an siRNA directed against cMyc (ON-TARGETplus SMARTpool siRNA, no. B-004500–100; Dharmacon) administered in the same concentration. B, This results in an approximate 50% reduction in [³H]thymidine incorporation in Ins1 cells. C, To confirm that proliferation in Ins1 and RIN cells is cMyc dependent, 1RH (also called “10058-F4”), a pharmacological inhibitor of cMyc-Max interactions, was added to Ins1 and RIN cells for 24 h at increasing concentrations, from 0–40 μm. [³H]thymidine incorporation declined by approximately 60% in Ins1, and 50% in RIN cells in response to pharmacological inhibition my 1RH. D–J, Associated with a reduction in proliferation, cMyc inhibition with 1RH also leads to a 20–80% reduction in each of the seven signature G₁/S molecules. Bars indicate sem. Asterisks indicate P < 0.05. Each experiment was performed three to five times. These results indicate that cMyc is upstream of the seven signature G₁/S molecules and is also required for full proliferation in Ins1 and RIN cells.

Fig. 5.

cMyc expression up-regulates the seven signature G1/S molecules in rat islets. A, Ad.cMyc driving cMyc under the control of the CMV promoter (“CMV-Myc”) was overexpressed in whole rat islets (700 moi for 1 h) and immunoblots were performed 72 h later. As can be seen, this resulted in cMyc expression at levels comparable to Ins1 cells, and higher than in uninfected (uninf) and Ad.lacZ-transduced rat islets. These same rat islet preparations were used for the experiments in the panels below. B and C, cMyc overexpression at Ins1 levels leads to increases in each of the seven signature G₁/S molecules, as demonstrated by immunoblot with quantitative densitometry shown in the bar graphs. Bars indicate sem. Asterisks indicate P < 0.05. Each experiment was performed three to five times. These results indicate that cMyc is capable of increasing all seven signature G₁/S molecules in rat islets and complement the data in Supplemental Fig. 2 indicating that cMyc is upstream of these molecules.

Fig. 6.

cMyc induces proliferation without inducing cell death in rat β-cells. A–F, Laser confocal imaging of BrdU incorporation into isolated dispersed rat β-cells transduced with no adenovirus, or with Ad.lacZ or with increasing moi of Ad.cMyc for 2 h, then incubated for 72 h, with BrdU labeling for the final 18 h. G–M, TUNEL assay for apoptosis in isolated dispersed rat β-cells handled exactly as in panels A–F, showing a positive control (Pos Ctrl) (deoxyribonuclease treatment) (M), uninfected β-cells (G), β-cells transduced with Ad.lacZ (H) or with increasing moi of Ad.cMyc (I–L) for 2 h, then incubated for 72 h. N, Quantification of BrdU incorporation into isolated rat β-cells. O, Quantification of TUNEL histochemistry. Bars indicate sem. Asterisks indicate P < 0.05. Each experiment was performed three to six times. These results indicate that cMyc can induce rat β-cell replication and can do so in doses that are below the threshold that induces β-cell death.

Fig. 7.

cMyc also induces proliferation in human β-cells without inducing cell death. Human islets were dispersed, plated on coverslips, and transduced with 200 moi Ad.lacZ (A) or Ad.cMyc at 50–200 moi (B–D) for 2 h, cultured for 72 h, and labeled with BrdU for the final 18 h before fixing, and visualized using laser confocal microscopy. Green indicates insulin staining and red indicates BrdU. A minimum of 2000 β-cells were counted for BrdU. TUNEL staining in human islets treated identically as in panels A–D with 200 moi Ad.lacZ (E), 50–200 moi Ad.cMyc (F–H). Red indicates insulin staining, and green indicates TUNEL positivity. A minimum of 900 β-cells were counted for TUNEL. I, Quantification of five experiments with different human islet preparations under conditions identical to panels A–D showing that cMyc expression reproducibly increases human β-cell BrdU incorporation. J, Quantification of five experiments with different human islet preparations showing that cMyc expression at 100 moi has no effect on TUNEL, but that a higher moi reproducibly increases TUNEL staining. K, Glucose-stimulated insulin secretion from three different preparations of control human islets (untransduced) or transduced with Ad.lacZ or Ad.cMyc. Bars in I–K indicate sem, and asterisks indicate P < 0.05. These studies indicate that cMyc can induce human β-cell proliferation, and can do so at doses that do not induce cell death, and allow human islets to retain the ability to sense glucose and secrete insulin.

Fig. 8.

A, A simplified working model for cMyc regulation of the seven G₁/S molecules in rat insulinoma lines. In this working model, cMyc expression in insulinoma is increased by an as yet unknown mechanisms. This leads, also through as yet undefined mechanisms indicated by the “?” symbols, to an increase in the three early G₁/S molecules. These lead to initial phosphorylation of the retinoblastoma protein (pRb) to its partially phosphorylated form (ppRb). This allows derepression of E2F activity, which activates late G₁/S molecule transcription, with resultant hyperphosphorylation of ppRb to pppRb, and further release of E2F repression. This then permits activation of the myriad genes involved in cell cycle progression. The model is linear and sequential, and it is likely that additional features are relevant. For example, direct activation of late G₁/S molecules by cMyc has been reported. Key questions from this model are: 1) What leads to the increase in cMyc?; and 2)How does cMyc increase the early G₁/S molecules at the protein, but not mRNA level? B, A model for the differential dose-responses of β-cell proliferation and death in rat and human β-cells in response to increasing amounts of cMyc. Physiological amounts of cMyc in β-cells are very low, but measurable (Refs. , and Figs. 3 and 5), and increase transiently by approximately 1.5- to 2-fold in response to glucose and partial pancreatectomy, as described by Jonas et al. (21). They are approximately 5- to 7-fold elevated in Ins1 and RIN cells (Figs. 3 and 5), sufficient to activate proliferation without inducing β-cell death (Figs. 6 and 7). Higher levels of expression, e.g. 50- to 150-fold, as discussed in the text, activate both proliferation as well as cell death in rat and human β-cells as shown by Murphy et al. (18), Pelengaris and Khan (19), Jonas et al. (21), Cano et al. (22), Demeterco et al. (23), and Evan and co-workers (24), and shown herein for human β-cells (Figs. 6 and 7).