| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Other Original Articles |
Cancer Center, University of California (C.D., P.I.-A., B.T., F.L.), San Diego, California 92037; and Ambion, Inc. (L.P.F., R.A.J.), Austin, Texas 78744
Address all correspondence and requests for reprints to: Dr. Fred Levine, University of California Cancer Center, 10901 North Torrey Pines Road, Building Seven, First Floor, La Jolla, California 92037. E-mail: . flevine{at}ucsd.edu
Abstract
Using immortalized human pancreatic endocrine cell lines, we have shown previously that differentiation into hormone-expressing cells requires cell-cell contact acting in synergy with the homeodomain transcription factor pancreatic duodenal homeobox-1 (PDX-1). Although differentiation is associated with a decrease in cell proliferation, the mechanisms behind this relationship are not known. Using TRM-6, a 
cell line, and ßlox5, a ß-cell line, we show here that cell-cell contact and subsequent endocrine differentiation lead to a down-regulation of the c-myc protooncogene. Overexpression of c-Myc obtained with an inducible c-Myc-estrogen receptor fusion protein results in an increase in cell proliferation and the ablation of hormone expression. Moreover, we show that although c-Myc is expressed in a subset of cells from the human fetal and adult pancreas, it is absent in differentiated endocrine cells. The mechanism by which c-Myc interferes with hormone expression may be through effects on the homeodomain transcription factor PDX-1, as immunostaining for PDX-1 in cells with activated c-Myc revealed a redistribution of PDX-1 from the nucleus to the cytoplasm. These results suggest that c-Myc plays a central role in a cell-cell contact-mediated switch mechanism by which cell division vs. differentiation in endocrine cells is determined.
THE DEVELOPMENT of the islets of Langerhans in the mammalian pancreas has been intensively studied as an example of coordinated tissue morphogenesis and because it is hoped that an understanding of this process will facilitate the development of a cell transplantation therapy for diabetes (1, 2). The islets that comprise the endocrine compartment of the pancreas contain four cell types, each producing a distinct hormone. 
-, -, and pancreatic polypeptide (PP) cells secrete glucagon, somatostatin, and PP, respectively. Insulin production is limited to ß-cells, which comprise the majority of the adult islet cells and are key regulators of glucose homeostasis.
It is known that all four types of endocrine cells as well as the pancreatic exocrine cells originate from the same set of epithelial cells that bud off from the early gut endoderm (3). Insulin-expressing cells arise at different stages of embryogenesis in a process thought to be dependent on three pathways controlled by distinct molecular programs (4). A minor pathway occurs early in development in which a small number of insulin-expressing cells coexpress glucagon (5). This is followed by a massive burgeoning of differentiated ß-cells, called the secondary transition (6), and finally by a late proliferation of preexisting ß-cells (7). It has been proposed that the great majority of ß-cells come from the secondary transition and develop from nonhormone-expressing progenitor cells (4, 8, 9).
A cardinal feature of the endocrine pancreas is the inverse relationship between cell division and differentiation, with mature endocrine cells exhibiting a very low proliferative index (10, 11). Although a great deal of effort has been dedicated to studying the transcription factors that control pancreatic morphogenesis and the differentiated phenotype of endocrine cells, especially the ß-cells (4), the regulatory interactions that control the inverse relationship between endocrine cell division and differentiation have not been studied extensively. It has been demonstrated that pancreatic endocrine primary cells and cell lines lose their differentiated phenotype when stimulated to divide in vitro, suggesting that the pathways involved in proliferation and differentiation are inversely linked, but a mechanism by which linkage may occur has not been demonstrated (12, 13).
We have been developing and characterizing cell lines derived from the human endocrine pancreas by infection with a retroviral vector expressing the simian virus 40 (SV40) T antigen and H-rasval12 genes. TRM-6, derived from human fetal islets, differentiates along the -cell lineage upon expression of the transcription factor pancreatic duodenal homeobox-1 (PDX-1) and promotion of cell-cell contact (14). ßlox5, a cell line derived from purified adult ß-cells, differentiates into cells exhibiting glucose-responsive insulin secretion in response to PDX-1, cell-cell contact, and activation of the glucagon-like peptide-1 (GLP-1) receptor (15). With both of these cell lines, differentiation is associated with exit from the cell cycle. Thus, the cell lines provide a model for studying the relationship between endocrine cell proliferation and differentiation.
Many studies have pointed to the importance of the c-myc protooncogene in distinct cellular processes, such as proliferation, differentiation, and apoptosis, and have described its role in tumorigenesis when it is overexpressed (16, 17, 18, 19). Myc is a helix-loop-helix/leucine zipper protein that binds to DNA and recognizes CAC(A/G)TG elements (20, 21). In vivo, c-Myc forms a heterodimeric complex with Max (22) and binds to the E-box enhancer, 5'-CACGTG-3' (20, 23), to activate transcription. It has also been suggested that Myc may have a repressive effect on gene expression, because cells transformed by constitutive expression of Myc exhibit down-regulation of many genes (24). In the endocrine pancreas, the level of c-Myc is increased in settings such as partial pancreatectomy in which proliferation of pancreatic cells occurs (25). In addition, it has been shown to be present in the pancreas of midgestation mouse embryos, mainly in the extension and folding of epithelial cell layers (26). However, whether c-Myc is playing a direct role in those processes is not known.
In this study we show that c-Myc is expressed in human fetal and adult pancreatic tissue, but not in differentiated endocrine cells. Moreover, we demonstrate in our cell lines that the cell cycle exit that occurs upon differentiation of human pancreatic endocrine cells is associated with down-regulation of c-Myc expression. Furthermore, conditional activation of c-Myc using a fusion protein between c-Myc and a hormone-binding domain derived from the estrogen receptor leads to the ablation of hormone expression and an increase in cellular proliferation. Therefore, we believe that c-Myc plays a role in the switch mechanism that controls the inverse relationship between proliferation and differentiation in human pancreatic endocrine cells.
Materials and Methods
Human fetal pancreases at 15–21 wk gestation were provided by the Advanced Bioscience Resources (Alameda, CA). Informed consent for tissue donation was obtained by the procurement center. In addition, approval for the use of human fetal tissue was obtained from the University of California-San Diego institutional review board.
Cells and cell culture
Islet-like cell clusters (ICCs) from the human fetal pancreas were prepared as previously described (27). Cells were maintained in 10% fetal bovine serum/DMEM with 5.5 mM glucose for TRM-6 and ßlox5, and 11 mM for HeLa. Aggregation of cells into three-dimensional clusters and use of the retroviral vector expressing PDX-1 have been described previously (14, 15). Exendin-4 (Sigma, St. Louis, MO) was used at a concentration of 100 nM in PBS. For the c-Myc induction studies, we exposed transduced cells to 100 nM 4-hydroxytamoxifen (4-OHTM; Sigma) for 4 d. Uninduced controls were treated with an equivalent volume of ethanol, the solvent for 4-OHTM. Cell viability was assessed using calcein AM and ethidium homodimer-1 according to protocols supplied with the LIVE/DEAD viability/Cytotoxicity Kit (L-3224, Molecular Probes, Inc., Eugene, OR) that we have used previously (28).
Expression of Myc-estrogen receptor fusion protein (MycER) in ßlox5/PDX-1 and TRM-6/PDX-1 cell lines
The retroviral vector pBABE-puro containing MycER (a gift from Dr. J. Michael Bishop, University of California, San Francisco, CA) was transiently transfected into the 293 GP cell line. Recombinant retrovirus pseudotyped with VSV-G (29) was harvested after 48 h and used to infect ßlox5/PDX-1 and TRM-6/PDX-1 cell lines in the presence of 8 µg/ml Polybrene. Puromycin-resistant clones were selected in the presence of 0.5 µg/ml puromycin (Sigma).
Transfection and gene silencing assays
HeLa S3 cells were transfected with 100 nM short interfering RNAs (siRNAs) specific to the 3'-untranslated region (3'UTR) or a scrambled siRNA (Silencer c-myc siRNA Kit, Ambion, Inc., Austin, TX) or a phosphorothioate antisense oligonucleotide complementary to the start codon of the c-myc mRNA (30). The cells were transfected using oligofectamine (Invitrogen, Carlsbad, CA) according to the manufacturer’s instructions.
TRM-6 and TRM-6/PDX-1 cells were transfected with 150 nM siRNA to c-myc 3'UTR or scrambled siRNA. The transfection was performed with the TransMessenger Transfection Reagent (QIAGEN, Valencia, CA) according to the manufacturer’s instructions. Cells were transfected at a 70–80% confluence in two-well chamber glass slides (Nalge Nunc International, Naperville, IL). The RNA final concentration of 2 µg/well was achieved using tRNA as a carrier together with siRNA to c-myc 3'UTR or scrambled siRNA. To enhance the intensity of somatostatin staining, the secretion blocker monensin (Sigma) was added to cells after the transfection incubation period (3 h) at a 10-µM final concentration in the culture medium. The cells were harvested 24 h after transfection for somatostatin and c-Myc staining.
Cell proliferation assay
Transfected HeLa S3 cells were analyzed using Alamar Blue (BioSource Technologies, Inc., Camarillo, CA) at 24-h intervals. Alamar Blue is a compound that, when reduced by cellular metabolism, changes from a nonfluorescent blue color to a fluorescent red form that is easily quantified. The amount of Alamar Blue reduced is directly proportional to the cell number, providing a rapid method for assessing cell proliferation. To perform the assay, the Alamar Blue reagent was added to the tissue culture medium at a 10% final concentration. The mixture was incubated for 3–6 h under growth conditions, after which fluorescence was quantified using a SpectraMax GeminiXS (Molecular Devices, Sunnyvale, CA).
Western blot
Whole cell extracts (10 µg protein) from monolayer cultures and aggregates of TRM-6 and ßlox5 cells as well as whole cell extracts of the MycER-expressing clones were used. Expression of c-Myc and MycER was determined by Western blot analysis with the human c-Myc-specific monoclonal antibody Myc-9E10 (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) at a dilution of 1:250. The same membrane was reprobed with antiactin rabbit antibody A2066 (Sigma) to verify the protein equivalence when comparing monolayers to aggregates. Western blot for phospho-cAMP response element-binding protein (phospho-CREB; Ser133) was performed using a polyclonal antibody [PhosphoPlusCREB (Ser133) Antibody Kit, Cell Signaling Technology, Beverly, MA] at a dilution of 1:500. Bound antibody was detected with horseradish peroxidase-linked antimouse or antirabbit Ig (Amersham Pharmacia Biotech, Little Chalfont, UK) and enhanced chemiluminescence (Amersham Pharmacia Biotech).
Immunohistochemistry
Immunostaining were performed on 5-µm-thick cryostat sections prepared from snap-frozen fetal pancreases. Sections were mounted on glass slides, dried at room temperature for 1 h, and fixed in freshly made 4% formaldehyde (from paraformaldehyde) for 20 min at 4 C. The human adult pancreas samples [a gift from Dr. Kamen Anachkov (Navy Hospital, Sofia, Bulgaria) and Dr. Sergio Atala Dib (Federal University of Sao Paulo, Sao Paulo, Brazil)] were fixed in formalin and embedded in paraffin. The adult pancreas samples, after paraffin removal and hydration, were permeabilized in 0.1% Triton X-100 in PBS for 15 min and treated for antigen retrieval as previously described (29). The fetal pancreas samples were also permeabilized in 0.3% Triton X-100 in PBS for 15 min. Nonspecific binding was blocked in all samples by incubation of sections in PBS containing 2% donkey serum (DS; Jackson ImmunoResearch Laboratories, Inc., West Grove, PA) and 1% BSA fraction V (Sigma) for 1 h at room temperature. After extensive washes in PBS-DS (0.2% DS, 0.1% BSA, and 5 mM glycine), sections were incubated overnight at 4 C with a mixture of primary antibodies (shown in Table 1
). For the secondary antibodies, the fluorophore-labeled donkey antimouse (Alexa Fluor 488, Molecular Probes, Inc.) or donkey antigoat (Alexa Fluor 488, Molecular Probes, Inc.) and donkey antirabbit (Alexa Fluor 594, Molecular Probes, Inc.) were used at a 1:250 dilution. Nuclear counterstained was performed with 300 nM 4',6-diamidino-2-phenylindole (DAPI; Molecular Probes, Inc.) in PBS. Slides were analyzed with a combined light and fluorescent microscope (ECLIPS E600, Nikon, Melville, NY) equipped with a digital camera (SPOT RT Digital Camera and SPOT RT version 3.0 software, Diagnostic Instruments, Inc., Sterling Heights, MI). For p38 and Ki-67 immunostaining, the peroxidase substrate kit (Vector Laboratories, Inc., Burlingame, CA) was used for color development as previously demonstrated (31). Confocal microscopy studies were performed in a laser scanning confocal microscope MRC 1024 MP (Bio-Rad Laboratories, Inc., Richmond, CA) equipped with kryton/argon laser and Two-photon Ti-Sapphire femtosecond laser system Millenia-Tsunami (Spectra-Physics, Mountain View, CA) to visualize the UV excited fluorophores (DAPI). To quantify c-Myc protein expression levels in HeLa S3 cells using immunofluorescence, we acquired pictures using Adobe Photoshop and quantified using Scion Image (Scion Corp., Frederick, MD).
|
RT-PCR for insulin, somatostatin, glucagon, PP, BETA 2/NeuroD1, Pax 6, and the housekeeping gene porphobilinogen deaminase (PD) has been described previously (14, 15, 32). RT-PCR for cyclin E and cyclin A was performed using the following primers: cyclin E: forward, 5'-tgataatgtggagagggcag-3'; and reverse, 5'-gaggcgtgcgtttgctttta-3'; and cyclin A: forward, 5'-ttggccgtaaggtttgggat-3'; and reverse, 5'-ggttgggaccaagaattcct-3'. PCR amplifications conditions were 5 min at 94 C, followed by 40 cycles of 94 C for 45 sec, 60 C for 45 sec, and 72 C for 45 sec. All of our primers, with the exception of BETA 2/NeuroD1 primers, flank intronic sequence to exclude contamination with genomic DNA. The housekeeping gene PD shows a different band size for genomic DNA, which was used to detect any genomic contamination in PCRs using BETA 2/NeuroD1 primers.
Electrophoretic mobility shift assay analysis (EMSA)
EMSA for PDX-1 and RIPE3b in whole cell extracts was performed using probes derived from the human insulin promoter A5 and C1 elements, respectively (14, 33). EMSA for BETA2/NeuroD1 was performed using a probe derived from the human insulin promoter E2 Box 5'-cagcccccagccatctgccgacc-3' (14, 33).
ELISA
Medium from ßlox5/PDX-1/MycER aggregates that were cultured with exendin for 4 d and exposed, or not exposed, to 4-OHTM was harvested and assayed for insulin by ELISA (Ultrasensitive human insulin ELISA, Mercodia, Uppsala, Sweden).
Results
Cell-cell contact correlates with down-regulation of c-Myc protein expression in ßlox5 and TRM-6 cell lines
Previously we found that induction of -cell differentiation in the human cell line TRM-6 was associated with a decrease in the rate of cell proliferation (14). Both cell-cell contact and the homeodomain transcription PDX-1 factor resulted in decreased proliferation. However, the mechanism by which that inhibition of cell division occurs and whether it is important in the process of differentiation are unknown. Because c-Myc plays a central role in cell cycle control (34) and is regulated by cell-cell contact, we examined its expression in monolayer culture compared with three-dimensional cell aggregates. Western blot analysis of whole cell extracts from monolayer cultures and aggregates revealed greatly decreased c-Myc expression in both TRM-6 and ßlox5 cells expressing PDX-1 when grown as aggregates compared with monolayer cultures (Fig. 1).
|
To test the hypothesis that decreased c-Myc expression in cell aggregates plays a causal role in promoting differentiation by cell-cell contact, we took advantage of the inducible properties of a fusion protein between an estrogen receptor and c-Myc (35). MycER was expressed in TRM-6 and blox5 by retroviral infection, followed by puromycin selection. Selected ßlox5 (Fig. 2) and TRM-6 (not shown) cells expressed MycER, as shown by Western blot analysis. To determine whether the fusion protein was active, the expression of cyclin E and cyclin A was determined because they are known downstream targets of c-Myc (36, 37, 38). The addition of 4-OHTM to cells expressing the fusion protein led to an increase in the expression of cyclin E mRNA in ßlox5/PDX-1/MycER aggregates, as demonstrated by RT-PCR, but did not affect cyclin A expression (Fig. 3).
|
|
Similar to what we have shown previously in the TRM-6 cell line (14), PDX-1 and cell-cell contact caused a decrease in cell proliferation in ßlox5/PDX-1/MycER cells, as shown by immunostaining for the proliferation-associated nuclear antigen Ki-67 (39). In monolayer culture, 77% of ßlox5 cells expressed Ki-67, whereas only 16% of the cells in ßlox5/PDX-1/MycER aggregates were Ki-67 positive (n = 300 cells in each case). Addition of 4-OHTM to the aggregates increased the percentage of Ki-67-positive cells to 57%. In the absence of 4-OHTM, few cells in the aggregates were positive for Ki-67 (Fig. 4A), whereas in the presence of 4-OHTM many cells throughout the aggregates were positive (Fig. 4B). This indicates that c-Myc is playing a causal role in cell-cell contact-induced growth arrest.
|
To determine whether the induction of differentiation in the TRM-6 and ßlox5 cell lines requires down-regulation of c-Myc expression, the effect of inducing c-Myc activity on hormone expression in cells expressing the MycER fusion protein was tested. 4-OHTM or the same amount of ethanol carrier (not shown) was added to TRM-6/PDX-1 and ßlox5/PDX-1 cells at the initiation of the 4-d period of aggregation required for hormone induction. In addition, 100 nM of the GLP-1 agonist exendin-4 was added to ßlox5/PDX-1 cell cultures, as activation of the GLP-1 receptor is necessary for ß-cell differentiation in that cell line (14, 15).
In the absence of 4-OHTM, ßlox5 aggregates expressed high levels of insulin after 4 d of aggregation (Fig. 5A, lane 2), whereas TRM-6 aggregates expressed high levels of somatostatin (Fig. 5B). However, in the presence of 4-OHTM, neither insulin (Fig. 5A, lane 3) nor somatostatin mRNA (Fig. 5B, lane 2) was present in ßlox5 or TRM-6 cells, respectively. Insulin peptide (273 ± 70.5 pmol/liter) was detected by ELISA in supernatant from ßlox5/PDX-1/MycER aggregates in the absence of 4-OHTM, but was undetectable in the presence of 4-OHTM. Data were obtained from four independent experiments. To determine whether c-Myc activation produced ablation of hormone expression as opposed to a switch to a different endocrine lineage, the expression of the other major pancreatic hormones, somatostatin, glucagon, and pancreatic polypeptide, was measured in 4-OHTM-treated ßlox5/PDX-1/MycER cells and was found to be absent (Fig. 5A).
|
Although c-Myc is a potent stimulator of entry into the cell cycle, it can also act to induce apoptosis in vitro and in vivo (40, 41, 42, 43). In a transgenic mouse expressing c-Myc under control of the insulin promoter, prolonged activation of c-Myc in ß-cells led to extensive cell death (43). Both of the cell lines that we are studying express SV40 T antigen and so are disrupted in the p53 pathway that is required for apoptosis induction by c-Myc (45). Furthermore, the level of mRNA for the housekeeping gene PD remained constant with c-Myc activation (Fig. 5A
, line 3, and Fig. 2B, line 2), arguing against the occurrence of extensive cell death. However, to rule out the possibility that overexpression of c-Myc led to cell death, we assessed the viability of the aggregated cells in the presence and absence of 4-OHTM. Both ßlox5/PDX-1/MycER control aggregates and the ones that overexpressed c-Myc were found to be viable (Fig. 6, A and B). Some cells at the center of the control aggregates were dead, as determined by uptake of ethidium homodimer-1 (red). We commonly observe this, particularly with large aggregates, and attribute it to ischemic necrosis. The same analysis was performed on TRM-6/PDX-1/MycER aggregates with the same result (not shown).
|
Induction of insulin gene expression in ßlox5 cells is a complex process that involves changes in the pattern of expression of a large number of genes, particularly transcription factors involved in the establishment and maintenance of ß-cell differentiation (14, 15). To determine whether the effect of c-Myc activation is global, i.e. inhibition of the entire ß-cell differentiation program that is activated in ßlox5, or restricted to a subset of genes that includes the hormones, we studied the expression in ßlox5 cells of transcription factors that are important in ß-cell differentiation. By focusing on genes involved in the control of insulin gene expression, we also hoped to gain insights into the mechanism by which c-Myc inhibits insulin gene expression.
ßLox5 cells express Pax6 and RIPE3b in both the uninduced and induced states. BETA2/NeuroD is present only when the cells are induced. The level of CREB protein increases dramatically with induction of ß-cell differentiation (15). As shown by EMSA, although not quantitative, we found binding of PDX-1, BETA2/NeuroD1, or RIPE3b to the insulin promoter in the samples from ßlox5/PDX-1/MycER aggregates cultured with and without 4-OHTM (Fig. 7A). Moreover, c-Myc activation did not result in a change in the level of mRNA for Pax6 or in the induction of BETA2/NeuroD1 gene expression that occurs when ßlox5 cells are induced to differentiate, as determined by RT-PCR (Fig. 7B). Western blot demonstrated that CREB expression was not affected by c-Myc activation (Fig. 7C). Therefore, the effects of c-Myc appear to be selective for a subset of genes in the ß-cell.
|
PDX-1 plays a central role in the establishment and maintenance of ß-cell differentiation. In both the ßlox5 and TRM-6 cell lines, it is absolutely required for the induction of hormone expression. Therefore, it was a good candidate for being a downstream target for c-Myc in the inhibition of hormone expression. As in the cell lines that we are studying, PDX-1 is introduced exogenously using a retroviral vector and is transcribed from the retroviral long terminal repeat promoter, it was unlikely that c-Myc would be acting at the transcriptional level. Therefore, we investigated PDX-1 cellular localization in ßlox5/PDX-1/MycER cells in the presence and absence of 4-OHTM (Fig. 8, A and B). 4-OHTM-induced activation of c-Myc caused a translocation of PDX-1 from the nucleus to the cytoplasm. However, many cells exposed to 4-OHTM continued to express PDX-1 in the nucleus, and PDX-1 binding to a target DNA sequence could still be detected by EMSA (Fig. 7A). This could be due to the length of time that the cells were exposed to 4-OHTM or to other pathways, such as p38 (reactivating kinase/stress-activated protein kinase-2) that influences PDX-1 localization and function (46).
|
Suppression of endogenous c-Myc induces pancreatic hormone expression
Based in the RNA interference, a phenomenon in which a double-stranded RNA specifically suppresses the expression of a gene bearing its complementary sequence (47), we used an siRNA targeted to the c-Myc 3'UTR. Suppression of c-Myc activity in HeLa S3 cells was observed by a reduction in cell proliferation and protein expression (Fig. 9, A and B). Differences in proliferation rates were first noted 48 h after cells had been transfected with the siRNAs. Figure 9A depicts the relative cell proliferation of HeLa S3 cells transfected with the siRNAs targeted to the c-Myc 3'UTR, the c-Myc scrambled RNA, as well as the positive control, antisense phosphorothioate oligonucleotide and the phosphorothioate scrambled oligonucleotide (30). The reduction in cell proliferation observed with siRNA to 3'UTR was similar to that found using the optimized antisense phosphorothioate oligonucleotide complementary to the start codon of the c-Myc mRNA (30). All of the transfections and cell proliferation assays were reproduced in independent experiments, and the differences in cell proliferation rates were shown to be statistically significant (P < 0.001).
|
, C and D) or with scrambled RNA (Fig. 9, E and F) and stained for somatostatin and c-Myc. We have shown that TRM6 cells requires the combination of cell-cell contact induction and PDX-1 transcription factor to express somatostatin protein (14). TRM6/PDX-1 cells grown in monolayer culture, however, do not show detectable levels of somatostatin protein, but have lower levels of somatostatin mRNA. Here we show that the suppression of endogenous c-Myc in monolayer culture of TRM6/PDX-1 cells induced somatostatin protein expression (Fig. 9, C and D). This result is consistent with a causal role of c-Myc in regulating endocrine differentiation. c-Myc expression in fetal and adult human pancreatic tissue
To extend the relevance of the results with the cell lines to pancreatic growth and differentiation in vivo, we used immunohistochemistry to examine the expression of c-Myc in human fetal and adult pancreatic tissue. Five individual samples of human adult pancreas and 10 samples of fetal pancreases at different stages of development were studied. In adult pancreas a small number of cells stained positively for c-Myc (Fig. 10A), whereas in fetal pancreas (Fig. 10, C, D, E, F, G, and H) and fetal ICCs (Fig. 10B) many positive cells were seen. Interestingly, in all samples the staining was predominantly cytoplasmic, which has been seen by others as well (48). Consistent with the results from the cell lines, no cells that stained positively for hormones including insulin (Fig. 10, A and B), or somatostatin, glucagon, or pancreatic polypeptide-positive cells (Fig. 10D) were positive for c-Myc.
|
, F–H). However, many HNF3ß-positive cells showed colocalization with c-Myc (Fig. 10E). HNF3ß expression has been shown to be expressed in the earlier steps of pancreas differentiation, which could explain its coexpression with c-Myc in less differentiated pancreatic endocrine cells (48). Discussion
The principal conclusion of the studies presented here is that c-Myc plays a central role in controlling the inverse relationship between proliferation and differentiation in human pancreatic endocrine cells. In vivo, there are no examples of primary pancreatic endocrine cells that continuously divide and express hormones. In vitro, stimulating primary human pancreatic endocrine cells to divide using growth factors and extracellular matrix leads to a rapid decline in hormone expression, which can be partially recovered by aggregation of the cells into ICCs in which cell division is inhibited (13, 50). Rodent insulinoma cell lines do divide while expressing hormones, but even in that setting there is a marked tendency for the most differentiated cells to grow slowly and lose hormone expression over time in culture as the cells are unavoidably selected for more rapid proliferation (51). In our system of transformed cell lines derived from primary human pancreatic endocrine cells, neither of the two cell lines that we have characterized extensively produce hormones until almost complete growth arrest has been induced by the combination of PDX-1 expression and cell-cell contact.
There is considerable evidence from other cell types that c-Myc plays an important role in controlling the cellular differentiation state by affecting the expression of key transcription factors (52, 53, 54, 55). For example, enforced expression of v-Myc in primary quail myoblasts was associated with the down-regulation of MyoD and myogenin (56). However, few data from the endocrine pancreas address the effects of c-Myc on cell division and differentiation. Transgenic mice in which the MycER gene is expressed under the control of the insulin promoter exhibit islet hyperplasia when 4-OHTM is administered intermittently or when apoptosis is inhibited by expression of the potent antiapoptotic gene bcl-xL (43). However, prolonged 4-OHTM administration in the absence of a block of apoptosis led to ß-cell loss. It has been shown that sodium butyrate inhibits cellular proliferation and promotes differentiation of rat insulinoma cells in association with a decrease in c-Myc mRNA levels (58). Chronic hyperglycemia induced by partial pancreatectomy has been associated with a decrease in ß-cell differentiation and an increase in c-Myc mRNA (59). Consistent with the negative effects of c-Myc on hormone expression that we observed in vitro, we demonstrated that there is no coexpression of c-Myc and hormones in the human adult or fetal pancreas. With the exception of the early marker of endodermal differentiation HNF3ß, the great majority of cells positive for the transcription factors known to be important in endocrine differentiation did not colocalize with c-Myc in the human fetal pancreas, suggesting that the regulation of c-Myc expression during pancreatic development occurs at a stage before the formation of the mature islet.
Although previous studies had shown an inverse correlation between endocrine differentiation and c-Myc expression, we took advantage of human pancreatic endocrine cell lines to demonstrate a causal role for c-Myc. Conditional activation of c-Myc was sufficient to ablate hormone expression in both - and ß-cell lines. While this manuscript was in revision, a study was published in which it was demonstrated that adenovirus-mediated c-Myc overexpression in normal rat islets suppressed both insulin gene transcription and glucose-stimulated insulin secretion by a mechanism that included inhibition of BETA2/NeuroD1-mediated transcriptional activation, but no changes in BETA2/NeuroD1 expression (60). A caveat to that study is the extremely high level of c-Myc in cells infected with the adenoviral vector. In our study activation of c-Myc resulted in a change in the subcellular localization of PDX-1, and in agreement with the previous study, no change in BETA2/NeuroD1 expression was found. In addition, we did not detect changes in the expression or function of Pax6, CREB, or RIPE3b, but as the downstream targets of those factors are not completely known, direct effects on the promoters of those downstream genes cannot be ascertained at present. As there are certainly multiple downstream targets of c-Myc, it is possible that the interference with endocrine differentiation and hormone gene expression occurs through indirect effects on as yet unknown downstream c-Myc targets as well as direct effects on hormone promoters.
It has been demonstrated that the control of PDX-1 localization is under the influence of different signaling pathways. Glucose stimulation has been shown to promote translocation of PDX-1 from the cytoplasm to the nucleus in human islets via the p38 pathway (46). However, in our study p38 activation was unaffected by c-Myc in ßlox5 (data not shown). In MIN6 insulinoma cells, glucose-stimulated translocation of PDX-1 from the nuclear periphery to the nucleoplasm required PI3K activation (61). It has been demonstrated that PDX-1 shuttles between the nuclear periphery and nucleoplasm in response to changes in glucose and insulin concentrations and that these events are dependent on phosphoinositol 3-kinase, stress-activated protein kinase-2/p38, and a nuclear phosphatase(s) (62). GLP-1 was shown to regulate PDX-1 through activation of adenylyl cyclase. The increase in intracellular cAMP activates protein kinase A, which ultimately leads to increases in PDX-1 protein levels and translocation of the protein to the nuclei of ß-cells (63). In one report c-Myc has been shown to induce a change in the localization of a transcription factor by direct binding and sequestration (64). Determining the pathway by which c-Myc affects PDX-1 localization may provide new insights into the mechanism by which c-Myc controls cell division and differentiation in endocrine cells.
c-Myc has been found to be expressed in a variety of normal tissues (65, 66, 67, 68), including the pancreas of midgestation mouse embryos (26). In some settings, including the adult murine pancreas model described above (43), deregulated c-Myc expression promotes apoptosis (69, 70, 71, 72, 73, 74). In our system no apoptosis was observed. It has been suggested that the c-Myc mediated apoptotic response depends on the p53 pathway via induction of alternative reading frame, which sequesters the p53-degrading protein, Mdm-2 (45). Our cell lines express SV40 large T antigen, which interferes with both p16/Rb and p53/p21 pathways (75, 76). The compromised p53 pathway may attenuate c-Myc apoptotic activity, thereby revealing its effects on endocrine differentiation.
The current study is the first to directly demonstrate that down-regulation of c-Myc is sufficient under some circumstances for endocrine cell differentiation to proceed. The question then arises as to which pathways are controlling c-Myc expression in the endocrine pancreas. In our cell lines cell-cell contact is required for endocrine differentiation and is associated with exit from the cell cycle. Here, we have shown that both of those effects require the down-regulation of c-Myc. Cell-cell contact activates cell adhesion molecules, including cadherins, which can down-regulate c-Myc through the ß-catenin pathway (77). In pancreatic endocrine cell differentiation, E-cadherin has been shown to be important (78), but a direct connection between E-cadherin and c-Myc expression in the pancreas has not been demonstrated. In addition to cadherins, ß-catenin signaling is also affected by the wnt pathway (79). Although proteins that are part of this pathway are expressed in the pancreas (80), there are no data about their function in pancreatic morphogenesis or endocrine differentiation.
The data presented here have implications for the generation of increased numbers of ß-cells for cell transplantation therapies for diabetes. Any attempt to promote proliferation of ß-cells or their precursors in vitro or to stimulate expansion in vivo inevitably involves altering the balance between proliferation and differentiation that is a central feature of almost all cell types. Understanding that relationship in the human endocrine pancreas should facilitate efforts to control cell expansion without irreversible effects on differentiated function, thereby allowing large quantities of cells to be produced for therapeutic purposes.
Acknowledgments
We thank Drs. Kamen Anachkov and Sergio Atala Dib for the samples of human adult pancreas, Dr. Chris Wright for the PDX-1 antibody, and Dr. Michael Bishop for the retroviral vector pBABE-puro containing MycER.
Footnotes
This work was supported by grants from the NIDDK (DK-55065 and DK-55283), the Juvenile Diabetes Foundation (197035), the Stern Foundation, and PanCel (to F.L.) and by the Swedish Academy of Pharmaceutical Sciences and the Swedish Medical Association (to B.T.). FL holds equity in PanCel, Inc.
Abbreviations: CREB, cAMP response element-binding protein; DAPI, 4',6-diamidino-2-phenylindole; DS, donkey serum; EMSA, electrophoretic mobility shift assay analysis; GLP-1, glucagon-like peptide-1; HNF, hepatocyte nuclear factor; ICC, islet-like cell cluster; MycER, Myc-estrogen receptor fusion protein; 4-OHTM, 4-hydroxytamoxifen; PD, porphobilinogen deaminase; PDX-1, pancreatic duodenal homeobox-1; PP, pancreatic polypeptide; siRNA, short interfering RNA; SV40, simian virus 40; UTR, untranslated region.
Received September 26, 2001.
Accepted April 8, 2002.
References
-cell development in a human pancreatic endocrine precursor cell line. Mol Endocrinol 14:814–822This article has been cited by other articles:
 |
|
 |
|
  K. Matsumura, B. H.-J. Chang, M. Fujimiya, W. Chen, R. N. Kulkarni, Y. Eguchi, H. Kimura, H. Kojima, and L. Chan Aquaporin 7 Is a {beta}-Cell Protein and Regulator of Intraislet Glycerol Content and Glycerol Kinase Activity, {beta}-Cell Mass, and Insulin Production and Secretion Mol. Cell. Biol., September 1, 2007; 27(17): 6026 - 6037. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
