Search

Glucagon-like peptide-2 (GLP-2) regulates proliferative and cytoprotective pathways in the intestine; however GLP-2 receptor (GLP-2R) signal transduction remains poorly understood, and cell lines that express the endogenous GLP-2R have not yet been isolated. We have now identified several expressed sequence tags from human cervical carcinoma cDNA libraries that correspond to GLP-2R nucleotide sequences. GLP-2R mRNA transcripts were detected by RT-PCR in two human cervical carcinoma cell lines, including HeLa cells. GLP-2 increased cAMP accumulation and activated ERK1/2 in HeLa cells transiently expressing the cloned human HeLa cell GLP-2R cDNA. However, the GLP-2R-induced activation of ERK1/2 was not mediated through Galphas, adenylyl cyclase, or transactivation of the epidermal growth factor receptor, but was pertussis toxin sensitive, inhibited by dominant negative Ras, and dependent on betagamma-subunits. GLP-2 also induced a significant increase in bromodeoxyuridine incorporation that was blocked by dominant negative Ras. Furthermore, GLP-2 inhibited HeLa cell apoptosis induced by LY294002 in a protein kinase A-dependent, but ERK-independent, manner. These findings demonstrate that the HeLa cell GLP-2R differentially signals through both Galphas/cAMP- and Gi/Go-dependent pathways, illustrating for the first time that the GLP-2R is capable of coupling to multiple heterotrimeric G proteins defining distinct GLP-2R-dependent biological actions.

Peptide hormones are secreted from endocrine cells and neurons and exert their actions through activation of G protein-coupled receptors to regulate a diverse number of physiological systems including control of energy homeostasis, gastrointestinal motility, neuroendocrine circuits, and hormone secretion. The glucagon-like peptides, GLP-1 and GLP-2 are prototype peptide hormones released from gut endocrine cells in response to nutrient ingestion that regulate not only energy absorption and disposal, but also cell proliferation and survival. GLP-1 expands islet mass by stimulating pancreatic beta-cell proliferation and induction of islet neogenesis. GLP-1 also promotes cell differentiation, from exocrine cells or immature islet progenitors, toward a more differentiated beta-cell phenotype. GLP-2 stimulates cell proliferation in the gastrointestinal mucosa, leading to expansion of the normal mucosal epithelium, or attenuation of intestinal injury in experimental models of intestinal disease. Both GLP-1 and GLP-2 exert antiapoptotic actions in vivo, resulting in preservation of beta-cell mass and gut epithelium, respectively. Furthermore, GLP-1 and GLP-2 promote direct resistance to apoptosis in cells expressing GLP-1 or GLP-2 receptors. Moreover, an increasing number of structurally related peptide hormones and neuropeptides exert cytoprotective effects through G protein-coupled receptor activation in diverse cell types. Hence, peptide hormones, as exemplified by GLP-1 and GLP-2, may prove to be useful adjunctive tools for enhancement of cell differentiation, tissue regeneration, and cytoprotection for the treatment of human disease.

Tissue-specific proglucagon gene transcription is achieved through combinations of transcription factors expressed in pancreatic A cells and enteroendocrine L cells of the small and large intestine. Cell transfection and electrophoretic mobility shift assay experiments previously identified Pax-2 as a regulator of islet proglucagon gene expression. We examined whether Pax-2 regulates gut proglucagon gene expression using enteroendocrine cell lines and Pax2(1NEU) mutant mice. Immunoreactive Pax-2 was detected in STC-1 enteroendocrine cells, and Pax-2 activated proglucagon promoter activity in transfected baby hamster kidney and GLUTag cells. Pax-2 antisera diminished the formation of a Pax-2-G3 complex in electrophoretic mobility shift assay studies using nuclear extracts from islet and enteroendocrine cell lines. Surprisingly, Pax-2 mRNA transcripts were not detected by RT-PCR in RNA isolated from adult rat pancreas, rat islets, embryonic d 19 or adult murine pancreas and gastrointestinal tract. Furthermore, embryonic d 19 or neonatal d 1 Pax2(1NEU) mice exhibited normal islet A cells and gut endocrine L cells, and no decrement in pancreatic or intestinal glucagon gene expression. These findings demonstrate that Pax-2 is not essential for the developmental formation of islet A or gut L cells and does not play a role in the physiological control of proglucagon gene expression in vivo.

The primary function of islet A cells is the synthesis and secretion of glucagon, an essential hormonal regulator of glucose homeostasis. The proglucagon gene is also expressed in enteroendocrine L cells of the intestinal epithelium, which produce glucagon-like peptide 1 (GLP-1) and glucagon-like peptide 2 (GLP-2), regulators of insulin secretion and intestinal growth, respectively. We show here that Pax6, a critical determinant of islet cell development and proglucagon gene expression in islet A cells, is also essential for glucagon gene transcription in the small and large intestine. Pax6 is expressed in enteroendocrine cells, binds to the G1 and G3 elements in the proglucagon promoter, and activates proglucagon gene transcription. The dominant negative Pax6 allele, SEYNeu, represses proglucagon gene transcription in enteroendocrine cells. Mice homozygous for the SEYNeu mutation exhibit markedly reduced levels of proglucagon mRNA transcripts in the small and large intestine, and GLP-1 or GLP-2-immunopositive enteroendocrine cells were not detected in the intestinal mucosa. These findings implicate an essential role for Pax6 in the development and function of glucagon-producing cells in both pancreatic and intestinal endodermal lineages.

The proglucagon gene is expressed in a highly cell-specific manner in islet and enteroendocrine cells. DNA sequences within the proximal proglucagon G1 promoter region bind the homeobox protein cdx-2/3, and cdx-2/3 activates the proglucagon promoter in fibroblasts. We show here that cdx-2/3 activates the proglucagon promoter in both islet (InR1-G9) and enteroendocrine (STC-1 and GLUTag) cell lines. Furthermore, transfected cdx-2/3 increased the levels of endogenous proglucagon mRNA transcripts in both transient and stable transfections of InR1-G9 islet cells. The cdx-2/3-dependent induction of endogenous proglucagon mRNA transcripts in stable islet lines was associated with a corresponding increase in the transcriptional activity of proglucagon promoter-luciferase plasmids. An amino-terminally truncated cdx-2/3 derivative containing the homeodomain and carboxy-terminal region of the molecule inhibited both the cdx-2/3 activation of the proglucagon promoter and the induction of endogenous proglucagon mRNA transcripts. These observations demonstrate that cdx-2/3, acting through the proximal G1 element, is a major transcriptional determinant of cell-specific proglucagon gene expression in pancreatic islet cells.

The amylin (IAPP) and insulin genes are coexpressed in the pancreatic beta-cell and share related promoter elements that may bind similar islet transcription factors. The observation that these promoter elements contain AT-rich subdomains suggests that homeobox proteins may be important for the regulation of both insulin and amylin gene transcription. We show here that the LIM domain homeobox protein isl-1 activates the rat amylin promoter in both fibroblast and islet cell lines. Mutation of the rAMY promoter TAAT motifs was associated with a marked reduction in both basal and isl-1 -dependent transcriptional activity. The isl-1 homeodomain binds to the AT-rich AMY element (-156 to -137) in the human amylin (hAMY) gene promoter, and electrophoretic mobility shift assay experiments using isl-1 specific antiserum detected the formation of an hAMY-isl-1 complex using nuclear extract from InR1 -G9 islet cells. Although isl-1 binds to both the insulin and amylin gene promoter elements in vitro, these sequences display marked differences in their relative transcriptional properties when ligated adjacent to a heterologous promoter and transfected into InR1 -G9 islet cells. The insulin gene E2 sequence that binds isl-1 (-230 to -208) functions as a negative element, whereas the hAMY sequence activates the thymidine kinase promoter in islet, but not nonislet, cell lines. Transfection of isl-1-depleted isl-1 (AS)InR1 -G9 cell lines demonstrated that the E2 element continued to repress thymidine kinase promoter activity, whereas the positive transcriptional activity mediated by the AMY element was considerably reduced in isl-1 (AS)-InR1-G9 cell lines. These dat2 demonstrate that highly similar elements in islet hormone gene promoters display differential functional properties and support a role for the isl-1 homeodomain protein in the regulation of amylin, but not insulin, gene transcription.

The gene encoding proglucagon is restricted in expression to the central nervous system, endocrine pancreas, and intestine. Transgenic experiments indicate that the proglucagon gene upstream enhancer (GUE) element is a principal determinant of both the tissue specificity and the relative level of proglucagon gene transcription. We have now sequenced the rat proglucagon GUE and analyzed the transcriptional properties of proglucagon-luciferase fusion genes (that contain 5'- and 3'-deletions in the GUE) after transfection of islet (InR1-G9) and enteroendocrine (STC-1 and GLUTag) cell lines. The GUE contains both positive and negative elements that are recognized differentially in islet vs. intestinal cell lines. The transcriptional properties of the GUE sequences were more similar in cell lines of intestinal (STC-1 and GLUTag) compared with islet (InR1-G9) phenotypes. The electrophoretic mobility shift assay was used to identify specific domains of the GUE that interacted with nuclear proteins from islet and intestinal cells. Several GUE sequences recognized proteins present in both fibroblast and endocrine cell lines. In contrast, electrophoretic mobility shift assay experiments also identified 1) GUE-protein complexes common to both islet and intestinal cell lines and 2) GUE-protein complexes specific to either islet or intestinal lineages. One of the GUE subdomains, designated GLUE1, displayed enhancer-like activity in InR1-G9 and GLUTag, but not BHK, cell lines. Taken together, these observations demonstrate that the proglucagon GUE is comprised of multiple positive and negative domains that likely function in a combinatorial fashion to regulate islet and intestinal-specific proglucagon gene transcription.

The human gene encoding PTH-related peptide (hPTHrP) is a complex transcriptional unit that gives rise to multiple messenger RNA (mRNA) transcripts in normal and neoplastic tissues. The utilization of three different transcription start sites and alternative splicing of exons encoding 5'-untranslated sequences accounts for some of the heterogeneity in hPTHrP gene expression; however, the pattern and potential significance of alternative exon splicing at the 3'-end of the hPTHrP gene have not been systematically examined. We have used exon-specific primers and reverse transciptase-polymerase chain reaction to analyze hPTHrP gene expression in nonneoplastic human tissues and human tumors. PTHrP mRNA transcripts encoding PTHrP-(1-139), -(1-173), or -(1-141) were identified in the majority of tissues analyzed. PTHrP mRNAs containing either exon 6 [PTHrP-(1-139)] or exon 8 [PTHrP-(1-141)] sequences were considerably more abundant than mRNAs encoding PTHrP-(1-173) (exon 7) in nonneoplastic tissues. In contrast, a switch in the profile of the different hPTHrP mRNAs was detected in some neoplastic tissues, with an increase in mRNA transcripts encoding PTHrP-(1-173) and a decrease in mRNA transcripts encoding PTHrP-(1-141) observed in several tumors. The differential properties of specific PTHrP 3'-untranslated sequences encoded by unique exons were examined by creating a series of luciferase-PTHrP fusion genes and examining the contribution of PTHrP sequences to relative luciferase activity after transfection of heterologous cell lines. LUC-PTHrP fusion genes containing exon 6 or exon 8 sequences generated consistently higher luciferase activity than fusion genes containing sequences from exon 7. Furthermore, distinct differences in the relative profiles of luciferase activities were observed after transfection of several human, but not rodent, cell lines. These studies suggest that exon splicing at the 3'-end of the hPTHrP gene may be an important determinant of the biological complexity of hPTHrP gene expression.

The gene encoding proglucagon is expressed predominantly in the pancreas and intestine. The physiological importance of glucagon secreted from the islets of Langerhans has engendered considerable interest in the molecular control of proglucagon gene transcription in the endocrine pancreas. In contrast, little is known about the molecular control of proglucagon gene expression in the intestine. The recent demonstration that glucagon-like peptide-1 (GLP-1) secreted from the intestine is a potent regulator of insulin secretion and glucose homeostasis has stimulated renewed interest in the factors that control GLP-1 synthesis in the intestinal L-cell. To develop a model for the analysis of intestinal proglucagon gene expression, we have targeted expression of a proglucagon gene-simian virus-40 large T-antigen fusion gene to enteroendocrine cells in transgenic mice. These mice develop intestinal tumors that were used to derive a novel cell line, designated GLUTag, that expresses the proglucagon gene and secretes immunoreactive GLP-1 in vitro. GLUTag cells demonstrate morphological characteristics of enteroendocrine cells by electron microscopy and are plurihormonal, as shown by immunocytochemistry and RNA analyses. GLUTag cells express the proglucagon and cholecystokinin genes, consistent with the pattern of lineage-specific enteroendocrine differentiation described for mouse intestine. Proglucagon gene expression was induced by activators of the protein kinase-A pathway, and a combination of messenger RNA half-life and nuclear run-on experiments demonstrated that the protein kinase-A-induction is mediated by an increase in proglucagon gene transcription. In contrast, activators of protein kinase-C stimulated secretion, but not biosynthesis of the PGDPs in GLUTag cell cultures. Analysis of proglucagon processing in GLUTag cells demonstrated the liberation of glucagon, oxyntomodulin, glicentin, and multiple forms of GLP-1. These observations provide evidence for the direct induction of proglucagon gene transcription by a cAMP-dependent pathway and suggest that the GLUTag cell line represents a useful model for the analysis of the molecular determinants of enteroendocrine gene expression.

Transgenic mice that express a glucagon gene-simian virus-40 large T-antigen (GLUTag) fusion gene develop neuroendocrine carcinoma of the large bowel. This glucagon-producing tumor was implanted sc and reproducibly formed tumors in nude mice. The transplanted GLUTag tumor expressed large amounts of proglucagon mRNA transcripts, and the levels of proglucagon mRNA transcripts remained constant during 2-8 weeks of tumor growth. The posttranslational processing of proglucagon in the transplantable tumor resembled that detected in the original transgenic tumor, with the liberation of glicentin, oxyntomodulin, glucagon, glucagon-like peptide (1-37) [GLP-1-(1-37)] and GLP-1-(7-37). Tumor-bearing mice demonstrated progressive elevations in the plasma levels of proglucagon-derived peptides. Elevated plasma levels of glucagon-like immunoreactive peptides and immunoreactive glucagon were associated with a marked reduction in the levels of pancreatic glucagon mRNA transcripts by 4 weeks, and after 8 weeks of tumor growth, the levels of glucagon mRNA transcripts in the pancreas were not detectable by Northern blot analysis. Synthesis of the proglucagon-derived peptides was also significantly suppressed at 4-8 weeks in the pancreas of tumor-bearing animals. Histological examination of the endocrine pancreas in mice carrying the GLUTag tumor for 6-8 weeks demonstrated a marked reduction in the number and size of the islets of Langerhans and a disproportionately greater decrease in the number of cells exhibiting glucagon immunoreactivity. By electron microscopy, the residual A-cells were small, compressed at the periphery of the islets, and had poorly developed cytoplasmic organelles. In contrast, no changes in mouse glucagon gene expression or islet morphology were detected in control animals without tumors or mice carrying a sc v-jun-induced fibrosarcoma. The suppression of pancreatic A-cell function and islet size in mice with elevated plasma levels of the proglucagon-derived peptides raises the possibility that a proglucagon-derived peptide may participate in a negative feedback loop, inhibiting expression of the glucagon gene in the A-cells of the endocrine pancreas.

The gene encoding PTH-related peptide (PTHrP) is expressed in a wide variety of normal and neoplastic tissues. Increased PTHrP gene expression in and secretion of PTHrP by specific tumors directly contributes to the development of malignancy-associated hypercalcemia in vivo. To define the genetic elements important for the control of PTHrP gene transcription, we used the reverse transcription polymerase chain reaction to delineate the control of promoter utilization and the splicing patterns of the exons encoding 5'-untranslated sequences. The majority of normal and neoplastic human tissues contained PTHrP mRNA transcripts initiating from both the up-stream (P1) and down-stream (P2) human PTHrP promoters. Furthermore, the downstream promoter was preferentially used by a factor of more than 30-fold. P1-initiated transcripts contained RNA species both with and without exon 2 (E2) sequences, except in the pancreas, adrenal, and stomach, where E2-containing sequences predominated. The transcriptional activities of P1, P2, and P1 + P2 were assessed by transfection of the corresponding PTHrP-chloramphenicol acetyltransferase (CAT) fusion genes into heterologous cell lines. Fusion genes containing P2 sequences were more transcriptionally active than fusion genes containing P1 sequences. The transcriptional activities of P1 + P2 in their natural tandem orientation were additive in rat keratinocytes and human JEG choriocarcinoma cells. In contrast, the activity of P1 + P2 was less than that of P2 alone in hamster BHK fibroblasts and InR1-G9 cells, and human HeLa cells. Analysis of the transcriptional properties of 5'-deleted human PTHrP-CAT constructs revealed the presence of multiple positive and negative DNA sequences (within both P1 and P2) functionally important for human PTHrP gene transcription. Distinct positive and negative DNA elements were also identified from analysis of 5'-deleted rat PTHrP-CAT fusion genes. The results of these experiments provide evidence for cell- and tissue-specific utilization of 1) distinct human PTHrP transcription start sites and specific patterns of 5'-exon splicing and 2) multiple positive and negative DNA control elements, important for the regulation of human and rat PTHrP gene transcription.

The phenotypically distinct cell types of the islets of Langerhans express a number of different hormone genes in a regulated and cell-specific manner. Analysis of the molecular factors important for the control of insulin gene transcription has led to the identification of a number of insulin gene-binding proteins. The cDNAs for several of these proteins have been isolated, including isl-1, a LIM domain homeobox protein reported to be preferentially expressed in islet cells. We now report that isl-1 mRNA transcripts are present not only in islet cells, but also in pheochromocytoma (PC-12) cell lines. Isl-1 mRNA transcripts were also detected in several different regions of rat brain as well as in rat kidney. Analysis of the patterns of gene expression in rat islet cell lines and human pancreatic endocrine tumors demonstrated a variable relationship between the hormonal phenotype of the tumor and expression of the isl-1 and insulin genes. Immunocytochemical studies demonstrated both cytoplasmic and nuclear staining for isl-1 in human pancreatic endocrine tumors, in normal pancreatic A, B, D, and PP cells, and in distal tubular cells of the kidney. Isl-1 immunoreactivity was also widely distributed in neurons in both the central and peripheral nervous system, pituitary, and several human pituitary tumors and pheochromocytomas. The results of these studies implicate a potential role for isl-1 outside the endocrine pancreas in the nervous system and kidney.

The PTH-like peptide (PLP) gene is expressed in a diverse number of normal and neoplastic cell types, and induction of PLP gene expression has been observed after the induction of cellular differentiation. The differentiation of islet cells can be studied in vitro, after the exposure of rat islet cell lines to sodium butyrate. The present work found that rat RIN 1056A islet cells express the PLP gene and treatment with sodium butyrate resulted in rapid (detectable by 30 min) induction of PLP gene expression. PLP gene expression was rapidly and transiently induced by serum and cycloheximide, but the butyrate induction of PLP mRNA transcripts was not dependent on serum or new protein synthesis. Dexamethasone inhibited PLP gene expression and blocked the butyrate induction of PLP mRNA transcripts. The rapid induction of PLP gene expression after the exposure of islet cells to sodium butyrate, serum, and cycloheximide suggests that PLP may be a member of a class of early response genes involved in the regulation of cellular growth or differentiation.

Glucagon and the glucagon-like peptides are encoded within a larger precursor, proglucagon. Transcription of the glucagon gene in pancreas, intestine, and brain gives rise to identical proglucagon mRNA transcripts, after which tissue-specific post-translational processing produces different profiles of proglucagon-derived peptides in each tissue. The importance of glucagon gene 3'-untranslated and 3'-flanking sequences in the control of glucagon mRNA production was studied by transfecting a series of 3'-deleted glucagon genes into fibroblast and islet cell lines. Glucagon genes containing 2 kilobases of 3'-flanking sequences gave rise to accurately processed mRNA transcripts in both baby hamster kidney fibroblasts and InR1-G9 islet cell lines. Deletion of all but 50 basepairs of 3'-flanking sequence had no effect on glucagon mRNA 3'-end formation. In contrast, additional deletion of 3'-flanking and 3'-untranslated sequences resulted in the production of read-through mRNA transcripts with aberrant 3'-ends. The results of these studies define a 50-basepair region in the 3'-flanking sequence of the glucagon gene important for the accurate processing of proglucagon mRNA transcripts.

PTH-like peptide (PLP) is produced by a number of tumors commonly associated with the development of hypercalcemia. Analysis of the expression of the PLP gene has demonstrated that a variety of non-neoplastic endocrine and nonendocrine tissues contain PLP mRNA transcripts. Using a combination of Northern blot analysis, immunohistochemistry, and RIAs, we have demonstrated that the PLP gene is expressed in normal human and rat fetal and adult islets of Langerhans. PLP gene expression was not confined to cells containing a single pancreatic islet hormone, but was found in cells of all four major endocrine subtypes. PLP mRNA transcripts were also detected in RNA prepared from isolated rat islets, and small amounts of PLP immunoreactivity were secreted by cultured rat islets. Fifteen human pancreatic endocrine tumors not associated with hypercalcemia were analyzed and PLP-immunopositive tumor cells were found in 13. These observations demonstrate that the PLP gene is expressed in the normal and neoplastic islets of Langerhans, and suggests a possible role for this peptide in the growth or function of the endocrine pancreas.