Niels Blume

Potent inhibitory effects of transplantable rat glucagonomas and insulinomas on the respective endogenous islet cells are associated with pancreatic apoptosis.

Effects of transplantable rat insulinomas (IN) and glucagonomas (GLU) on the endogenous pancreas were analyzed using morphometry, immunocytochemistry, in situ hybridization, and staining for apoptotic cells. Hyperinsulinemia (IN-rats) and hyper-GLP-1/glucagonemia (GLU-rats) were both associated with marked islet atrophy (67 and 76% of control average planimetrical islet area, respectively). Selective islet B cell inhibition of proinsulin (I and II) genes as well as of expression of the insulin gene transcription factor, IPF1/STF1, was found in IN-rats. Moreover, these islets were characterized by significant B cells apoptosis in the absence of infiltrating lymphocytes. In GLU-rats selective islet A cell inhibition was observed at the level of glucagon mRNA. These islets contained small, highly condensed but clearly active B cells with prominent IPF1/STF1-positive nuclei, surrounded by densely packed glucagon-negative cells with reduced cytoplasm. Furthermore, an active apoptotic process was found exclusively in the exocrine pancreas of GLU-rats. Thus, in IN-rats, islet B cell mass reduction is distinguished by non-immune-mediated programmed cell death, while GLU-rats exhibit A cell mass reduction by cytoplasmic retraction and selective exocrine apoptosis.

Transplantable rat glucagonomas cause acute onset of severe anorexia and adipsia despite highly elevated NPY mRNA levels in the hypothalamic arcuate nucleus.

We have isolated a stable, transplantable, and small glucagonoma (MSL-G-AN) associated with abrupt onset of severe anorexia occurring 2-3 wk after subcutaneous transplantation. Before onset of anorexia, food consumption is comparable to untreated controls. Anorexia is followed by adipsia and weight loss, and progresses rapidly in severity, eventually resulting in reduction of food and water intake of 100 and 80%, respectively. During the anorectic phase, the rats eventually become hypoglycemic and hypothermic. The tumor-associated anorexia shows no sex difference, and is not affected by bilateral abdominal vagotomy, indicating a direct central effect. The adipose satiety factor leptin, known to suppress food intake by reducing hypothalamic neuropeptide Y (NPY) levels, was not found to be expressed by the tumor, and circulating leptin levels were reduced twofold in the anorectic phase. A highly significant increase in hypothalamic (arcuate nucleus) NPY mRNA levels was found in anorectic rats compared with control animals. Since elevated hypothalamic NPY is among the most potent stimulators of feeding and a characteristic of most animal models of hyperphagia, we conclude that the MSL-G-AN glucagonoma releases circulating factor(s) that overrides the hypothalamic NPY-ergic system, thereby eliminating the orexigenic effect of NPY. We hypothesize a possible central role of proglucagon-derived peptides in the observed anorexia.

Islet expression of Rhombotin and Isl-1 suggests cell type specific exposure of LIM-domain epitopes.

The homeodomain protein Isl-1 and the proto-oncogene Rhombotin (a LIM-only protein), share a double zinc-binding LIM domain and have both been implicated in neural and possibly endocrine development. Isl-1 is expressed in all endocrine cell-types of the islet of Langerhans while Rhombotin mRNA expression was reported in rat insulinoma cells. We have cloned and sequenced Rhombotin cDNA from rat insulinoma (99.4% identical to human and mouse sequences) and demonstrate that it is expressed in normal islets, intestinal tissue, and testis, in addition to the brain; but absent in all other organs tested. Rhombotin mRNA is expressed in phenotypically distinct islet tumours (α-, β, and δ-tumours) at levels comparable to that of normal islets. Antisera raised against two distinct epitopes contained within a short synthetic peptide representing part of the N-terminal LIM domain of Rhombotin surprisingly stain α- and δ-cells, respectively, on sections of rat pancreas. Rhombotin is undetectable by immunocytochemistry using LIM-domain antisera on intact monolayer islet tumor cells or transfected fibroblasts while readily detectable when equipped with a FLAG epitope, as detected with FLAG antiserum. In contrast, recombinant FLAG-Rhombotin is efficiently recognised by Western blotting or immunoprecipitation with all LIM-specific antisera. Almost identical results were obtained with LIM-specific versus homeodomain/C-terminal Isl-1 antisera staining α-cell cytoplasm or all islet nuclei, respectively. We conclude that Rhombotin in addition to Isl-1 is expressed in the islet of Langerhans and propose that the differential staining patterns obtained with antisera towards the LIM domains versus flanking epitopes of both proteins reflect (1) cell-specific protein-protein interactions of these domains or, alternatively, (2) islet cell type specific expression of novel homologous LIM domain proteins.

Activation of insulin receptors and IGF-1 receptors in COLO-205 colon cancer xenografts by insulin and insulin analogue X10 does not enhance growth under normo- or hypoglycaemic conditions

Aims/hypothesisRecent studies with normal rats and mouse allograft models have reported that insulin and insulin analogues do not activate the IGF-1 receptor in vivo, and that this characteristic therefore cannot be responsible for the increased incidence of mammary tumours observed for the insulin analogue X10 in chronic toxicity studies with Sprague Dawley rats. This is in clear contrast to reports of insulin and insulin analogues in vitro. Clarification of this is important for understanding the mechanisms behind possible growth-promoting effects of insulin analogues, and will have implications for the development of novel insulin analogues.MethodsWe established a xenograft model in BALB/c nude mice with the human colon cancer cell line COLO-205, which expresses human insulin and IGF-1 receptors, and explored the acute and chronic effects of treatment with supra-pharmacological doses of human insulin, insulin analogue X10 and human IGF-1. With a novel antibody, acute IGF-1 receptor activation was also examined in various tissues from normal rats treated with human insulin, insulin analogue X10 or human IGF-1. Finally, the effects of pharmacologically relevant doses of human insulin and insulin analogue X10 on receptor activation and growth of COLO-205 xenograft were explored in BALB/c nude mice with alloxan-induced hyperglycaemia.ResultsIn normal rats and in BALB/c nude mice bearing a COLO-205 cell xenograft, treatment with supra-pharmacological doses of human insulin, insulin analogue X10 or human IGF-1 resulted in activation of insulin receptors as well as IGF-1 receptors. Treatment of diabetic nude mice with pharmacologically relevant doses of human insulin or insulin analogue X10, which decreased blood glucose from hyperglycaemic levels to the normoglycaemic range, did not increase IGF-1 receptor activation. Furthermore, repeated treatment with supra-pharmacological as well as pharmacological doses of human insulin or insulin analogue X10 did not influence the growth of COLO-205 xenografts.Conclusions/interpretationThis study demonstrates that activation of IGF-1 receptors in cancer cells by insulin and insulin analogues cannot be considered as a purely in vitro phenomenon. It does occur in vivo in animal models, although only after treatment with supra-pharmacological doses. Furthermore, treatment with insulin or insulin analogue X10 did not influence the growth of COLO-205 xenografts under normo- or hypoglycaemic conditions. Further studies are needed before a conclusion can be reached on whether IGF-1 receptor activation by insulin analogues correlates with increased growth in vivo.