Secretory defects induced by immunosuppressive agents on human pancreatic beta-cells.

Despite the considerable interest for islet and pancreas transplantation, remarkably little is known about the direct effects of immunosuppressive drugs on human beta-cell function. We measured different insulin secretory parameters and insulin gene expression of human islets cultured for 5 days in the presence of mycophenolate mofetil (MMF), cyclosporin A (CsA), tacrolimus (FK506) or a mixture of 3 cytokines. Basal insulin release after exposure to cytokines and FK506 was significantly higher than in control islets. Responsiveness to an acute glucose stimulus did not differ significantly between control and treated islets. However, absolute incremental insulin responses (delta-AUCs) of islets exposed to cytokines or FK506 were significantly higher compared to islets exposed to CsA or MMF, mainly because of the higher basal release. Indeed, maximal over basal release (stimulation index, SI) tended to be lower in islets exposed to FK506 than in control islets. Insulin gene expression was significantly reduced only in islets exposed to CsA. FK506 was, among those tested, the immunosuppressive drug that most profoundly altered the normal insulin secretory pattern of human beta-cells, whereas CsA was the only inhibiting insulin gene expression. Although the abnormalities induced by the immunosoppressive drugs utilized in this study were modest, these in vitro data are consistent with the reported in vivo diabetogenicity of CsA and FK506 and point to MMF as the ideal immunosuppressive agent from a pancreatic beta-cell point of view.

Development and characterization of pituitary GH3 cell clones stably transfected with a human proinsulin cDNA.

Successful beta-cell replacement therapy in insulin-dependent (type I) diabetes is hindered by the scarcity of human donor tissue and by the recurrence of autoimmune destruction of transplanted beta cells. Availability of non-beta cells, capable of releasing insulin and escaping autoimmune recognition, would therefore be important for diabetes cell therapy. We developed rat pituitary GH3 cells stably transfected with a furin-cleavable human proinsulin cDNA linked to the rat PRL promoter. Two clones (InsGH3/clone 1 and 7) were characterized in vitro with regard to basal and stimulated insulin release and proinsulin transgene expression. Mature insulin secretion was obtained in both clones, accounting for about 40% of total released (pro)insulin-like products. Immunocytochemistry of InsGH3 cells showed a cytoplasmic granular insulin staining that colocalized with secretogranin II (SGII) immunoreactivity. InsGH3 cells/clone 7 contained and released in vitro significantly more insulin than clone 1. Secretagogue-stimulated insulin secretion was observed in both InsGH3 clones either under static or dynamic conditions, indicating that insulin was targeted also to the regulated secretory pathway. Proinsulin mRNA levels were elevated in InsGH3 cells, being significantly higher than in betaTC3 cells. Moreover, proinsulin gene expression increased in response to various stimuli, thereby showing the regulation of the transfected gene at the transcriptional level. In conclusion, these data point to InsGH3 cells as a potential beta-cell surrogate even though additional engineering is required to instruct them to release insulin in response to physiologic stimulations.

Insulin-secreting pituitary GH3 cells: a potential beta-cell surrogate for diabetes cell therapy.

In a companion article, we describe the engineering and characterization of pituitary GH3 cell clones stably transfected with a furin-cleavable human insulin cDNA (InsGH3 cells). This article describes the performance of InsGH3 (clones 1 and 7) cell grafts into streptozotocin (STZ)-induced diabetic nude mice. Subcutaneous implantation of 2 x 10(6) InsGH3 cells resulted in the progressive reversal of hyperglycemia and diabetic symptoms, even though the progressive growth of the transplanted cells (clone 7) eventually led to glycemic levels below the normal mouse range. Proinsulin transgene expression was maintained in harvested InsGH3 grafts that, conversely, lose the expression of the prolactin (PRL) gene. Elevated concentrations of circulating mature human insulin were detected in graft recipients, demonstrating that proinsulin processing by InsGH3 cells did occur in vivo. Histologic analysis showed that transplanted InsGH3 grew in forms of encapsulated tumors composed of cells with small cytoplasms weakly stained for the presence of insulin. Conversely, intense insulin immunoreactivity was detected in graft-draining venules. Compared to pancreatic betaTC3 cells, InsGH3 cells showed in vitro a higher rate of replication, an elevate resistance to apoptosis induced by serum deprivation and proinflammatory cytokines, and significantly higher antiapoptotic Bcl-2 protein levels. Moreover, InsGH3 cells were resistant to the streptozotocin toxicity that, in contrast, reduced betaTC3 cell viability to 50-60% of controls. In conclusion, proinsulin gene expression and mature insulin secretion persisted in transplanted InsGH3 cells that reversed hyperglycemia in vivo. InsGH3 cells might represent a potential beta-cell surrogate because they are more resistant than pancreatic beta cells to different apoptotic insults and might therefore be particularly suitable for encapsulation.