Table of Contents 21-3

Cell Replacement for Diabetes

The last decade saw major improvements in the management of type 1 diabetes (T1DM). However, as noted by Cozar-Castellano and Stewart, there are still several obstacles to realizing ideal therapy, which involve the ability of systems to intelligently and appropriately sense glucose and release insulin in response to glucose levels. Pancreas transplants and especially pancreatic islet transplantation offers much promise, but this approach faces 2 primary hurdles. The first is the immune response associated with transplanting cells from another individual or a different species. The second is acquiring adequate cells for transplantation even if the immune problems could be overcome. As these authors pointed out, successful islet cell transplant takes up to 4 cadaver - pancreas donors to treat a single person with T1DM and there are many times more diabetic patients that there are available donors. Sapir et al¹ report a strategy that may obviate these problems; it relies on generating functional insulin-producing cells from adult human liver cells.

Based on evidence that the homeobox gene designated &emdash; pancreatic and duodenal homeobox gene 1 (PDX-1) &emdash; has a central role in regulating pancreatic organogenesis, adult beta-cell function and expression of multiple beta-cell specific genes, that transient expression of PDX-1 in mice induces transdifferentiation of liver cells, and that PDX-1 has the capacity to convert adult liver cells to pancreatic cells in frogs and mice, Sapir et al used PDX-1 to transdifferentiate fetal and adult human liver cells along the endocrine pancreas lineage. The harvested cells were expanded in primary culture and infected with recombinant adenovirus encoding rat PDX-1. About half of adult cells infected with the virus were capable of activating the insulin promoter in a PDX-1 dependent fashion. Expression of 2 other pancreatic genes that are normally silent in hepatic cells - glucagon and somatostatin - was induced by PDX-1. Addition of nicotinamide and EGF, previously shown to promote endocrine pancreatic differentiation, enhanced the induction, although the levels of the 3 peptides were substantially lower than those detected in freshly isolated pancreatic islet cells.

When a battery of pancreatic genes was tested, they too showed increased expression in cells in which PDX-1 was expressed. The battery included neuroendocrine vesicle markers secretogranin 2 and secretory granule neuroendocrine 1, and prohormone convertase 2. These results strongly suggested that PDX-1 induces cellular transdifferentiation of adult liver cells rather than simply expression of a few key genes.

Next, light and electron microscopy were used to document that the transdifferentiated adult liver cells produced and stored insulin in granules that somewhat resembled granules observed in intact islet cells. This observation raised the possibility that the cells were forming the machinery pancreatic cells utilize to store and release insulin in response to glucose. To test this notion, the cells were treated with glucose. The glucose caused an immediate and profound increase in insulin C-peptide secretion, which was not induced by cells treated with a glucose analog that does not elicit an insulin response.

Finally, to determine if the transdifferentiated adult liver cells could function in an in vivo setting, they were implanted under the kidney capsule of immunodeficient mice made diabetic by streptozotocin treatment. The implantation caused a gradual and significant decrease in plasma glucose levels that persisted over the 60 day duration of the experiment. The drop in glucose levels correlated well with a substantial increase in human C-peptide levels. The transplanted animals responded to a glucose load with a clearance rate comparable to that of healthy mice, although the glucose levels were higher. Removal of the transdifferentiated cells by nephrectomy resulted in the return of hyperglycemia and loss of human C-peptide validating that the transplanted cells were the source of insulin in these mice.

The authors concluded that their work demonstrates the potential use of adult liver cells as pancreatic progenitor tissue. They pointed out that their protocol did not lead to induction of pancreatic exocrine genes as investigators in the field had feared based on similar experiments with rodent liver cells.

Sapir T, Shternhall K, Meivar-Levy I, et al. Cell-replacement therapy for diabetes: generating functional insulin-producing tissue from adult human liver cells. Proc Natl Acad Sci. 2005;102:7964-7969.

Editor’s Comment: This is a proof of concept paper. The work needs to be repeated by others and other issues such as how long the therapeutic effect can be sustained and what problems arise in the long term need to be resolved. The editorial by Cozar-Castellano and Stewart addresses these issues well. Nevertheless, the results are encouraging since the strategy potentially circumvents the 2 major hurdles mentioned above: immune response and inadequate cell numbers. This is because the liver cells would come from the patient him/herself and they could be expanded during primary cell culture to insure adequate cell number not only for initial treatment but for keeping the beta-cell tank full as needed (Figure).

William A. Horton, MD

Reference - (linked to )

1. Cozar-Castellano I, Stewart AF. Molecular engineering human hepatocytes into pancreatic beta cells for diabetes therapy. Proc Natl Acad Sci. 2005;102:7781-7782.