Stem cell therapy for diabetes

this month, it was reported that scientists at Harvard University have successfully made insulin-secreting beta cells from human pluripotent stem cells. This is an important milestone towards a “stem cell therapy” for diabetes, which will have huge effects on human medicine.

Diabetes is a group of diseases in which the blood glucose is too high. In type 1 diabetes, the patients have an autoimmune disease that causes destruction of their insulin-producing cells (the beta cells of the pancreas). Insulin is the hormone that enables glucose to enter the cells of the tissues and in its absence the glucose remains in the blood and cannot be used. In type 2 diabetes the beta cells are usually somewhat defective and cannot adapt to the increased demand often associated with age and/or obesity. Despite the availability of insulin for treating diabetes since the 1920s, the disease is still a huge problem. If the level of blood glucose is not perfectly controlled it will cause damage to blood vessels and this eventually leads to various unpleasant complications including heart failure, stroke, kidney failure, blindness, and gangrene of limbs. Apart from the considerable suffering of the affected patients, the costs of dealing with diabetes is a huge financial burden for all health services. The prevalence of type 2 diabetes in particular is rising in most parts of the world and the number of patients is now counted in the hundreds of millions.

To get perfect control of blood glucose, insulin injections will never be quite good enough. The beta cells of the pancreas are specialised to secrete exactly the correct amount of insulin depending on the level of glucose they detect in the blood. At present the only sources of beta cells for transplantation are the pancreases taken from deceased organ donors. However this has enabled a clinical procedure to the introduced called “islet transplantation”. Here, the pancreatic islets (which contain the beta cells) are isolated from one or more donor pancreases and are infused into the liver of the diabetic patient. The liver has a similar blood supply to the pancreas and the procedure to infuse the cells is surgically very simple. The experience of islet transplants has shown that the technique can cure diabetes, at least in the short term. But there are three problems. Firstly the grafts tend to lose activity over a few years and eventually the patients are back on injected insulin. Secondly the grafts require permanent immunosuppression with drugs to avoid rejection by the host, and this can lead to problems. Thirdly, and most importantly, the supply of donor pancreases is very limited and only a tiny fraction of what is really needed.


This background may explain why the production of human beta cells has been a principal objective of stem cell research for many years. If unlimited numbers of beta cells could be produced from somewhere then at least the problem of supply would be solved and transplants could be made available for many more people. Although there are other potential sources, most effort has gone into making beta cells from human pluripotent stem cells (hPSC). These resemble cells of the early embryo: they can be grown without limit in culture, and they can differentiate into most of the cell types found in the body. hPSC comprise embryonic stem cells, made by culturing cells directly from early human embryos; and also “induced pluripotent stem cells” (iPSC), made by introducing selected genes into other cell types to reprogram them to an embryonic state. The procedures for making hPSC into beta cells have been designed based on the knowledge obtained by developmental biologists about how the pancreas and the beta cells arise during normal development of the embryo. This has shown that there are several stages of cell commitment, each controlled by different extracellular signal substances. Mimicking this series of events in culture should, theoretically, yield beta cells in the dish. In reality some art as well as science is required to create useful differentiation protocols. Many labs have been involved in this work but until now the best protocols could only generate immature beta cells, which have a low insulin content and do not secrete insulin when exposed to glucose. The new study has developed a protocol yielding fully functional mature beta cells which have the same insulin content as normal beta cells and which secrete insulin in response to glucose in the same way. These are the critical properties that have so far eluded researchers in this area and are essential for the cells to be useful for transplantation. Also, unlike most previous procedures, the new Harvard method grows the cells as clumps in suspension, which means that it is capable of producing the large number of cells required for human transplants.

These cells can cure diabetes in diabetic mice, but when will they be tried in humans? This will depend on the Food and Drug Administration (FDA) of the USA. The FDA has so far been very cautious about stem cell therapies because they do not want to see cells implanted that will grow without control and become cancerous. One thing they will insist on is extremely good evidence that there are absolutely none of the original pluripotent cells left in the transplant, as they would probably develop into tumours. This highlights the fact that the treatment is not really “stem cell therapy” at all, it is actually “differentiated cell therapy” where the transplanted cells are made from stem cells instead of coming from organ donors. The FDA will also much prefer a delivery method which will enable the cells to be removed, something which is not the case with current islet transplants. One much discussed possibility is “encapsulation” whereby the cells are enclosed in a semipermeable membrane that can let nutrients in and insulin out but will not allow cells to escape. This might also enable the use of immunosuppressive drugs to be avoided, as encapsulation is also intended to provide a barrier against the immune cells of the host.

Stem cell therapy has been hyped for years but with the exception of the long established bone marrow transplant it has not yet delivered. An effective implant which is easy to insert and easy to replace would certainly revolutionize the treatment of diabetes, and given the importance of diabetes worldwide, this in itself can be expected to revolutionize healthcare.