Back

Insulin secretion, islet biology

Three aspects of this topic have drawn my attention: the use in diabetes management of stem cells, the role of the mitochondrium in insulin secretion and sulfonylureas and ATP-dependent potassium channels of the pancreatic ß cell receptors.

Symposium : " Stem cells as a source of insulin-producing cells for transplantation "
Diabetes specialists are showing considerable interest in the potential use of stem cells with the aim of trying to replace pancreatic ß cells destroyed by auto-immune attack in type 1 diabetics. Grafts still raise many problems: for allografts, the lack of donors, and for xenografts the risk of transferring viruses to the recipient rendering it unusable.
Genetic manipulations of ß and non-ß cells have not yielded very notable information.
Use of these stem cells seems potentially useful to try to recreate ß cells and replace the endocrine pancreas in humans.
Three types of cells which are stem cells are found during embryonic development

- Cells already present in the zygote, the very first to appear, which are totipotent, i.e. can give rise to all the different cells of the body.

- Pluripotent embryonic cells, which have already undergone an imprint, a slight differentiation which enables them to go in one direction or in another, e.g. endodermic cells.

- Adult cells, which are multipotent, which are even slightly more differentiated and which cannot be directed with the same degree of differentiation to all cells types.

Can these stem cells be used to recreate a pancreas?
The first thing to have been discussed (there are recent articles for and against) is the possibility of producing adult stem cells, obtained from an adult progenitor, notably using 90% pancreatectomies or pancreatic duct ligation. The possibility of producing normal ß cells from a pancreas which regenerates after 90% ablation, or from pancreatic duct cells, has been described in the literature. These authors feel that there are stem cells in human pancreas which can produce ß cells usable for later transplant. Very recently in the USA, Melton published an extremely provocative article stating that no stem cells existed in adult endocrine pancreas. This left many current projects and hopes in the literature and diabetes community.

Melton says that ß cells are pre-existent, i.e. the process is not the recruitment of stem cells but simply the awakening of pre-existent ß cells.

This view was criticized by Madsen during the symposium, who said that Melton had used in his studies for its demonstration 70% pancreatectomies in the rat or mouse, which do not regenerate in the same way as the 90% type. Hence controversy remains on the fact that it may be possible to recruit stem cells in 90% pancreatectomies, which would not have been noticed by Melton. A major question mark remains regarding adult stem cells, which is important since French legislation only permits the use of these adult stem cells and not embryonic stem cells.

The second lecture by Pedersen, biologist from Greenwich, England, concerned embryonic stem cells and raised two points which I found interesting.

- The first is that it is possible to culture in vitro these embryonic cells taken from an embryo very early during development. These cells show formidable chromosomal stability, without cross over and without transfer of one end of chromosome to another, suggesting that they retain a normal chromosomal structure.

- The other point raised is that after harvesting, these stem cells must be cultured for a certain time so as to proliferate and they are then made to differentiate so as to produce ß cells. The problem here is the epigenetic aspect. You know that in epigenetics, part of the paternal or maternal genome is transferred to the embryo and it has been shown that while there was no problem concerning re-expression of paternal alleles in these stem cells in culture, on the contrary there could be re-expression of the maternal allele giving rise to cells in culture potentially showing abnormalities and unusable later.

The third paper was that of Shimon Efrat of Tel Aviv who uses non-pancreatic stem cells. There are adult and even embryonic stem cells, for example in the liver or bone marrow (the most used, see MeDia News 12, Papers in brief) or the intestine, but according to Melton this does not apply to ß cells. Efrat took non-pancreatic stem cells from mouse embryo liver and introduced a differentiation gene, PDX1, which enables these cells to orientate to a ß cell phenotype (if this gene is invalidated, there is no endocrine pancreas nor ß cell). He is able to convert these liver cells into cells which secrete insulin. The interest is that the liver is a quite large organ, with cells which multiply well and hence it is possible using one liver cell to produce a cell which secretes insulin.

When these differentiated cells are taken from liver cells and are introduced into a mouse which has no rejection immune system, it is possible to correct streptozocin-induced diabetes in the animal concerned. The first problem raised by introduction of a cell manipulated in this way into the body is to know whether it will resist auto-immune attack.

The objective is introduce these new ß cells into a body which attacks them quite vigorously and it was first thought that by taking these stem cells quite early it would be possible to avoid auto-immune attack. The problem is actually not yet solved.

Other problems arise, i.e. whether immunotherapy should be applied to these cells, and finally whether they should be encapsulated. All these extremely delicate approaches of immunotherapy and encapsulation of ß cells have been tried by a large number of colleagues, in particular Gérard Reach at the Hôtel-Dieu Hospital in Paris, and have not led to the creation of an effective pancreas in the long term.

The greatest problem in this type of work is that while transformation of a liver cell into a ß cell which secretes insulin is all very well, at the very least the kinetic aspect of this secretion must be preserved. Efrat was asked this question, and answered that the insulin secretion obtained was not similar to that of a normal ß cell, which raises an important question: would it not be better to treat a diabetic by standard insulin treatment rather than play the sorcerer's apprentice by introducing ß cells manufactured from liver to obtain baseline insulin secretion. I hence feel that this perspective is biologically interesting but not yet of short term usefulness regarding its therapeutic use.

The fourth aspect, developed by Marcus Stoffell of New York, is the possibility of differentiating ß cells from the pancreas, about which he does not agree with Melton. The problem arising is that while these cells are capable of acquiring an adult ß cell phenotype, by expressing all the genes found in a classical ß cell, they express only 1% of the insulin present in a normal cell and they do not secrete insulin.

At the outcome of this symposium, which I found very interesting, and which contained no double-talk, it emerges that these studies, very highly funded by the international JDM in the USA, certainly arrive at very good understanding of the embryonic pancreatic stem cell and trans-differentiation of the liver cell to ß cell, but that in the final analysis it is still far from possible to use this technology in the treatment of type 1 diabetes. It was nevertheless a very good update of the situation.

Role of the mitochondrium in insulin secretion and possibly in type 2 diabetes
The ß cells mitochondrium elaborates the synthesis of ATP which closes an ATP-dependent potassium channel, leading to membrane depolarisation and the secretion of insulin granules. The ß cell is known to possess very special metabolic characteristics. Notably, when glucose is metabolised to pyruvate, one of the characteristics of the ß cell is not to have lactate-dehydrogenase, such that the pyruvate produced can only go into the mitochondrium and it is easy to understand the importance of the mitochondrial oxidation of pyruvate in generating ATP.

One of the speakers showed that ß cell mitochondria are heterogeneous, not all having the same metabolic capacities. Histology and immunohistochemistry techniques have shown the existence of two types of mitochondria

- One functioning like usual mitochondria, i.e. which metabolise pyruvate, produce NADH and synthesise ATP. It is this type which is found in biochemistry books.

- The other which is believed to function in the opposite direction and on the contrary hydrolyse ATP.

When a ß cell is stimulated by glucose, a small number of mitochondria functioning classically are used initially, as in all good biochemistry manuals. The ATP produced is then used for other mitochondria alongside, functioning in the opposite direction, with resultant triggering of an extremely effective signal for producing insulin secretion. This is a very unusual way of looking at things, based upon arguments clearly showing this heterogeneity of mitochondria which seem to be part of a veritable intracellular network enabling the potentialisation of insulin secretion.

Another quite surprising aspect of the question described in this symposium concerns the transcription factors responsible for ß cell differentiation, especially that which I mentioned earlier, when a liver cell is used to produce a ß cell, the PDX1 gene, and another called PAX4. These two transcription factors not only endow an endocrine pancreas cell with the capacity of being a ß cell but also play an important role in controlling mitochondrial function, and hence in controlling the genes inside the mitochondrium.

Two examples
- In MODY 4 diabetes, there is a negative dominant mutation of this PDX1 factor. When this mutated factor is expressed in animals or in cells, this inhibits mitochondrial genes which synthesise NADH, at the origin of the respiratory chain, decreasing the polymerase factor which controls the biosynthesis of RNA of mitochondrial DNA and which encodes the genes of ATP generation
.
This mechanism has shown for this gene described as the MODY 4 gene, that not only does its mutation have repercussions responsible for an identified monogenic disease, but that this involves the intermediary of mitochondria genes and it is very well known that MODY 4 individuals do not secrete insulin well, which explains their diabetes.

- There is another mechanism in PAX4: it increases intramitochondrial calcium, which stimulates dehydrogenases and completely depletes the cytoplasm of ATP. In the absence of ATP in the cytoplasm, it is no longer possible to close the ATP-dependent potassium channel, depolarise the membrane and secrete insulin. You can see the connection which exists between transcription factors and mitochondrial function.

These examples are interesting since the targets of these transcription factors within the mitocondrium are not known and this fits quite well with what is known about relations between mitochondrial diabetes and type 2 diabetes.

ATP-dependent potassium channel and sulfonylurea receptors
The last aspect I found interesting came from a poster presented by the recipient of the Minkowski Prize, G. Rutter of Bristol1, and the results of which were confirmed to me by an eminent and extremely demanding Islets of Langerhans specialist, Jean Claude Henquin.

It has long been known that sulfonylurea receptors and the ATP-dependent potassium channel are not just found in the membrane but also inside the ß cell. In this poster, Rutter shows that this intracellular localisation concerns more than 80% of these receptors. It was presumed for many years that these intracellular channels were located on the mitochondrium, while this presentation shows that this is not the case at all and that these channels, i.e. SUR 1 and Kir 6.2 are inside the insulin secretion granule. It is as if, when these cells are stimulated, as is seen in glucose transporters, there is translocation of an insulin granule involving both the ATP-dependent potassium channel and the sulfonylurea receptor expressed in the membrane, and it is a mechanism which enables activation of the depolarisation system.

Reference
1. Varadi A, Mitchell KJ, Tsuboi T, Schwappach B, Rutter GA. Subcellular localisation of endogenous and overexpressed ATP-sensitive K+ channel subunits SUR1 and Kir6.2 in clonal MIN6 b-cells. (Abstract). Diabetologia 2004;47(Suppl1):A155.