“I’ve always understood that the job of insulin was to move glucose from the blood into the cell, which is why in the absence of sufficient insulin, glucose ends up hanging out in the blood, picking fights and causing trouble. So what do the inside-the-cell reactions you describe have to do with diabetic complications?”
This question, posed in the comment to my post about sorbitol (the best question of the semester so far), deserves a full answer so let’s see if I can give it.
First of all, you will have had to read the previous two posts to make sense of this post. In the first, I talk about how glucose outside the cell causes problems (Unfettered RAGE). In the second post, I discuss how glucose inside the cell causes problems (The Sorbitol Blues).
Let’s start by considering outside the cell reactions mediated by glucose. Glucose reacts spontaneously with proteins. This reaction is really slow so very few proteins get glycosylated under normal conditions. Furthermore, the proteins in our bodies are the infrastructure of the cell. Like our roads and bridges, these things are constantly getting repaired and replaced. Eventually, a glycosylated protein will get replaced with a new protein, and things remain in balance. If you do not have diabetes and you get your hemoglobin A1c measured it will be some very low number but it will not be zero. A few proteins get screwed up by glycosylation but they get replaced before the function of the tissue is compromised. When glucose levels rise, the rate of glycosylation smoothly rises with the glucose and the amount of glycosylated proteins increases.
Compare this to rust. Think about a cast iron skillet – something I use for cooking all the time and something that gets in contact with water plus air a lot. If I wash the skillet but leave some water in it for a long period of time then rust forms. If, however, I wash the skillet and perhaps an hour later get around to drying it, no rust forms. Why? The oxidation of the iron takes a very long time. Each molecule that reacts requires a specific set of conditions that takes a while to happen so it takes many hours of contact with water before appreciable rust forms.
Hopefully you see the connection between the rust reaction with metal and the glucose reaction with protein. When we eat a meal, glucose levels shoot up in the blood but glycosylation does not appreciably increase because the exposure time is short – on the order of perhaps an hour or so.
Now let’s move inside the cell. Glucose moves into the cell at a rate determined by its relative concentration inside the cell and its relative concentration outside the cell. We call this passive diffusion. I often use the concept of water travelling down a sloped pipe when I describe this to students. The greater the difference in concentrations between outside and inside is sort of like the speed at which the water is flowing – which would be determined by the angle at which the pipe is sloped. Glucose travels through glucose shaped holes called glucose transporters. GLUT4 is the name of the gene that encodes the particular glucose transporter that insulin regulates when we eat a meal. The amount of GLUT4 protein on the surface of the cell will also determine how fast glucose gets into the cell. Returning to our water analogy, this would correspond to the width of the pipe.
When glucose enters the cell, it is met by two enzymes; hexokinase and aldose reductase. Hexokinase has one main job – get glucose prepared for the energy conversion and storage pathway that we call the Krebs Cycle that ultimately involves the generation of ATP in the mitochondria. Aldose reductase has a bunch of jobs. Aldehydes which can get into the cell can be toxic and aldose reductase transforms them into a less reactive form. Aldose reductase also takes a little bit of the glucose that comes into the cell and converts it to sorbitol for several purposes including osmoregulation as we discussed in the previous post. Unlike the non-enzymatic glycosylation reaction that glucose undergoes outside the cell these enzymatic reactions are very rapid. Outside the cell the bolus of glucose that comes with the meal has little effect on protein glycosylation however inside the cell it is a different matter.
Now, the cell is prepared for meal sized waves of glucose. It has a good amount of hexokinase and as I mentioned in the previous post, hexokinase has a high affinity for glucose. For the diabetic patient or the insulin resistant patient things are different. Glucose levels are high all the time. How high? Perhaps at the level seen during the peak of the meal. The reason for this is the liver. The liver is “designed” to make and supply glucose in between meals. Insulin acts to modulate this along with glucagon. Insulin blocks the liver production of glucose while glucagon promotes the production of glucose. In the diabetic patient or the insulin resistant patient, this balance is screwed up and the liver pumps out lots of glucose all the time. This is why diabetic people can be really good in terms of their diet and yet still have a high blood glucose reading. At any rate, when a person with this problem eats a meal, the level of glucose in the blood is now much higher than before. Using the water pipe analogy, glucose floods the cell, overwhelming the enzymes that are there to meet it. Hexokinase happily churns away, converting glucose to glucose-6-phosphate as quickly as it can. It is not fast enough unfortunately. This state of affairs probably has to do with the need of aldose reductase to make some sorbitol. The factor floor of metabolism was tuned to a certain range of glucose.
The result, as I mentioned in my previous post, is the production of too much sorbitol. Note here how outside the cell, it is the constant exposure of glucose to proteins over a long period of time that causes the problem while inside the cell it is the bolus of glucose during the meal that causes the problem. If fasting glucose levels are under control, the glucose level in the cell will always remain within “design tolerance parameters” and no problems with sorbitol will develop. Diabetes patients on Actos, counter the liver problem by having this glucose constantly taken up into cells. Actos upregulates the amount of GLUT4 on the surface of the cell and in between meals, liver generated glucose is taken care of so that blood glucose levels are down and the flood of glucose into the cell never gets too great.
Why is it that people are not made aware of this issue? Well consider: according to a recent survey most folks do not understand the relationship between glucose and diabetes. We are totally consumed with just trying to get that one message across to the public. The sorbitol issue, in comparison is subtle and confusing. As you saw above, I had to use an engineering argument to explain it fully.
Are people using this knowledge to treat diabetes? The answer is yes – sort of. There are drugs out there but they are not yet good enough for prime time. Tolrestat was marketed by Wyeth but it was pulled from the market in 1997 due to severe liver toxicity. Epalrestat showed a reasonable efficacy in the treatment of diabetic neuropathy in a clinical trial which was published in 2006. It is being marketed in Japan and India but has not been approved by the FDA for use in the US. Ranirestat is an experimental aldose reductase inhibitor that is still undergoing clinical trials. The results of one such trial were recently published by a Japanese group in Diabetes Care.
I hope this answers your question. It was a REALLY good one!
Thank you!
Hi Mary, You have asked an interesting question. Decreased GLUT4 levels should decrease the rate at which glucose enters the cell thus allowing the cell to process glucose via hexokinase, however…..When we create the state of type 1 diabetes in rats (using a drug called streptozotocin) and do not treat the diabetes they get the sorts of complications that we associate with too much sorbitol production. Also, a colleague of mine has genetically engineered mice to be missing the gene for aldose reductase (the enzyme that makes sorbitol) and when these mice are given diabetes (and are not treated with… Read more »
Does this apply to type 1 diabetics? I.e. If people with untreated type 1 diabetes are completely insulin deficient, surely they will get a reduced amount of GLUT4 on cell surfaces? in which case will their cells be flooded by glucose resulting in sorbitol production?