Glycosylation and Glycation

In light of recent discussions about increased protein intake producing a rise in blood sugar, this seems to be a good time to repeat a post from 2008. It helps explain why elevated blood sugar can present potential long-term health risks.
---------------------------------------------------
When proteins are assembled in our cells, sometimes specific sugar molecules are attached to them in carefully-defined ways. This is called glycosylation. Enzymes add the sugar molecules to help proteins fold properly and to route proteins to various places inside and outside the cell. Glycosylation patterns also help our bodies to distinguish proteins that are "self" versus "not-self" and are useful in immune responses. Glycosylation results from controlled reactions and is important for our biochemical wellbeing.

When we have glucose in our blood (and if we're alive, we do), sugar molecules are also added to proteins in a random fashion. The random addition of sugar molecules to proteins is called glycation. If only single glucose molecules have been added to a protein, when the blood sugar level drops, the glucose can detach and the protein will again be normal. But if blood glucose remains high, more sugars will be added. These will rearrange and crosslink, eventually producing something called an Advanced Glycation Endproduct or AGE. One example of an AGE is hemoglobin A1c, the form of hemoglobin found elevated amounts in the red blood cells of poorly-controlled diabetics. Evidence suggests that many other proteins in our bodies are also converted into Advanced Glycation Endproducts by elevated blood sugar. Glucose and fructose in the blood interact with and crosslink these other proteins in our bodies, forming AGEs that accumulate in our eyes, kidneys, arteries, nerve endings, joints and skin. The end result of AGE accumulation can be retinal disease, kidney failure, atherosclerosis, peripheral neuropathy, frozen joints and cracked skin.

Although our bodies have mechanisms to cope with the identification and disposal of AGEs, the AGEs gradually accumulate and stiffen our tissues. The elasticity of youth is slowly replaced by the physical degeneration of old age. In other words, crosslinked AGE proteins produce in us the symptoms we associate with old age. This happens in all people, but the process is made worse and happens more quickly in the presence of elevated blood sugar.

(The illustration is taken from the cover of the journal Science, March 23, 2001.)

How Can Eating Excess Protein Raise Blood Glucose?

It is almost an article of faith among low-carbers that the low-carb lifestyle is able to lower blood glucose values in diabetics and pre-diabetics. It would be logical to assume that the lower the carbohydrate intake, the lower the corresponding blood glucose. But recent observations in a limited sample of people who were doing something very close to zero-carbing suggest that this is not necessarily the case.

Donald K. Layman has done some interesting work on the effect of dietary protein on glycemic control that may help explain this phenomenon. In an article in The Journal of Nutrition, he presents a diagram of the glucose-alanine cycle, which appears in modified form above.

For those who are not familiar with this type of diagram, here is a brief explanation. Ingested protein enters the gut and is digested into amino acids. The amino acids are taken up in the blood and proceed to the liver, where many of them are metabolized. However the branched-chain amino acids leucine, isoleucine and valine are unique. Although they constitute 15-25% of protein intake, they experience very little metabolism in the liver. Most of the branched-chain amino acids, abbreviated BCAA, continue to move through the circulation and are eventually absorbed by muscle cells.

In muscle cells the branched-chain amino acids have two possible fates. First, when branched-chain amino acids enter a muscle cell, they promote protein synthesis. Our muscle tissue is continually undergoing repair, and because of this each of us has an individual daily protein need. If sufficient high-quality protein is consumed, this repair is able to take place without loss of lean muscle tissue.

Second, if there is an excess of amino acids in the muscle cells, the surplus branched-chain amino acids enter the pathway of energy production. In order to do this, they must have their amino group (NH₃)removed in a process called transamination. The amino group from a BCAA is transferred to a molecule called alpha-keto-glutarate to form the amino acid glutamate. Next, another transamination transfers the amino group from the glutamate to pyruvate, transforming the pyruvate into the amino acid alanine. The alanine leaves the muscle cell and travels to the liver, where it is turned into pyruvate by removal of the amino group, and then the pyruvate is turned into glucose by gluconeogenesis. The liver sends the newly-synthesized glucose into the blood, where it can be taken up by muscle cells and broken down once again into pyruvate. Each pyruvate is ready to accept another amino group from one of the branched-chain amino acids, and the cycle repeats itself until the branched-chain amino acids have been used up.

The glucose-alanine cycle explains why it is possible to have an elevated blood glucose while eating essentially only meat and fat. Normally, leucine signals the muscle cells to synthesize protein and maintain lean body mass. When an excess of branched-chain amino acids is available, leucine serves as a metabolic signal to muscle cells telling them to upregulate their use of BCAA as a fuel, while simultaneously downregulating their use of glucose as a fuel. Any glucose that appears in the cell is preferentially broken down into pyruvate, which is used to accept excess amino acid nitrogen (NH₃ groups) and allow them to be removed them from the cell in the form of alanine. In the liver, the alanine is recycled into glucose, and the glucose is returned to the blood until it is no longer needed to mop up excess NH₃ groups in peripheral tissues.

If this pathway is correct, it shows that excess amino acids not only provide the raw materials for glucose synthesis in the liver, but they also require additional glucose synthesis in the liver in order to allow branched-chain amino acids to be converted into energy.

Metabolic regulation is a huge topic, and this post presents only a small piece of it. Once again, please do not modify your lifestyle in accordance with what you read here. In the overall context of a human organism, it may be incomplete or even incorrect. However the glucose-alanine cycle does provide a possible explanation for what some people have seen with regard to a higher-than-normal blood sugar while eating essentially zero carbohydrates.

Observations on Protein Intake in Low-Carbers

Last week I asked if people doing low-carb or zero-carb might be willing to test their blood glucose before and after meals and report their results. Many finger sticks later, we have a few tentative observations. Please note, this was NOT a scientific study in any way. Don't change your life or your eating habits based on what you read here. The purpose of this post is to consider ideas and to raise possibilities, particularly if you have been having trouble succeeding on low-carb or zero-carb. That said, here are the patterns that seemed to emerge from the data.

1. Some people, particularly people over 50, do have an increase in blood glucose following meals that are either entirely or mostly meat and fat. Dr. Bernstein says the optimum level of blood glucose is 83 mg/dl. For zero-carbers over 50, the fasting blood glucose was often somewhere between 95 and 110 mg/dl and could even go as high as the high teens. For long-time low-carbers over 50, fasting blood glucose was usually somewhere in the 80's. In both low-carbers and zero-carbers over 50, it was not unusual to have a 30-40 mg/dl rise in blood glucose after consuming a large amount of protein, such as a 12-ounce ribeye. Because protein is slowly digested, blood glucose levels sometimes stayed elevated for three to five hours or longer. It is important to remember that at blood sugars above about 100 mg/dl, insulin is secreted and its presence keeps fat in the fat cells. This may explain why low-carbers over 50 have such a hard time losing weight if they eat as much protein as they want. Insulin levels stay elevated for long periods, forcing most of what they eat into storage, and keeping it there until insulin levels finally come down again.

2. Most people under age 50 do not have a rise in blood glucose following a meal, even a large meal, that is mostly meat and fat. I had three participants in the under-50 group whose blood sugars stayed approximately in the 80's following meals ranging from a 1/3 pound hamburger to a ribeye steak. Two of them told me that they occasionally see rises to near 100 mg/dl, but often there is no rise at all.

3. Decreasing protein intake in two participants over 50 to the amount recommended at Blood Sugar 101 caused a decline in average pre-meal blood glucose to the low 90's and post-meal glucose values between about 90 and 110 mg/dl. In fact, both of them started losing weight again after several months of eating as much protein as they wanted and gradually gaining weight.

4. And then there were the outliers, which I shall address below.
Two participants occasionally experienced a fall in blood glucose following a low-carb meal. Neither has been diagnosed with diabetes. Nevertheless (unless they were eating more carbs than usual), their blood glucose sometimes declined after they had eaten a low-carb meal of meat and vegetables. One was a man and one was a woman. One was under 40 and one was over 60. The woman, SC, suggested to me that it might have something to do with the fact that she is a super-taster. When I checked with the other one, who happens to be Jimmy Moore, it turned out that he is also a super-taster. Just to be sure, I checked with super-taster Cleochatra. She did not have blood glucose data to give me, but she said, "I can tell when I've eaten a carrot, even when it's been hidden in a dish, because my stomach is growling within minutes and I want to dive face first into various vats of puddings. I can say in all honesty, artificial sweeteners made me starve...and when I'm VLc I feel fantastic. No woobly or feelings of hunger at all." Later she specified that Splenda and the sugar alcohols are the artificial sweeteners that affect her.

[In the comments, Mariasol asked what made a person a super-taster. Although there are tests for this ability, I simply used an informal question as a criterion: If I poured out five unlabeled dixie cups of Diet Rite, Diet Pepsi, Diet Coke, Coke Zero and Splenda Coke, could you correctly label each cup with the brand, based on taste alone? If your answer to that question is yes, you probably are a super-taster. Subsequently, I have been told that when a super-tasters are cooking something and then add in the salt, they can smell the salt. Just like everything else in this post, the super-taster information has been collected in a non-rigorous manner, so please do not take it as settled science.]

From a limited sample size of three, I can speculate that super-tasters are the ones whose insulin is on a hair-trigger. As soon as they eat, or maybe even before they eat, they secrete enough insulin to nail any food that might appear in the stomach. And if that food happens to be diet soda, it's possible that the insulin secretion occurs anyway. This can either trigger hunger pangs, or if the diet soda is consumed continuously, can keep insulin levels relatively high and thus prevent fat mobilization and weight loss.

All of this is anecdotal. It didn't come down on tablets at Mt. Sinai, so various parts of it could be wrong. But I present it as something worth thinking about in the context of a low-carb lifestyle.

Very many thanks to Cleochatra, ES, D, Jimmy Moore, K, KM, LR, SC, SG, SO, V, VS, P and U for providing data that was used in this blogpost.

Protein Intake and Blood Glucose Levels

Low-carbers know that when a person eats foods that contain carbohydrates, his blood glucose will rise. As the pancreas releases insulin in response, the blood glucose levels will gradually return to normal.

What happens when a person eats protein? Insulin is released in response to protein as well, enabling the amino acids to be removed from the blood and stored in the tissue. The cells don't know the insulin is there to remove amino acids from the blood, so they will take up glucose from the blood as well. To prevent hypoglycemia, the liver gradually releases glucose into the blood to replace the glucose that has been stored.

In the graph above, the white lines show us that when a normal person eats 50 grams of protein, the blood glucose remains the same out to five hours after the meal, even though a significant amount of insulin has been released. The person with type 2 diabetes is represented by the yellow lines. His blood glucose levels start out at a much higher level, but when he eats 50 grams of protein, his blood glucose levels also stay steady out to two hours and then actually begin to drop because a great deal of insulin has been released. These graphs are found at Metabolic response of people with type 2 diabetes to a high protein diet.

It is important to realize that the response to protein in both the diabetic and non-diabetic person are happening in people who are not low-carb-adapted. Low-carbohydrate-adapted people are able to make all the carbs they need through gluconeogenesis. Their brains and muscles have switched over to the use of ketones and fatty acids for fuel, and the 40 or so grams of glucose they need for glucose-requiring tissues are readily converted from glycogenic amino acids and the glycerol backbones of triglycerides. So, what happens when a person who eats very low carbs has a meal of protein? For a rather extreme example, look at the graph below.

Lex Rooker is a very dedicated and meticulous individual who posts at the Raw Paleo Forum. (In no way do I either support or condemn what Lex does regarding his diet, but his journal certainly makes fascinating reading.) For about two years, Lex ate a single daily meal in the afternoon, at the time marked by an asterisk on the graph. This meal contained 150 grams of protein and consisted of 68% fatand 32% protein. As you can see, his blood glucose remained rock-steady at about 106 mg/dl throughout the day. But a couple of hours before he ate, it would drop to 95 mg/dl. After he ate a meal consisting solely of meat and fat, his blood glucose would rise about 25 mg/dl, returning to baseline in about four hours. (The graph shows a rise of 15 mg/dl, but he refers to the amount of the rise several times, so this may be an error in the graph.)



At one point Lex decided to switch things up a bit. He kept his calories the same, but ate only 90 grams of protein per day, making the ratio 80% fat and 20% protein. His baseline blood glucose dropped into a range between 68 and 78. After his single daily meal of meat and fat, his blood glucose would rise about 15 mg/dl, though it would take longer than before to come down to baseline. It appears that decreasing the amount of protein intake also decreases the amount of glucose released into the blood of a low-carb-adapted person.

People who are not low-carb-adapted do not do much gluconeogenesis because they get plenty of glucose from their diet. People like Lex Rooker who eat no carbs at all, apparently do quit a bit of gluconeogenesis. Low-carbers fall somewhere in between those two points. This provokes a question to which I do not currently have an answer: What does a normal blood glucose curve look like in a low-carber? If he chooses to eat only meat and fat at a particular meal, does his blood glucose rise or does it stay steady? If he eats a few carbs with each meal, does it rise less, or does it rise more than it would without the carbs?

In other words, this time it's not a blog, it's a bleg. If anybody has data on what a normal (or abnormal) daily blood glucose curve looks like in a low-carber, would you please share that information in the comments? Thanks!

(If any of the graphs are too fuzzy to read, just click on them and you'll get a clearer version.)