The Science Behind the Glycine Protocol

This is the science behind the Glycine Protocol.

Signs you need to pay attention to your glycine status to fix a health problem include any of the following: anxiety, poor sleep (lack of deep sleep, waking up in the middle of the night, trouble falling asleep or long sleep latency), poor memory, excessive startle response, increased muscle tension and discomfort, poor muscle tone leading to poor posture, poor joint alignment, poor joint health, poor skin quality, excessive aging of the skin, osteopenia or osteoporosis, slow wound healing, intolerance to methylated B vitamins, intolerance to glycine supplements, blood sugar problems, brain fog, lack of alertness, unusually persistent hiccups or episodes of temporary breathing cessation (during the day or as sleep apnea), attention deficit, hyperactivity, schizophrenia, weakness, jerks, or seizures.

One or more of these are sufficient to suspect the relevance of glycine. Some are common, others are rare and extreme. No one will have all of them.

Animal experiments show glycine extends lifespan, so if you are healthy and simply looking to optimize for longevity, this alone should be a reason to optimize your glycine status.

Glycine supplementation is very safe according to existing studies. However, many people anecdotally report adverse effects from glycine supplements such as insomnia, overstimulation, or headaches.

Even if you do not notice any adverse effect from a glycine supplement, it could be generating oxalate, which is a mitochondrial toxin and probably nullifies the benefits of glycine toward lifespan. Oxalate could crystallize in the kidney, joints, or brain, contributing to long-term loss of function in those organs over time.

Hypothetically, excess glycine could also generate methylglyoxal, which could aggravate oxidative stress, accelerate aging, and cause diabetes.

Most glycine is synthesized from carbohydrate using an input of nitrogen from the amino acid glutamate and an assortment of vitamin and mineral cofactors. Thus, it is important to consume enough protein and carbohydrate to maintain endogenous glycine synthesis, and it is probably much safer to default to promoting robust endogenous synthesis than to jump straight into taking glycine supplements.

The ability to tolerate glycine supplements depends on robust nutritional support for two primary ways to convert glycine to carbon dioxide instead of oxalate, ensuring also that excessive glycine does not accumulate, which could lead to an awkward mix of excess neurological excitation and inhibition and general neurotoxicity.

This Article Accompanies The Glycine Protocol

The Glycine Protocol is my 7-Page quick guide to optimizing glycine status that guides you step-by-step through all the reasons your glycine levels or glycine tolerance could need improvement.

Get the Glycine Protocol here:

This Article Is Free For Only 24 Hours

This article is freely available until Monday, March 10, 2025, at 12:30 PM Eastern.

After that it is reserved for Masterpass members. You can learn more about the Masterpass here. You can subscribe to the Masterpass here.

Non-members can use the share button to access the article on the web site:

This Article Is Likely Truncated In Your Email

Masterpass members should also use the share button to read this on the web site because your email service is likely cutting off the end of the article (unless you want to download the PDF below):

Masterpass Members Can Download the PDF Below

The download button is immediately below the table of contents in the email.

In This Article

• Glycine: So Simple, So Important

• What We Can Learn From Human Trials

• What We Can Learn From Disorders of Glycine Synthesis, Clearance, and Receptor Function

• The Relation of Glycine to Serine, Glucose, and Its Derivatives

• How the Conversion of Glucose to Glycine is Regulated

• The Importance of Total Protein

• How We Break Down Excess Glycine

• How We Dispose Of Extra One-Carbon Units

• Balancing Methionine and Glycine

• The Possible Conversion of Glycine to Methylglyoxal

This is educational in nature and not medical or dietetic advice. Any products linked herein may be affiliate links. See terms for additional and more complete disclaimers.

Glycine: So Simple, So Important

Glycine is the simplest stable amino acid. In free form, it tastes sweet, so it is named after the Greek word for “sweet.”

It represents 11.5% of the total body amino acids and 20% of the total body nitrogen. The average person synthesizes about 25 grams of glycine per day, but consumes only about three grams per day. As discussed below, glycine synthesis can easily reach over 40 grams per day depending on diet.

Glycine is important in many proteins, but is especially abundant in collagen, making up one-third of its amino acids and 20-25% of its weight. Collagen represents about 30% of total body protein, but is especially abundant in bone (90% of the protein); skin (70-80% of protein); cartilage, ligaments, and tendons (60-70% of the protein); and extracellular matrix generally (70-95% of the protein).

Glycine makes up a third of the glutathione molecule, which is the master antioxidant of the cell and protects us from wear and tear as we age. Glutathione is also important to mucous fluidity, bronchial dilation, and general protection of the lungs.

Glycine is one of the two amino acids that join together to form creatine, which is like the assistant of the mitochondria, absolutely central to energy metabolism throughout the body, famous for improving strength and vastly underappreciated as a brain booster that combats mental fatigue and mental disorders while promoting healing after traumatic brain injury.

Glycine is also a direct input into the synthesis of heme, and thus to detoxification enzymes, steroid-synthesizing enzymes, the mitochondrial respiratory chain that synthesizes 90% of our ATP, and the hemoglobin that carries oxygen.

Glycine directly detoxifies many foreign substances, as well as toxic intermediates of blocked metabolic pathways.

Glycine is a building block of purines, which makes it important to the synthesis of DNA, RNA, and energy carriers like ATP, NAD+, FAD, and coenzyme A.

Glycine is, along with taurine, an important component of bile salts that promote fat digestion.

Glycine is our endogenous buffer of excess methyl groups, preventing swings of overmethylation. Without sufficient glycine, for example, methyl supply after a meal could lead to mood instability by causing excessive degradation of dopamine and histamine.

Finally, glycine is a critically important neurotransmitter.

What We Can Learn From Human Trials

Human trials support the following:

• Glycine can improve NMDA receptor activation. This likely explains why 100 milligrams of sublingual glycine has been shown to slightly improve memory task performance 30 minutes later. 30-60 grams per day reduces symptoms in schizophrenia on the order of 20-35%.
  

• Glycine-mediated neuroinhibition promotes healthy sleep. This is supported by studies showing 3 grams of glycine one hour before bed reduces daytime sleepiness, helps people fall asleep faster, and helps people feel more rested and perform better cognitively during the day, while 15 grams of collagen peptides (which are rich in glycine) taken one hour before bed has been shown to help people stay asleep without waking up in the middle of the night.
  

• Glycine improves glucose utilization. 5 grams of glycine cuts the blood sugar response in half when it is taken alongside 25 grams of glucose. This does not appear to be driven by insulin and the mechanisms are not clear. This could be due to an anti-inflammatory effect: many immune cells have glycine receptors that can be used to inhibit them in the way that glycine receptors inhibit neurons. It could also be due to an antioxidant effect: glycine makes up a third of the master antioxidant of the cell, glutathione, along with two other amino acids, glutamate and cysteine, and oxidative stress and inflammation can each cause each other in a vicious cycle. 10 grams of N-acetyl-cysteine and 7.5 grams of glycine per day have been shown to improve glutathione status and improve aging markers, and decrease inflammation in older adults. Inflammation and oxidative stress hurt glucose metabolism, so an acute improvement in these phenomena probably explains glycine’s acute improvement of glucose handling.
  

• Glycine-rich collagen promotes collagen synthesis, and 1000 Calories of collagen for four weeks has been shown to help severe burn victims heal nearly four times faster.
  

• Glycine-rich collagen improves skin quality and reduces skin aging. 8 weeks of dissolving 5 grams of marine collagen in hot liquid daily after dinner improved skin moisture and elasticity, and decreased roughness and wrinkles.

On the other hand, while glycine can probably become limiting for creatine synthesis, human studies suggest arginine is typically limiting, and methyl group supply will become limiting when sufficient arginine is supplied.

I cannot find any human trials showing glycine has clear anti-anxiety effects, but as described below deficient glycine signaling will increase the startle response and muscle tension, so I suspect it will also increase general anxiety.

What We Can Learn From Disorders of Glycine Synthesis, Clearance, and Receptor Function

Glycine serves as a neurotransmitter that can act on glycine receptors, serving a classical inhibitory role analogous to that of GABA in the brainstem and spinal cord, serving a weak excitatory role when found outside of traditional synapses, and acting as an enabler for the excitatory role of glutamate on a specific subtype of glutamate receptor known as the N-methyl-D-aspartate (NMDA) receptor.

The role of glycine at NMDA receptors is as a permissive co-agonist, and this is dependent on “ambient” glycine rather than on specific release of glycine at the site in order to stimulate the receptor itself. D-serine can also do the same thing, and the ability of glycine and D-serine to carry out the same function allows different parts of the nervous system to control the activity of NMDA receptors differently.

Glycine receptors often have one beta-subunit that anchors them to an organizational protein known as gephyrin in the membrane, encoded by the GLRB gene, and three active versions of the glycine-binding alpha-subunit encoded by the GLA1, GLA2, and GLA3 genes. There is a GLA4 gene encoding a fourth alpha-subunit, but it is an inactive, non-expressed pseudo-gene in humans. They can form from five identical alpha subunits (which are called “pentameric” because there are five subunits and “homomeric” because there is only one type of subunit), or they can form from an assortment of three alpha subunits and two beta subunits (which are also “pentameric” but are “heteromeric” rather than “homomeric” because there are multiple different types of subunits).

The existence of different alpha subunits allows different combinations of heteromeric receptors with different sensitivities, speed of opening and closing, and rates of desensitization. That homomeric receptors cannot bind gephyrin means that they are more diffusely organized and tend to be found outside of synapses, whereas heteromeric receptors do bind gephyrin and tend to be well-organized in clusters at the synapses of glycine-inhibited neurons.

There are two glycine transporters, 1 and 2, encoded by the SLC6A9 (glycine transporter 1 or GlyT1) and SLC6A5 (glycine transporter 2 or GlyT2) genes. GlyT1 is primarily associated with controlling the ambient glycine around NMDA receptors, while GLYT2 is primarily associated with controlling the glycine level at the synapses of glycine-inhibited neurons.

The dominant effect of glycine is as an inhibitory neurotransmitter that has a calming, cooling, relaxing effect.

It’s ability to act as an excitatory neurotransmitter outside of synapses is very weak and on its own cannot lead to the release of glutamate, but might lower the threshold for other stimulators of glutamate release to have their effect. On the other hand, glycine can be taken up via GlyT1 at the site of glutamate-releasing neurons and lead to their release of glutamate, possibly playing a signaling role letting the neuron know that there is adequate glycine in the environment to allow the NMDA receptors to be activated.

Genetic impairments in glycine receptor subunits, transporters, gephyrin, or the gephryin-anchoring protein collybistin lead to hyperekplexia. This disorder involves dramatically increased startle reflex in newborns, where otherwise trivial sound and touch stimuli elicit a massive increase in muscle tension that can restrict breathing and lead to joint dislocations. Often these disorders are also associated with epilepsy.

It is difficult to isolate the effect of impaired glycine synthesis from other impairments. For example, the enzyme that converts serine to glycine, serine hydroxymethyltransferase (SHMT) is also responsible for interconverting the folate forms tetrahydrofolate (THF) and 5,10-methylene-THF. SHMT has a cytosolic form encoded by the SHMT1 gene and a mitochondrial form encoded by the SHMT2 gene. Loss of the mitochondrial form cuts the intracellular glycine-to-serine ratio in half, so it does constitute a disorder of glycine synthesis, but it also alters folate metabolites and causes a secondary impairment in the mitochondrial respiratory chain, likely driven by altered folate status interfering with mitochondrial protein translation. This disorder does cause spasticity (stiffness) in the limbs, which could be from reduced glycine neurotransmission leading to elevated muscle tone, but it also causes ataxia and peripheral neuropathy, which are more likely due to respiratory chain impairment, and developmental defects that are more likely due to altered folate metabolism.

An analogous disorder of SHMT1 impairment has not been established.

The enzyme 3-phosphoserine phosphatase, encoded by the PSPH gene, catalyzes the final step in the synthesis of serine from glucose, just upstream from the use of SHMT to convert serine to glycine.

Frank deficiency of PSPH has only been documented in 9 cases worldwide, and only in one adult. The one adult diagnosed suffered from contractures, which are progressive shortening of the tissues around the joints over time leading to the progressive loss of joint mobility. She also suffered from an axonal neuropathy that led to chronic neuropathic pain and a chronic non-healing foot ulcer. Treatment with 2.5 grams of serine three times a day for four months completely healed the foot ulcer, allowed her to reduce her pain medication, and improved her energy levels and her sensation in her foot.

Contractures can be due to prolonged, chronic, excessive muscle contraction that leads to remodeling around the joints where tissues have been maintained in a state of constant tightness.

The non-healing foot ulcer could be due to deficient collagen synthesis, and the neuropathy could be due to the loss of serine-containing phospholipids that are critical to myelin formation.

If we synthesize the various signs of muscle tightness across the loss of glycine neurotransmission, serine and glycine interconversion, and serine synthesis, we can arrive at the conclusion that low glycine neurotransmission leads to excessive stiffness in the muscles taking various forms, with or without an exaggerated startle response playing a role.

While medicine treats these with a useful reality distortion filter known as “diagnosis” to triage treatment decisions, these disorders are clearly just showing us the extremes of various spectra. Anxiety isn’t usually mentioned in these reports, but that is probably because it is so common that it isn’t useful in diagnosis. Most likely increased startle reflex and tight muscles are clustered with anxiety, insomnia, and general overstimulation due to impaired inhibitory neurotransmission.

On the other end of the spectrum, impaired clearance of glycine through the glycine cleavage system to break it down into carbon dioxide, ammonia, and a one-carbon unit that enters the folate pool, causes a disorder characterized by frequent hiccups that persist for more than two days at a time; frequent apnea, brief cessation of breathing that can occur even while awake; low muscle tone throughout the body, which progresses to low muscle tone in the trunk and weakness or paralysis of the limbs; attention deficit and hyperactivity; lethargy; jerks, seizures, and coma.

Abnormally persistent hiccups are not isolated to impaired glycine clearance, but they are a rare tell-tale sign that should point one in the direction of looking for elevated glycine. As discussed in this paper, this is thought to be due to aberrant NMDA receptor activation and neurotoxicity in the brainstem. Excessive NMDA activation could be in and of itself sufficient for neurotoxicity. This probably underlies attention deficit and hyperactivity as well.

Despite some signs of NMDA receptor hyperactivation, glycine cleavage deficiency involves a preponderance of signs indicating excessive neuroinhibition. For example, rather than spasticity, elevated startle reflex, and contractures, we find hypotonia, weakness, lethargy, paralysis, and coma.

That seizures can occur in both disorders of deficient and excess glycine signaling is an interesting puzzle. Most epilepsy involves excessive neuroexcitation. More often than not this is from too much glutamate relative to GABA, and one could suppose that loss of glycine could cause seizures similarly to how loss of GABA can. However, in some cases epilepsy is caused by GABA function being reversed and causing excitation instead of inhibition.

As I covered in my lesson, How Neurons Get Excited, GABA and glycine both act as inhibitory neurotransmitters whose receptors are chloride channels that allow negatively charged chloride ions to flow into the neuron, thereby altering the charge differential across the membrane in favor of inhibition:

If the inhibitory receptors are activated too frequently, without the chance for the chloride differential to recover, all of the chloride will wind up in the cell and eventually opening the channel will cause it to come out because there will be no chloride ions left to go in. Whether the neuron gets excited or inhibited is all about the change in charge differential across the membrane, so if the flow of chloride is flipped, then the function of the receptor is also flipped.

It is well established that in the first few weeks of life, GABA acts as an excitatory neurotransmitter that promotes nerve growth. It is suspected that GABA can be excitatory in some cases of epilepsy due to overuse of GABA or to injury to the neuronal membrane that causes the chloride gradient to collapse.

We could suppose, then, that the epilepsy that occurs with elevated glycine is driven by overactivation of glycine receptors and consequent collapse of chloride gradients. However, we cannot dismiss that too much glycine could cause excessive excitation at the normally excitatory extra-synaptic glycine receptors, contribute to glutamate release through inward transport into glutamate-releasing neurons, or contribute to excessive NMDA receptor activation as a co-agonist.

Nevertheless, with a handful of exceptions the disorders overwhelmingly favor an interpretation that glycine is inhibitory, that too little glycine leads to an overly revved up nervous system, and that too much leads to an overly suppressed nervous system.

Therefore, I think that people who get overstimulated by supplementing glycine are unlikely to have some kind of genetically embedded excitatory response to glycine. I find it more likely in these cases that chronically low glycine has NMDA receptor volume turned up in compensation, and that temporarily normalizing glycine status causes this to lead to over-activation of those NMDA receptors because they have not been given a chance to turn their volume back down to normal. If this is the case, the solution is to normalize glycine status more slowly.

Alternatively this could be due to chronically elevated glycine — the solution of course is to measure plasma glycine instead of guessing — but if the extremely elevated glycine in glycine cleavage system disorders still has the preponderance of its manifestation being excessive inhibition, it is hard to believe this is the most probable cause if there is no weakness, low muscle tone, and lethargy accompanying it.

As I also covered in my lesson, How Neurons Get Excited, the clearance of glutamate, GABA, and glycine from synapses is mediated by sodium, chloride, potassium, and hydrogen ions, and these ions are all maintained in concentrations that can perform these activities using energy from ATP. Therefore, deficiencies in salt, potassium, and ATP production could also explain why the neurotransmitters would not clear from their synapses, leading to excessive effects in their normal direction or causing them to flip their normal activity to its opposite.

Watch, read, listen to, or study the How Neurons Get Excited lesson here:

The Relation of Glycine to Serine, Glucose, and Its Derivatives

Glycine is the simplest stable amino acid, consisting of two carbons (each shown for simplicity as the corner of two connected lines), one bound to a nitrogen-containing amino group (NH2), and the other forming an acidic carboxyl group (COOH). The amino and carboxyl groups are present on all amino acids. The twenty amino acids that make up proteins in biological systems all have some side chain proceeding from the central carbon. In the case of glycine, that side chain is a hydrogen atom. In the structure, hydrogens are “silent” for simplicity, so the side chain is not shown.

Glycine is closely related to the amino acid serine.

Serine has a hydroxymethyl group as its side chain. A methyl group is CH3. If a methyl group is hydroxylated one hydrogen is replaced with an OH group, making the hydroxymethylgroup CH2OH. In the figure, the carbons are corners between lines and the hydrogens are silent, so it is shown as an extra line extending to the left from the central carbon of what would otherwise be glycine, with an OH at the end.

The addition and removal of the hydroxmethyl group is mediated by serine hydroxymethyltransferase (SHMT).

The diagram below shows an overview of the ten steps of glycolysis:

Glucose is phosphorylated, split in half, and oxidized, resulting in glyceric acid, which initially has two phosphates attached, gets one removed, one shifted to a different carbon, and then gets dehydrated to phosphoenolpyruvate, which has its final phosphate removed to form pyruvate.

If we zoom in on glyceric acid, imagined with no phosphates attached, we can see that it looks just like serine except it has a hydroxyl (OH) group where serine has an amino (NH2) group:

As shown in the following figure, serine synthesis occurs when 3-phosphoglycerate exits the glycolytic pathway to replace this hydroxyl group with an amino group using nitrogen from glutamate in a three-step process:

3-phosphoglycerate dehydrogenase (enzyme 1) converts 3-phosphoglycerate to 3-phosphohydroxypyruvate with the help of NAD+.

If we look at 3-phosphoglycerate, we can see that, compared to glyceric acid, it has the leftmost carbon bound to a phosphate instead of an OH group:

This is essentially exchangeable with the OH group, because if you were to hydrolyze the phosphate off the molecule, H2O would break into OH and H, with the OH leaving to make the third OH of free phosphate, and the H joining the glycerate molecule to reconstitute its leftmost OH group.

3-phosphoglycerate dehydrogenase oxidizes the middle OH group to a carbonyl group (C=O), which, because it is located in the middle of the molecule, is considered a ketone or a keto group, making hydroxypyruvate a ketoacid:

Ketoacids and amino acids are interconverted using vitamin B6-dependent transaminases, also called aminotransferases. These enzymes switch the keto group of one molecule for the amino group of the other, converting one amino acid to a ketoacid while the original ketoacid becomes an amino acid. With some exceptions, most transaminases use glutamate and alpha-ketoglutarate as the base from which they interconvert another specific pair.

This general principle is explained in much more detail in Masterclass With Masterjohn Energy Metabolism Lesson 6: How Isocitrate Dehydrogenase Makes Oxalosuccinate Decarboxylate Itself. The role of B6 is that its ability to switch between pyridoxal and pyridoxamine allows it to take the keto group and amino group while they are being moved between the two molecules. When the carbonyl group moves onto the B6 molecule, it is on the end instead of the middle, so it is an aldehyde group rather than a keto group, and this is why the molecule is called pyridoxal.

Watch, read, listen to, or study the lesson here:

This general principle applies with no exceptions here. Phosphoserine aminotransferase (enzyme 2), encoded by the PSAT gene, converts 3-phosphohydroxypyruvate to 3-phosphoserine using the help of vitamin B6. In the process, glutamate is converted to alpha-ketoglutarate. The effect is that glutamate donates the nitrogen to form the amino group of serine, and the mechanism is that the keto group of 3-phosphohydroxypyruvate is switched out for the amino group of glutamate and vice versa.

The reactions are completely reversible, but the figure below shows the arrows moving in the direction of serine synthesis.

If you trace the arrows up from glutamate, glutamate is donating its amino group to pyridoxal, forming pyridoxamine, which donates it to 3-phosphoglycerate, forming 3-phosphoserine. Meanwhile, you can trace the arrows down from 3-phosphoglycerate to see that it donates its keto group to pyridoxamine, generating pyridoxal, which donates the keto group to glutamate, forming alpha-ketoglutarate.

Finally, 3-phosphoserine phosphatase, encoded by the PSPH gene, hydrolyzes 3-phosphoserine to serine and free phosphate. Water (H2O) provides the third hydroxyl (OH) group of free phosphate, while the left over H reconstructs the leftmost hydroxyl group of serine. This enzyme is dependent on magnesium and activated by copper.

Serine is then converted to glycine by serine hydroxymethyltransferase (SHMT), which depends on vitamin B6 and tetrahydrofolate (THF).

This is the pyridoxal 5’-phosphate (PLP) form of vitamin B6, shown with a red box highlighting its aldehyde group:

When something with an amino group binds to an aldehyde, the oxygen of the aldehyde leaves with the two hydrogens of the amino group to form water and an alidimine:

In the SHMT reaction, when viewed in the direction of glycine synthesis, serine binds to PLP to form the PLP-serine aldimine. In the picture below, the hydroxymethyl group is shown on the left in red. B6 holds on to the amino acid while the enzyme catalyzes the dehydration of the hydroxymethyl group, causing the net removal of the OH to form water, leaving behind the CH2 that can be added to THF to form 5,10-methylene-THF, shown on the right with the CH2 in red. The actual details of the reaction mechanism are not fully understood, but they involve the conversion of the hydroxymethyl group to formaldehyde and its subsequent conversion to water while it drops its carbon off on the THF molecule.

This reaction is reversible, and a buildup of glycine and/or 5,10-methylene-THF will favor serine synthesis, while a buildup of serine and/or THF will favor glycine synthesis.

Thus, niacin (vitamin B3) in the form of NAD+, vitamin B6, magnesium, copper and folate (vitamin B9) in the form of tetrahydrofolate are all needed to synthesize glycine from glucose, and anything that could excessively elevate the NADH/NAD+ ratio, such as a respiratory chain impairment, alcohol intake, obesity, and sedentariness could compromise the use of NAD+ for this purpose. NAD+ depletion due to DNA damage could act similarly. Finally, any impairments in the utilization of these nutrients or in the specific enzymes involved could hurt serine synthesis.

As we will see below, there are far more nutrients involved in maintaining healthy glycine status, and yet more involved in cleanly clearing out excess glycine to carbon dioxide without generating oxalate.

For this reason the Glycine Protocol has you track your diet in Cronometer using these instructions, these best practices, and these custom nutrient targets just in the Baseline Diet section. For the same reason, if you use the Comprehensive Approach, the protocol has you run the Comprehensive Nutritional Screening.

Learn more about tracking your micronutrients here:

Run the Comprehensive Nutritional Screening here:

How the Conversion of Glucose to Glycine is Regulated

To best understand these points, refer back to the figure on the ten steps of glycolysis, and study the glycolysis, pentose phosphate, and anaplerosis lessons of the MWM Energy Metabolism course.

Here is the glycolysis lesson:

Here is the pentose phosphate lesson:

Here is the anaplerosis lesson:

The flux out of the glycolytic pathway toward serine synthesis, and therefore toward glycine synthesis is regulated to be the third priority after producing cytosolic ATP and citric acid cycle inflow (priority one) and supporting antioxidant defense (priority 2).

These are regulated as follows:

• Phosphofructokinase-1 (PFK-1, encoded by tissue-specific genes PFKL, PFKM, and PFKP) and phosphoglycerate mutase (PGAM, encoded by tissue-specific genes PGAM1 and PGAM2) are major points of regulation.

• Inhibition of PFK-1 by citrate and a high ATP/AMP ratio preserves glucose 6-phosphate for the pentose phosphate pathway when energy needs have been met.

• PGAM is controlled by acetylation, which inhibits it, and deacetylation, which activates it. High energy status favors acetylation and inhibition; reactive oxygen species favor deacetylation and activation.

• The substrate of PGAM, 3-phosphoglycerate, inhibits 6-phosphogluconate dehydrogenase in the pentose phosphate pathway. Thus, inhibition of PGAM by acetylation causes its substrate to rise and inhibit the pentose phosphate pathway.

• The product of PGAM, 2-phosphoglycerate, activates 3-phosphoglycerate dehydrogenase, the first step of serine synthesis. This ensures that glucose going into serine synthesis is “extra” because there is plenty of the next downstream glycolytic metabolite.

• When energy status is low, PFK-1 is active, which feeds the glyocolytic pathway, and the pyruvate kinase reaction at the end of glycolysis is active, and draws the glycolytic intermediates to completion. Thus, glycolysis is used to produce cytosolic ATP and to feed citric acid cycle inflow.

• When energy status is high, but reactive oxygen species need additional defense, high energy status slows PFK-1. Reactive oxygen species activate SIRT2, which keeps PGAM deacetylated and active. Highly active PGAM clears 3-phosphoglycerate, which keeps the pentose phosphate pathway activated. Thus, glucose 6-phosphate is directed into the pentose phosphate pathway.

• When energy status is high and reactive oxygen species defense is adequate, the same things happen except SIRT2 is not activated, and mitochondrial citrate generates cytosolic acetyl CoA, which is used to acetylate PGAM. This makes 3-phosphoglycerate accumulate. It inhibits the pentose phosphate pathway, and acts as the first substrate of serine synthesis. Thus, glycolysis is used for serine synthesis.

Thus, abundance of glucose and energy, combined with scarcity of oxidative stress, is what drives the synthesis of serine and glycine.

For this reason, the Glycine Protocol prioritizes a titration of carbohydrate upwards toward the point of maximal well being, minimal signs of glycine deficiency, or to the plasma glycine at which one feels the best, prior to supplementing with glycine.

The importance of minimizing oxidative stress is why the Comprehensive Approach of the protocol uses the Antioxidant Vitamin and Mineral section of Testing Nutritional Status: The Ultimate Cheat Sheet.

You can get the Cheat Sheet here:

The Importance of Total Protein

While the regulation of serine and glycine synthesis is as described immediately above, the abundance of glutamate is necessary to supply the nitrogen needed.

Glutamate is the most abundant amino acid in the diet, and it is interconvertible with many other amino acids due to the common use of the glutamate/alpha-ketoglutarate pair in transaminases. Thus, this essentially means that we need enough total protein to support glycine synthesis.

In five healthy males, increasing protein intake from 0.4 to 1.5 grams per kilogram bodyweight per day increased the rate of glycine synthesis by 71% from 25.9 to 44.3 grams per day. In six elderly males, the same dietary regimen increased glycine synthesis by 51% from 25.7 to 38.8 grams per day.

In this study the caloric balance was maintained with variable amounts of “carbonated beverages, sucrose soft drinks, cornstarch deserts, and protein-free cookies,” so the protein came largely at the expense of carbohydrate.

This suggests that the nitrogen provided by glutamate is more important to glycine synthesis than the carbon backbone provided by glucose when protein intake is under the necessary threshold, which is somewhere between 0.4 and 1.5 grams per kilogram bodyweight per day.

One possibility is that 1.5 grams protein per kilogram bodyweight is necessary to achieve a glycine synthesis rate of 44 grams per day.

Another possibility is that the effect of protein was maxed out way before that. If we assume that the marginal increase in glycine synthesis is, across some threshold, directly proportional to the protein intake, but then plateaus as soon as the point of maximal impact of protein on glycine synthesis is reached, we could posit the sweet spot is 0.4*1.71=0.684 grams per kilogram bodyweight.

The only other study I know of that provides additional insight into the protein threshold is one that measured the urinary excretion of 5-oxoproline (pyroglutamate) on different protein intakes. This is a marker of not having enough glycine to make glutathione, so it is an implicitly inverse marker of glycine status.

This study showed that 70 grams of protein per day markedly suppresses urinary pyroglutamate compared to 25 grams of protein per day. These intakes likely reflected 1 gram per kilogram bodyweight and 0.35 grams per kilogram bodyweight based on the the average bodyweights reported in the paper, which were about 70 kilograms.

Further study of 40 grams of protein per day or the addition of 6.9 or 13.8 grams of urea per day to the 25 g/d diet suggested the urinary pyroglutamate decreases linearly and dose-dependently across this range and is largely a function of total nitrogen intake. The authors suggested that on very low protein intakes, urea is harvested to ammonia by intestinal microbiota and incorporated into human metabolism to spare the breakdown of amino acids.

In this study, the protein came at the expense of both carbohydrate and fat.

The figure above shows the comparison of the high protein, the low protein, and the two doses of supplemental urea.

If we focus on the effect at day 5 in the above graph, and we read the bars from right to left, and thus from lower protein to higher protein, we see what looks like a dose-dependent reduction in pyroglutamate as urea is supplemented and then as protein is increased to 70 grams per day. 6.9 and 13.8 grams of urea are the equivalent of 20 and 40 grams of protein. So this could be seen as 25, 45, 65, and 70 grams of protein per day.

Only the lowest protein intake is statistically significantly different from the others, so the strongest inference is that 0.35 grams of protein per kilogram bodyweight is not enough, which is consistent with the previous study showing that 0.4 grams per kilogram bodyweight is not enough.

The apparent linearity, however, makes it look like 70 is better than 65, which implies that the optimal dose of protein for glycine synthesis is at least 0.93-1.0 grams per kilogram bodyweight.

Separately, they plotted the low and high protein intakes along with a separate study in which they fed 40 grams of protein per day. This also looks very linear:

If we assume that pyroglutamate can be brought down to zero with high enough glycine status, we could take the apparent linearity and extrapolate it through the zero point from two of their graphs.

Their graph of the three protein intakes extrapolates to a theoretical maximal suppression of pyroglutamate at 107 grams of protein per day:

If I add in the levels found with the two doses of urea at the points of their protein equivalency, it does not seem to detract that much from the linearity:

This lines up with the first study remarkably well, almost stunningly well, because 107 grams per day is the equivalent of 1.53 g/kg bw, and the first study just happened to use 1.5 g/kg bw as the highest of the two intakes of protein.

So:

• The strongest inference is that 0.35-0.4 g/kg bw is inadequate.

• The next strongest inference is that at least 0.93-1.0 g/kg bw is required to maximize glycine synthesis.

• A weaker but quite plausible inference is that glycine synthesis is maximized at about 1.5 g/kg bw per day.

• Since higher levels of protein have not been tested, it is possible that glycine synthesis improves further at higher intakes.

The Baseline Diet of the Glycine Protocol uses 1.5 grams per kilogram bodyweight total protein because there is little reason to eat less protein than this for most people, eating this level of protein is probably going to benefit body composition, and because there are reasonable indications that this is where glycine synthesis is maximized.

How We Break Down Excess Glycine

These are the three primary ways excess glycine is broken down:

• The glycine cleavage system converts glycine to carbon dioxide and ammonia while donating one carbon to folate for use in mitochondrial protein translation, DNA synthesis, purine synthesis, or methylation. Energy released from the chemical bonds is burned for energy in the respiratory chain. This has all of its energy carried to complex I of the respiratory chain and none enters at complex II, which in excess could overwhelm complex I, raise lactate, and compromise all NAD+-dependent reactions.

• SHMT allows glycine to be converted to serine and then to pyruvate, which can be used to synthesize glucose under energy-rich, glucose-poor conditions, or to pyruvate that can be cleanly burned for energy in the respiratory chain. This withdraws one carbon from folate rather than donating one carbon to folate, so constitutes a net loss of two carbons from the folate pool compared to the glycine cleavage system.

• D-amino acid oxidase converts glycine to glyoxylate. If this is not quickly converted back to glycine it will generate oxalate. This route of clearance comes at the expense of forming the mitochondrial toxin oxalate.

The glycine cleavage system is a mitochondrial multienzyme complex consisting of four enzymes, only two of which are unique to the system. Glycine decarboxylase (also called the P protein) is encoded by the GLDC gene, and aminomethyltransferase (also called the T protein) is encoded by the AMT gene. The other two enzymes are shared with pyruvate dehydrogenase (used for glucose metabolism), alpha-ketoglutarate dehydrogenase (used in the citric acid cycle), the branched-chain ketoacid dehydrogenase (used in the catabolism of leucine, isoleucine, and valine), and possibly in the 2-ketoadipate dehydrogenase involved in lysine and tryptophan metabolism.

Most of these other enzyme systems have a unique thiamin-dependent “E1” enzyme and a shared pair of “E2” and “E3” enzymes that depend on lipoic acid, riboflavin, and niacin.

These are covered in MWM Energy Metabolism Lessons 7 (α-Ketoglutarate Dehydrogenase: A Massive Enzymatic Factory), 8 (7 Unforgettable Things About α-Ketoglutarate Dehydrogenase), and 13 (Pyruvate Dehydrogenase: Why Carbs Leave Your Thiamin Working Overtime).

You can watch/listen to/read/study these lessons with the links below.

Alpha-ketoglutarate dehydrogenase:

Pyruvate dehydrogenase:

In the glycine cleavage system, E2 and E3 are called the H and L proteins, and there are two enzymes taking the place of E1, neither of which depend on thiamin.

In the figure below, glycine is shown in red in the upper left:

In the transition of A to B, glycine joins to vitamin B6 as pyridoxal 5-phosphate (PLP) on the P protein in a dehydration reaction that forms an aldimine just like in the SHMT reaction.

In the transition of B to C, the carboxyl group of glycine is released as carbon dioxide. What remains is an aminomethyl group (CH2NH3) that has moved over to a sulfhydryl (SH) group of lipoic acid on the H protein.

In the transition of C to D, tetrahydrofolate (THF) picks up the remaining carbon to become 5,10-methylene-THF.

High-energy electrons from these reactions have been left on the lipoic acid molecule, whose sulfurs had started out oxidized and bonded together in a disulfide bond (see A and B) but are now reduced and shown as SH.

In the transition of D to E, riboflavin in the form of flavin adenine dinucleotide (FAD) takes those electrons away from lipoic acid to itself, becoming FADH2. FAD/FADH2 are bound to the protein and never leave it.

In the transition of E back to A, niacin in the form of nicotinamide adenine dinucleotide (NAD+) takes those electrons away from FADH2, regenerating FAD and itself becoming NADH + H+. Because NADH is mobile, it can now take those electrons to complex I of the mitochondrial respiratory chain.

The 5,10-methylene-THF can be used for mitochondrial protein translation, or to synthesize formate that leaves into the cytosol and is added to THF there to form 10-formyl-THF to support purine synthesis, or to reform 5,10-methylene-THF to support DNA synthesis, or to form 5-methyl-THF to support methylation.

In vitro, the glycine cleavage system is completely reversible, but in vivo it only operates in reverse in anaerobic microbes when NADH is heavily supplied. In humans, glycine cleavage system deficiency results in massive accumulation of toxic levels of glycine, not glycine deficiency.

The upside of this way of clearing glycine is that it contributes useful carbons for the synthesis of mitochondrial proteins, of purines and DNA, or methylation. The downside is that it generates NADH exclusively rather than, like other fuels, a mix of NADH and FADH2. This could overwhelm complex I of the respiratory chain, contribute to oxidative stress, and compromise all NAD+-dependent reactions. This would be reflected in elevated lactate, and this is why the Glycine Protocol uses lactate testing in the Data-Lite and Comprehensive Approaches.

Alternatively, the SHMT reaction described earlier in the direction of glycine synthesis can be reversed, with glycine converting to serine and 5,10-methylene-THF converting to THF:

The following figure shows the two ways of converting serine to pyruvate:

In the first, serine dehydratase (enzyme 4), encoded by the SDS gene, releases ammonia while converting serine to pyruvate. This enzyme depends on B6, and is stimulated by magnesium and potassium.

Alternatively, serine-pyruvate aminotransferase (enzyme 5) converts serine to hydroxypyruvate while converting pyruvate to alanine. This enzyme depends on vitamin B6 as well. It is now known to be identical to alanine-glyoxylate transaminase, encoded by the AGXT gene. Enzyme 6 is referred to as D-glycerate dehydrogenase in this figure but this is now considered to be performed by hydroxypyruvate reductase, encoded by the GRHPR gene. It uses NADH to generate D-glycerate. Then D-glycerate kinase, encoded by the GLYCTK gene, uses ATP to return the metabolite to the glycolytic pathway as 2-phosphoglycerate.

These pathways will be activated if either serine and glycine are present in excess, or if the need for energy or glucose outweigh the needs for these two amino acids. Under either condition, if energy is deficient then the pyruvate will be burned for energy using pyruvate dehydrogenase, the citric acid cycle, and the respiratory chain. If energy is abundant but glucose is scarce, the pyruvate will be used for gluconeogenesis.

The upside of breaking down glycine this way is that it can be used to synthesize glucose. The downside is that it robs the folate pool of a carbon that could have been used to synthesize purines, DNA, or methyl groups. When compared to the glycine cleavage system, it is both failing to donate a carbon and also stealing a carbon, leading to net loss of two carbons per glycine molecule from the folate pool.

The last major way to break down glycine is with D-amino acid oxidase, which is an enzyme of the peroxisome. Peroxisomes are in some ways assistants to mitochondria in that they synthesize mitochondrial lipids and make up for gaps in mitochondrial energy metabolism when mitochondria are overwhelmed or when certain things need to be burned for energy that mitochondria don’t have the enzymes for. In another way they are specialists in detoxification of many substances; for example, about 25% of alcohol is detoxified in peroxisomes. D-amino acid oxidase can act on a number of substances, but one of its reactions is to convert glycine to glyoxylate and ammonia. From molecular oxygen (O2), one oxygen atom combines with water to make hydrogen peroxide and the other is added to the glycine molecule in place of the amino group.

This reaction is dependent on riboflavin as FAD, whose role is to remove electrons from glycine and add them to what becomes the hydrogen peroxide molecule.

Glyoxylate then has two fates. One is for it to be converted by to glycine using alanine-glyoxylate transaminase (AGXT):

This is a B6-dependent transaminase and is one of the few human transaminases that does not use the glutamate/alpha-ketoglutarate pair. Rather than glutamate donating the nitrogen for glycine synthesis, alanine does. As glyoxylate is converted to glycine, alanine is converted to pyruvate.

The other is to be converted by L-lactate dehydrogenase to oxalate. Most likely LDH acts on the spontaneously formed hydrate of glyoxylate, dihydroxyacetic acid:

The AGXT reaction is totally reversible, but severe AGXT deficiency results severe hyperoxaluria, showing that the standard flux of this enzyme is toward glycine synthesis to support clearance of glyoxylate without oxalate generation.

However, I suspect that the reason thiamin deficiency raises oxalate in animals is that pyruvate cannot be cleared through pyruvate dehydrogenase, so it causes pyruvate to accumulate to the point that it drives the AGXT reaction backwards.

In conclusion:

• The glycine cleavage system depends on vitamin B6 as pyridoxal 5’-phosphate (PLP or P5P), folate in the form of tetrahydrofolate (THF), riboflavin in the form of FAD, and niacin in the form of NAD+. It is the ideal way of clearing out glycine unless glucose needs to be synthesized or complex I of the respiratory chain is overwhelmed.

• Conversion to glucose or energy via SHMT and serine catabolism requires B6, magnesium, and copper; if ATP, ammonia and/or NADH are abundant it will tend to follow a path still requiring B6 but requiring ATP and NADH instead of magnesium and copper. The main advantage of this pathway is that it can synthesize glucose and that burning pyruvate for energy is less taxing on complex I of the respiratory chain than burning glycine directly through the glycine cleavage system. The downside is it detracts one-carbon units from the folate pool.

• The more the above pathways are deficient, the more they encourage the production of oxalate.

Oxalate inhibits pyruvate kinase, the last enzyme of glycolysis, so can raise blood glucose. This is why the Glycine Protocol has you test your glucose in the Data-Lite and Comprehensive Approaches.

The diversity of nutrients involved is why the Glycine Protocol has you track your diet in Cronometer using these instructions, these best practices, and these custom nutrient targets just in the Baseline Diet section. For the same reason, if you use the Comprehensive Approach, the protocol has you run the Comprehensive Nutritional Screening.

Learn more about tracking your micronutrients here:

Run the Comprehensive Nutritional Screening here:

How We Dispose Of Extra One-Carbon Units

When methyl groups are in excessive abundance, SAM shuts down MTHFR, which causes 5-methyl-THF levels to drop. 5-methyl-THF acts as an off-switch for glycine N-methyltransferase (GNMT), so the drop in methylfolate causes GNMT to turn on and methylate glycine.

Glycine can be methylated once to sarcosine, or twice to dimethylglycine.

Sarcosine and dimethylglycine can both have their methyl groups harvested in the mitochondrion while converting THF to 5,10-methylene-THF using riboflavin as FAD, nonheme iron, and CoQ10. This regenerates glycine.

However, if the ratio of 5,10-methylene-THF to THF in the mitochondrion is very high, this indicates that there is no need to scavenge the methyl groups due to a sustained excess, which allows sarcosine and dimethylglycine to be peed out in the urine.

If the ratio of 5,10-methylene-THF to THF in the cytosol is very high, then these non-methyl one-carbon units can be disposed of by converting glycine to serine via SHMT.

Thus, excesses of one-carbon units in the form of 5,10-methylene-THF or SAM can both be eliminated using glycine, leading to glycine depletion in the process.

If glycine levels are not adequate to handle this, it will become a cause of low glycine.

This is why serine, sarcosine, and the inverse marker of THF levels, formiminoglutamate, factor heavily into the interpretation of lab data in the Comprehensive Approach of the Glycine Protocol.

Balancing Methionine and Glycine

Since glycine is used to dispose of excess methyl groups and can be depleted in the process, it is likely that methionine and glycine should be balanced in the diet.

Balancing Methionine and Glycine in Foods: The Database outlines the principles involved and has a searchable database of methionine-to-glycine ratios in foods.

Each glycine molecule is capable of absorbing the methyl groups from one to two methionine molecules.

If we add up all of the glycine uses that apply to a typical person, such as helping with sleep, blood sugar, and collagen synthesis, we could easily get to twenty grams a day.

If we look at different sources of protein, non-shellfish, non-collagen animal proteins really stand out as having a high ratio of methionine to glycine.

If you were to meet the protein recommendations here from foods like steak, eggs, and milk, matching 10 grams of collagen for every 100 grams of such protein would put you in the range where any excess methyl groups from methionine could be buffered by glycine, leaving over enough glycine to stay in the therapeutic range, and converging on the methionine-to-glycine ratios you would find if you ate a plant-based or shellfish-based diet.

Therefore, the Baseline Diet of the Glycine Protocol calls for collagen to balance non-collagen, non-shellfish animal protein.

The Possible Conversion of Glycine to Methylglyoxal

Glycine can condense with acetyl CoA to form free CoA and 2-amino-3-ketobutyrate, which spontaneously decarboxylates to form aminoacetone, which is converted by amine oxidase to methylglyoxal.

Excess methylglyoxal can deplete glutathione, cause oxidative stress, and contribute to diabetes.

I could not find any evidence demonstrating the relevance of this pathway in humans. However, it hypothetically could be a way of clearing out acetyl CoA to free up the CoA pool when the acetyl CoA cannot enter the citric acid cycle and exceeds the capacity for ketogenesis, or when it cannot enter the citric acid cycle in a tissue with little or no capacity for ketogenesis. Oxalate inhibits pyruvate carboxylase, which forms oxaloacetate, which is needed to allow acetyl CoA to enter the citric acid cycle. So it is possible that glycine generating oxalate would inhibit the citric acid cycle and thereby cause other glycine molecules to form methylglyoxal.

This does not have direct evidence in humans so does not feature into the Glycine Protocol. However, rising blood glucose could serve as a possible indicator of this occurring as well as the generation of oxalate.

Get the Glycine Protocol

Want this boiled down in to a practical set of to-do items?

The Glycine Protocol is my 7-Page quick guide to optimizing glycine status that guides you step-by-step through all the reasons your glycine levels or glycine tolerance could need improvement.

Get the Glycine Protocol here: