An unconscious child with keto-acidosis and a skin rash

His immediate problem is severe acidosis, so the most appropriate emergency treatment would be intravenous infusion of bicarbonate to permit respiratory compensation for the acidosis by exhaling carbon dioxide.

He regained consciousness rapidly. Over the next few days he continued to respire rapidly, and even after a meal he excreted ketone bodies in his urine. His plasma glucose remained in the low normal range, and his plasma insulin was normal both in the fasting state and in response to an oral load of 50g of glucose /kg body weight.

Rapid respiration is a response to acidosis - an attempt to breathe out carbon dioxide and so shift the equilibrium above to the right, and raise plasma pH.

The fact that he was excreting ketone bodies suggests that he was still suffering from keto-acidosis. Without further information it is difficult to account for ketosis in the fed state after a meal.

The normal fasting plasma insulin and the normal response to a glucose load means that we can rule out diabetes mellitus.

Analysis of his urine by high pressure liquid chromatography revealed that he was excreting relatively large amounts of a variety of organic acids that are not normally excreted. In addition to the compounds shown on the right, he was also excreting glycine conjugates of propionic, tiglic and methylcrotonic acids.

As shown below, tiglyl CoA is an intermediate in the metabolism of isoleucine, and normally is metabolised onwards, via propionyl CoA to yield succinyl COA, an intermediate of the citric acid cycle.

3-Methylcrotonyl CoA is an intermediate in the metabolism of isoleucine, and is normally metabolised onwards to acetyl CoA and acetoacetate.

If he cannot metabolise tiglyl CoA, propionyl CoA and 3-methylcrotonyl CoA onwards for some reason, then as these accumulate, they will be hydrolysed to salvage the CoA, leaving the free acids. These cannot be metabolised further, and will be excreted in the urine.

Many organic acids are conjugated with glycine (and sometimes also alanine) before excretion. Excessive intake of e.g. benzoic acid can outstrip the body's capacity for glycine synthesis, so impairing protein synthesis and rendering glycine a partially essential amino acid

As shown on the right, propionyl CoA is an alternative substrate for citrate synthase, forming methylcitrate. Methylcitrate cannot be metabolised further, and so is excreted in the urine.

There will be depletion of oxaloacetate, leading to a reduction of citric acid cycle activity, and hence a deficit in ATP formation.

As ADP accumulates, it undergoes reaction to form 5'AMP and ATP, catalysed by adenylate kinase.

This is a very minor source of ATP, but the 5'AMP acts as a metabolic signal; among other actions it results in increased beta-oxidation of fatty acids in the liver, for ATP formation.

5'AMP leads to increased beta-oxidation of fatty acids in the liver. However, because of the lack of oxaloacetate, the acetyl CoA arising from beta-oxidation cannot be metabolised through the citric acid cycle, and can only be used for synthesis of ketone bodies.

Oxaloacetate can be replenished by carboxylation of pyruvate, catalysed by pyruvate carboxylase.

He cannot synthesise oxaloacetate from pyruvate. At the same time, the accumulation of acetyl CoA from beta-oxidation of fatty acids inhibits the decarboxylation of pyruvate to acetyl CoA. This leads to accumulation of pyruvate that cannot be utilised. MUch will be reduced to lactate as a way of regenerating NAD+, as occurs in anaerobic metabolism, and much will be transaminated to alanine, as a means of reducing the acidosis due to lactate and pyruvate.

In the fed state he is metabolising mainly glucose, and would normally be expected to synthesise fatty acids to increase adipose tissue triacylglycerol reserves for use in the fasting state. He seems to be unable to do this, and the acetyl CoA from glucose metabolism that cannot be used for either citric acid cycle activity or fatty acid synthesis is used for ketone body synthesis in the liver.

The first reaction of fatty acid synthesis is the carboxylation of acetyl CoA to malonyl Coa, catalysed by acetyl CoA carboxylase.

There seem to be four defective enzymes: acetyl COA carboxylase, pyruvate carboxylase, propionyl CoA carboxylase and methylcrotonyl CoA carboxylase. It is extremely unlikely that he would have genetic defects of all four enzymes.

However, it is known that all four carboxylases use the coenzyme biotin as the carrier of carbon dioxide to catalyse the carboxylation reaction. Furthermore, Barry's skin rash and hair loss resemble the signs of biotin deficiency.

Dietary deficiency of biotin is extremely rare, but can result from consumption of relatively large amounts of uncooked egg white. There is a protein in egg white, avidin, which binds biotin, rendering it unavailable for absorption. When the dietitian interviewed Barry's mother she said that he was never given raw eggs, but did eat mashed hard-boiled eggs (the yolks of which are a rich source of biotin).

Avidin is a protein, and therefore it will be denatured when the eggs are cooked, so it will not be able to bind biotin.

The paediatrician still suspected biotin deficiency, despite the diet history, and sent plasma and urine samples for biotin assay. Barry's plasma biotin was 0.2 nmol /L (the normal level for a child of 9 months is > 0.8 nmol /L).

The urine sample showed that he was excreting a significant amount of biotin, mainly in the form of biocytin (biotiyl lysine) and small peptides containing biocytin.

Barry was treated with 5 mg of biotin /day. After 3 days the various abnormal organic acids were no longer detectable in his urine, and his plasma lactate, pyruvate and ketone bodies had returned to normal, although his excretion of biocytin and biocytin-containing peptides increased. At this stage he was discharged from hospital, with a supply of biotin tablets. After 3 weeks his skin rash began to clear, and his hair loss ceased.

Three months later, at a regular out-patient visit, it was decided to cease the biotin supplements for a trial period. Within a week the abnormal organic acids were detected in his urine again, and he was treated with varying doses of biotin until the organic aciduria ceased. This was achieved at an intake of 150 µg /day (compared with the reference intake of 10 - 20 µg /day for an infant under 2 years old). He has continued to take 150 µg of biotin daily, and has remained din good health for the last 4 years.

In carboxylases, biotin is covalently bound at the catalytic site of the enzyme to a lysine residue, as biocytin. It is this covalently bound biocytin that acts as the carbon dioxide carrier in carboxylation reactions, forming carboxybiocytin.

Most, if not all, of the biotin in foods is also present as biocytin, in enzymes in the food. The enzyme biotinidase catalyses hydrolysis of biocytin to free biotin, which can be absorbed.

A single enzyme, holocarboxylase synthetase, catalyses the incorporation of biotin into the apo-enzymes of the four carboxylases: acetyl COA carboxylase, pyruvate carboxylase, propionyl CoA carboxylase and methylcrotonyl CoA carboxylase.

When the carboxylases are catabolised in normal protein turnover, the biotin is released from biocytin and small peptides containing biocytin by hydrolysis, catalysed by biotinidase.

Barry suffers from defective activity of all four carboxylases, a condition that can be called multiple carboxylase deficiency. There are two genetic defects that can lead to multiple carboxylase deficiency:

• a defect of holocarboxylase synthetase
• a defect of biotinidase

High doses of biotin resolve Barry's metabolic problems and clinical signs. Either biotinidase deficiency or holocarboxylase synthetase deficiency could be the cause of his problem. An abnormal holocarboxylase synthetase with a very high Km for biotin would have little or no activity under normal conditions, but in the presence of very high concentrations of biotin (as would be expected after his high dose supplements) would have appreciable activity, perhaps enough to permit more or less normal carboxylation of the apo-enzymes.

Barry excretes biocytin and biocytin-containing peptides in his urine, and the amounts of these increased when he was given biotin supplements. This suggests that he is capable of utilising free biotin (from the supplements) to form biocytin in the carboxylases, but is not able to salvage the biotin when the enzymes are catabolised, so that biocytin and biocytin-containing peptides are excreted in his urine.

If the problem was a defect of holocarboxylase synthetase, the high dose of biotin would permit formation of active carboxylases, and when these are catabolised there would be biotinidase activity to permit hydrolysis of biocytin and liberation of free biotin. Therefore he would not excrete biocytin in is urine.

Multiple carboxylase deficiency due to holocarboxylase synthetase deficiency is normally apparent at or very shortly after birth.

At birth, Barry will have had more or less normal activity of the carboxylases, because free biotin is transported across the placenta. After birth, these enzymes will gradually be catabolised, and he will not now be able to salvage the biotin, so will develop functional biotin deficiency.

By contrast, an infant who lacks holocarboxylase synthetase will be unable to form the holocarboxylases in the first place without very high doses of biotin. This means that at birth it will be deficient in all four carboxylases.

Key points from this exercise:

• Biotin provides the prosthetic group for carboxylases; it is covalently bound to the side-chain amino group of a lysine residue in the protein as biocytin (biotinyl-lysine), and acts as the carrier for carbon dioxide in carboxylation reactions.
• There are four carboxylases involved in human metabolism: acetyl COA carboxylase, pyruvate carboxylase, propionyl CoA carboxylase and methylcrotonyl CoA carboxylase.
• A single enzyme, holocarboxylase synthetase, catalyses the insertion of biotin into all four apo-carboxylases, and genetic defects of holocarboxylase synthetase lead to failure of all four carboxylases, with clinical signs of functional biotin deficiency and urinary excretion of a variety of abnormal organic acids.
• Biotin is normally recovered from carboxylases when they are catabolised by hydrolysis of biocytin, catalysed by biotinidase.
• Genetic defects of biotinidase lead to failure of all four carboxylases, with clinical signs of functional biotin deficiency and urinary excretion of a variety of abnormal organic acids.
• Lack of pyruvate carboxylase activity leads to depletion of oxaloacetate and reduced citric acid cycle activity, and increased synthesis of ketone bodies, which cannot be metabolised because of the lack of oxaloacetate.
• Lack of acetyl CoA carboxylase means that acetyl COA arising form glycolysis in the fed state cannot be utilised for fatty acid synthesis - it will again be used for ketone body synthesis, leading to ketosis even in the fed state.

Click here for more information on holocarboxylase synthetase deficiency.

Click here for more information on biotinidase deficiency