Metabolism
on-line - the virtual tutorial room
copyright © 2008 - 2015 David A Bender
An unconscious child with hyperammonaemia and keto-acidosis
At the
age of 28 weeks Angela was admitted to the Emergency Department at her local
hospital in a coma, having suffered a convulsion after feeding. She had a mild
infection and slight fever at the time.
Since birth she had been a sickly child, and she frequently vomited and became drowsy after feeding. She was bottle fed and at one time cows’ milk allergy was suspected, although the problems persisted when she was fed on a soya-milk substitute. She had always shown poor muscle tone, and indeed at times was described as a "floppy baby", unable to hold her head up unsupported.
On admission she was moderately hypoglycaemic (plasma glucose 2.8 mmol /), ketotic and her plasma pH was 7.29. Analysis of a blood sample showed normal levels of insulin, but considerable hyperammonaemia (plasma ammonium ion concentration 500 µmol/L; reference range 40 - 80 µmol/L).
She was ketotic and acidotic. What treatment would be appropriate for this?
The standard treatment for acidosis is intravenous bicarbonate to permit respiratory compensation for the acidosis by shifting the bicarbonate / carbon dioxide equilibrium to the left
She was hypoglycaemic. What treatment would be appropriate for this?
Intravenous glucose.
What would be the appropriate treatment for her hyperammonaemia?
A rectal infusion of lactulose.
See the exercise on Hyperammonaemic coma due to liver failure for discussion of how intestinal bacterial fermentation of lactulose acidifies the intestinal contents and permits lowering of plasma ammonia.
Click here for a summary of the ammonia-lowering action of the products of lactulose fermentation. Press the space bar to advance the animation
A glucose tolerance test gave normal results, and she showed a normal increase in insulin secretion in response to the glucose load.
What conclusions can you draw from this information?
One possible cause of hypoglycaemia and keto-acidosis is diabetes mellitus. The normal glucose tolerance test and insulin secretion eliminate this as a likely diagnosis.
Defects of any of the the enzymes of urea synthesis lead to hyperammonaemia
after a moderately protein-rich meal. Therefore a liver biopsy sample was taken,
and the following enzyme activities were determined. The control samples were
from six infants of about the same age who were being investigated for conditions
that did not lead to hyperammonaemia. The results are shown in the table below:
µmol product formed /min
/mg protein |
||
Angela |
control samples |
|
| carbamoyl phosphate synthetase | 0.337 |
1.30 ± 0.40 |
| ornithine carbamoyl transferase | 29.0 |
18.1 ± 4.9 |
| argininosuccinate synthetase | 0.852 |
0.49 ± 0.09 |
| argininosuccinase | 1.19 |
0.64 ± 0.15 |
| arginase | 183 |
152 ± 56 |
What conclusions can you draw from these results?
The results suggest that she has a partial defect of carbamoyl phosphate synthetase, which would explain the hyperammonaemia after a moderately protein-rich meal. Note that the activities of the other enzymes of urea synthesis are moderately elevated, as might be expected in response to hyperammonaemia, when there is induction of these enzymes.
Lack of carbamoyl phosphate synthetase would not explain either the severe keto-acidosis and hypoglycaemia, or her poor muscle tone and muscle weakness.
She remained well on a high-carbohydrate, very low-protein feed for several days, although the poor muscle tone and muscle weakness persisted. A second liver biopsy sample was taken after four days when her plasma ammonia was well within the normal range. The results are shown in the table below:
µmol product formed /min
/mg protein |
||
Angela |
control samples |
|
| carbamoyl phosphate synthetase | 1.45 |
1.30 ± 0.40 |
| ornithine carbamoyl transferase | 28.6 |
18.1 ± 4.9 |
| argininosuccinate synthetase | 0.75 |
0.49 ± 0.09 |
| argininosuccinase | 0.95 |
0.64 ± 0.15 |
| arginase | 175 |
152 ± 56 |
What conclusions can you draw from these results?
The activity of carbamoyl phosphate synthetase is now normal, thus suggesting that the initial diagnosis of a urea cycle enzyme defect was incorrect. The activities of the other enzymes of urea synthesis have all fallen somewhat. This is as would be expected since they will no longer be being induced in response to hyperammonaemia.
Before the results from this second biopsy sample were available, the keto-acids of several of the essential amino acids (threonine, methionine, leucine, isoleucine and valine) were introduced into her feed. This is standard practice in cases of urea cycle defects where it is desirable to maintain the nitrogen intake as low as possible to prevent hyperammonaemia, yet ensure an adequate supply of essential amino acids for growth.
This precipitated another attack of ketosis and metabolic acidosis, although her plasma ammonia was normal this time.
What conclusions can you draw from this information?
This suggests that the underlying problem is in the metabolism of the carbon skeletons of one or more of these essential amino acids.
In order to investigate this further, a plasma sample and a urine sample were analysed by high pressure liquid chromatography. This revealed a number of abnormalities, as shown in the table below:
Angela |
reference range |
|
| plasma propionic acid (µmol /L) | 24 |
0.7 - 3.0 |
| urine methylcitrate (µmol /mg creatinine) | 1.1 |
not normally detectable |
| urine short-chain acyl carnitine, mainly propionyl carnitine (µmol /mg creatinine) | 28.6 |
5.7 ± 3.5 |
What conclusions can you draw from these results?
She obviously has a problem metabolising propionic acid; her plasma concentration is abnormally high and she is excreting a great deal of propionic acid as propionyl carnitine.
Propionyl CoA is the homologue of acetyl CoA, and can be expected to act as a weak substrate for, and therefore competitive inhibitor of, enzymes that utilise acetyl CoA. It is formed in the metabolism of the carbon skeletons of leucine, valine, methionine and threonine - those keto-acids she was given in an attempt to ensure an adequate supply of essential amino acids.
Therefore, assuming that her condition is due to accumulation of propionyl CoA that cannot be metabolised onwards to methylmalonyl CoA and then the citric acid cycle intermediate succinyl CoA, we can explain her adverse reaction to milk. It is likely that the especially severe reaction that led to her admission was the effect of protein intake at a time of fever, when she would anyway have been metabolising a significant amount of tissue protein.
CoA is derived from the vitamin pantothenic acid, any unmetabolised propionyl CoA is likely to undergo hydrolysis to free propionic acid, or acyl transfer to carnitine, which is not normally a dietary essential, in order to spare CoA and its vitamin precursor.