We demonstrated brain biochemical differences by 1HMRS in medically-stable adults with partial OTCD, even in individuals who were considered asymptomatic with normal anatomic MRI findings. We found a decrease in mI in PWM, FWM, PCGM, and tha as well as elevated gln in PCGM, PWM, FGM, and FWM in many asymptomatic subjects, and elevated NAA in PWM of subjects. The concentration of mI was inversely correlated with disease severity score () and plasma gln levels. Glutamine concentration was elevated in subjects, both in symptomatic and many asymptomatic females. Choline was decreased in thalamus and in FWM. Myoinositol and gln exhibit an inverse relationship in subjects with partial OTCD reflecting the presumed osmotic compensation with mI depletion in the face of increased gln production due to limitations of the astrocyte to continue to provide nitrogen buffering in the face of hyperammonemia in OTCD ().
This observation suggests unrecognized biochemical disturbances that may also underlie previous findings of cognitive impairments in a pattern indicating a white matter injury model. It may also serve as a biomarker to identify carriers of a gene mutation in OTCD, and those who may have low clinical reserve in the face of triggers for hyperammonemia. Subjects with repeated hyperammonemic episodes or neonatal mutations revealed the most striking reductions in mI.
The decrease of mI is supported by other studies [24
]. The role of mI in the brain is uncertain, but it plays a role as an osmolyte with importance in cell volume in astrocytes. The mI component in MRS is a composite signal with the bulk of the contributions coming from mI itself. Previous theories posit that decreases in mI are due to volume regulatory mechanisms with astrocytic mI release as a direct response to ammonia-induced astrocytic accumulation of gln [30
]. The finding of decreased mI also occurs in hepatic encephalopathy associated with the effects of hyperammonemia [31
]. In astrocyte cultures, mI is released through volume-sensitive anion channels in response to osmotic stress [30
Elevations of NAA in the white matter of subjects were unexpected. The possibility of either another compound resonating at the same ppm or increased axonal density has not been evaluated. It has been suggested that NAA may play a role as a molecular water pump or in brain nitrogen balance [32
The choline resonance contains contributions from phosphocholine and glycerophosphorylcholine, cell membrane precursors and degradation products. The observed reduction may reflect alterations in phospholipid metabolism, membrane alterations, membrane fluidity, or secondary changes in water content, glycerophosphorylcholine being a cerebral osmolyte.
There are several possible limitations of our methodologic approach that others experience in similar studies. First, overlap of the spectra of metabolites is a challenge with one-dimensional MR spectroscopy. The spectra for glutamine, N
-acetylaspartate, γ-aminobutyric acid, and aspartate overlap in the 2–3 ppm region, whereas mI, glutamate/glutamine, glucose, and aspartate spectral peaks overlap in the 3.5–4 ppm region. To account for and overcome these phenomena, we used time-domain fitting with ‘LCModel’ which is more accurate for quantitation of overlapping peaks. Although voxel placement was designed to maximize tissue homogeneity, the relatively large size of our voxel (2×2×2 cm3
) resulted in inclusion of both gray and white matter in some voxel locations. By using segmentation analysis as we describe, the contributions from gray versus white matter minus CSF is more accurately reflected in the total metabolite concentrations that are reported. Previous autopsy and neuroimaging studies show that OTCD results in white matter injury. Many previous studies reported small series of patients. The majority of these subjects had a history of hyperammonemic coma and were studied at different stages of disease, and emphasized CT which is not comparable to MRI [33
] in delineating white matter pathology.
When ammonia is not adequately detoxified by the urea cycle, there is an increase in scavenger amino acids, including glutamine [34
]. Ammonia entering the brain is rapidly incorporated into the formation of glutamine by glutamine synthetase, present in the astrocyte, which serves to rapidly detoxify and buffer excess ammonia [35
]. As a result, glutamine concentrations increase in hyperammonemic states [34
]. Previous 1
H magnetic resonance spectroscopy studies of patients with UCDs demonstrated elevations of the glutamine/glutamate complex and depletion of mI that may be detected non-invasively in various stages of disease.
Glutamine has been implicated in hyperammonemic encephalopathy [36
]. Previous studies demonstrated that 1
H MRS can be used to non-invasively detect elevated brain glutamine in chronic hepatic encephalopathy [37
] as well as experimentally-induced hyperammonemia [38
]. A rise in plasma glutamine levels precedes hyperammonemia [38
]. The role of glutamine in this process has been demonstrated by the temporal relationship between hyperammonemia, neurological dysfunction, and glutamine concentrations in CSF observed in patients with hepatic encephalopathy. A triad of biochemical changes that can be observed by 1
H MRS include increased Glx, decreased mI, and decreased choline, which is the reverse order of changes seen in hepatic encephalopathy (decreased choline, mI depletion, and increased Glx) [24
Inhibition of glutamine synthetase in hyperammonemic rats by treatment with enzyme inhibitors prevents a rise in cortical glutamine levels and cortical water content [40
]. Acute ammonia toxicity is mediated by excessive activation of the NMDA type of glutamate receptors that mediate glutamate neurotoxicity. Clearance of synaptic glutamate by glial cells is required for the normal function of excitatory synapses and to prevent further neurotoxicity. This process occurs in the astrocyte, which takes up glutamate from the synapse and returns it to the neurons in the form of glutamine, a non-toxic amino acid. Neurons subsequently reconvert glutamine to glutamate via the action of mitochondrial phosphate-dependent glutaminase.
Biochemical changes such as altered energy and membrane metabolism and intercellular metabolite trafficking (glutamate-glutamine shuttle) may be detected prior to structural MRI changes. Given the normal structural MRI scans in the majority of our subjects, these data suggest that biochemical changes precede structural changes, thus providing a window of opportunity for intervention before irreversible damage ensues.
We demonstrate that brain biochemistry differs in subjects with partial OTCD and may explain neurocognitive differences between these groups. These data are in agreement with previous studies, and support white matter pathology. We hypothesize that mI decrements mark prior hyperammonemic episodes and represent a marker of vulnerability. We observed elevations of glutamine in several brain regions despite normal plasma ammonia and normal plasma glutamine levels in some subjects. Further work is ongoing to clarify these aspects of brain metabolism in subjects with OTCD. Our findings have implications for clinical practice and dietary management to prevent cognitive sequelae of hyperammonemia or its effects. Future investigations are directed towards understanding elevation of NAA in white matter as well as the extent to which dietary factors may play a role in effects on brain function and biochemistry.