www.elsevier.com/locate/neuint

Neurochemistry International 50 (2007) 1042–1051

The metabolic role of isoleucine in detoxification of ammonia

in cultured mouse neurons and astrocytes

Maja L. Johansen a,1, Lasse K. Bak a,1, Arne Schousboe a, Peter Iversen b, Michael Sørensen b,c,Susanne Keiding b,c, Hendrik Vilstrup c, Albert Gjedde b, Peter Ott c, Helle S. Waagepetersen a,*

a Department of Pharmacology and Pharmacotherapy, Faculty of Pharmaceutical Sciences, University of Copenhagen, 2 Universitetsparken,

2100 Copenhagen, Denmarkb Positron Emission Tomography Center, Aarhus University Hospital, 8000 Aarhus, Denmark

c Department of Medicine V, Aarhus University Hospital, 8000 Aarhus, Denmark

Received 2 November 2006; received in revised form 15 January 2007; accepted 18 January 2007

Available online 6 February 2007

Abstract

Cerebral hyperammonemia is a hallmark of hepatic encephalopathy, a debilitating condition arising secondary to liver disease. Pyruvate

oxidation including tricarboxylic acid (TCA) cycle metabolism has been suggested to be inhibited by hyperammonemia at the pyruvate and

a-ketoglutarate dehydrogenase steps. Catabolism of the branched-chain amino acid isoleucine provides both acetyl-CoA and succinyl-CoA, thus

by-passing both the pyruvate dehydrogenase and the a-ketoglutarate dehydrogenase steps. Potentially, this will enable the TCA cycle to work in the

face of ammonium-induced inhibition. In addition, this will provide the a-ketoglutarate carbon skeleton for glutamate and glutamine synthesis by

glutamate dehydrogenase and glutamine synthetase (astrocytes only), respectively, both reactions fixing ammonium. Cultured cerebellar neurons

(primarily glutamatergic) or astrocytes were incubated in the presence of either [U-13C]glucose (2.5 mM) and isoleucine (1 mM) or [U-13C]iso-

leucine and glucose. Cell cultures were treated with an acute ammonium chloride load of 2 (astrocytes) or 5 mM (neurons and astrocytes) and

incorporation of 13C-label into glutamate, aspartate, glutamine and alanine was determined employing mass spectrometry. Labeling from

[U-13C]glucose in glutamate and aspartate increased as a result of ammonium-treatment in both neurons and astrocytes, suggesting that the TCA

cycle was not inhibited. Labeling in alanine increased in neurons but not in astrocytes, indicating elevated glycolysis in neurons. For both neurons

and astrocytes, labeling from [U-13C]isoleucine entered glutamate and aspartate albeit to a lower extent than from [U-13C]glucose. Labeling in

glutamate and aspartate from [U-13C]isoleucine was decreased by ammonium treatment in neurons but not in astrocytes, the former probably

reflecting increased metabolism of unlabeled glucose. In astrocytes, ammonia treatment resulted in glutamine production and release to the

medium, partially supported by catabolism of [U-13C]isoleucine. In conclusion, i) neuronal and astrocytic TCA cycle metabolism was not inhibited

by ammonium and ii) isoleucine may provide the carbon skeleton for synthesis of glutamate/glutamine in the detoxification of ammonium.

# 2007 Elsevier Ltd. All rights reserved.

Keywords: TCA-cycle; Aspartate; Glutamate; Glutamine; Metabolism; Hepatic encephalopathy; Energy; Glucose

1. Introduction

Cerebral hyperammonemia is thought to play a pivotal role

in hepatic encephalopathy (HE), a debilitating neurological

Abbreviations: AAT, aspartate aminotransferase; ALAT, alanine amino-

transferase; HE, hepatic encephalopathy; GDH, glutamate dehydrogenase; a-

KGDH, a-ketoglutarate dehydrogenase; PBS, phosphate-buffered saline; PC,

pyruvate carboxylase; PDH, pyruvate dehydrogenase; TCA, tricarboxylic acid

* Corresponding author. Tel.: +45 35306470; fax: +45 35306021.

E-mail address: hsw@farma.ku.dk (H.S. Waagepetersen).1 Contributed equally to this work.

0197-0186/$ – see front matter # 2007 Elsevier Ltd. All rights reserved.

doi:10.1016/j.neuint.2007.01.009

condition arising secondary to severe liver disease. The

majority of patients suffering from liver cirrhosis develop at

least one episode of HE during the cause of the disease, varying

from minor disturbances in personality to frank coma. The

pathogenesis of HE is largely unknown; however, cerebral

intoxication caused by ammonium derived from the intestines

(usually detoxified by the liver) is believed to play a role and

hyperammonemia is a hallmark of liver cirrhosis (e.g. review by

Albrecht and Dolinska, 2001). The glutamine synthetase (GS)

reaction, forming glutamine from glutamate by amidation, is

the quantitatively most important way for disposal of excess

ammonium in the mammalian brain (review, Hawkins et al.,

M.L. Johansen et al. / Neurochemistry International 50 (2007) 1042–1051 1043

2002). Interestingly, GS is selectively localized in astrocytes

(Norenberg and Martinez-Hernandez, 1979). In addition, the

mitochondrial glutamate dehydrogenase (GDH; both neurons

and astrocytes) reaction may also to some extent serve this

purpose, as it fixes ammonium while forming glutamate from

a-ketoglutarate by reductive amination, a mechanism shown to

operate in cultured neurons (Yudkoff et al., 1990).

Cerebral hyperammonemia has been suggested to inhibit

tricarboxylic acid (TCA) cycle metabolism at the level of both

pyruvate dehydrogenase (PDH) and a-ketoglutarate dehydro-

genase (a-KGDH) which are both rate-limiting steps in

oxidative metabolism (Lai and Cooper, 1986; Zwingmann

et al., 2003). The consequences of this might be an overall

reduction of oxidative (TCA cycle) metabolism, which in turn

may result in increased glycolysis. In a recently published

hypothesis, degradation of the carbon skeletons of the

branched-chain amino acids (BCAAs) particularly isoleucine

and valine, was suggested to be able to compensate for the

ammonium-induced inhibition of a-KGDH by introducing

anaplerosis subsequent to the a-KGDH step plus acetyl-CoA

(Fig. 1) (Ott et al., 2005). The anaplerotic function of

isoleucine/valine was suggested to provide the carbon skeletons

for glutamate and glutamine synthesis by GDH and GS,

respectively, both reactions consuming ammonium and thus

contributing to detoxification of excess ammonium. There are

important differences in the degradation of the BCAAs carbon

skeleton; leucine ultimately ends up as acetyl-CoA and may

then be oxidized in the TCA cycle, whereas valine ends up as

succinyl-CoA, and may act as an anaplerotic substrate.

Fig. 1. Schematic representation of the tricarboxylic acid (TCA) cycle plus the py

synthetase (GS) reactions. Glucose is catabolized to pyruvate in the process of glyco

condenses with oxaloacetate to form citrate. As indicated, the PDH step may be inhi

(a-KG) that can be reductively aminated to glutamate (Glu) by GDH producing NAD

by GS. As indicated, the formation of succinyl-CoA from a-KG by a-ketoglutarate d

the branched-chain amino acids is complex involving multiple steps. Leucine enter

succinyl-CoA and acetyl-CoA. See text for further details.

Isoleucine, on the other hand, is degraded into both acetyl-

CoA and succinyl-CoA. This would make isoleucine an ideal

substrate to counteract both the inhibition of a-KGDH and

PDH by ammonium.

In the present study, the role of isoleucine as a TCA cycle

substrate and metabolism of glucose during exposure to

ammonium chloride was investigated in cultured mouse

cerebellar neurons or astrocytes. The objective was: (i) to

detect any significant inhibitory effect of ammonium exposure

on TCA cycle metabolism; and (ii) to determine the ability of

isoleucine to support TCA cycle metabolism and formation of

glutamate/glutamine in the face of an ammonium challenge.

Both types of cell cultures were incubated in the presence of

different levels of ammonium chloride and either [U-13C]iso-

leucine or [U-13C]glucose. Cell extracts were analyzed by mass

spectrometry for incorporation of label into alanine, glutamate,

glutamine (in astrocytes) and aspartate. Release of glutamine

into the medium (astrocytes only) was quantified by HPLC.

2. Materials and methods

2.1. Materials

Seven-day-old NMRI mice were obtained from Taconic M&B (Ry, Den-

mark). Plastic tissue culture dishes were purchased from NUNC A/S (Roskilde,

Denmark), fetal calf serum from SeraLab Ltd. (Sussex, U.K.). Culture medium

and poly-D-lysine (MW > 300,000) were from Sigma Chemical Co. (St. Louis,

MO, U.S.A.). Penicillin was from Leo (Ballerup, Denmark). Isotopically

labeled compounds were either from Cambridge Isotopes Laboratories, Inc.

(Massachusetts, U.S.A.) or Isotec (a subsidiary of Sigma Chemical Co.), all

ruvate dehydrogenase (PDH). glutamate dehydrogenase (GDH) and glutamine

lysis and the resulting pyruvate is decarboxylated by PDH to acetyl-CoA which

bited by hyperammonemia. Citrate is in three steps converted to a-ketoglutarate

(P)+ in the process. In astrocytes, glutamate may be amidated to glutamine (Gln)

ehydrogenase (a-KGDH) may be inhibited by hyperammonemia. Catabolism of

s the TCA cycle as acetyl-CoA, valine as succinyl-CoA and isoleucine as both

Fig. 2. Simplified schemes illustrating tricarboxylic acid (TCA) cycle metabolism of [U-13C]glucose (A) or [U-13C]isoleucine (B). Labeled carbon atoms are

represented by black circles. (A) TCA cycle metabolism of [1,2-13C]acetyl-CoA or acetyl-CoA derived from [U-13C]glucose or glucose/lactate. respectively. In the

first turn, unlabeled oxaloacetate condenses with either labeled or unlabeled acetyl-CoA. The scheme shows all the possible combinations of label in glutamate and

aspartate during three turns of the TCA cycle. Insert: labeling patterns of glutamate (first turn) and direct formation of aspartate from labeled oxaloacetate arising from

pyruvate carboxylase activity on [U-13C]pyruvate. (B) TCA cycle metabolism of [U-13C]isoleucine and unlabeled acetyl-CoA, the latter derived from glucose/lactate.

[U-13C]isoleucine is metabolized to [U-13C]succinyl-CoA and [1,2-13C]acetyl-CoA (the latter assumed to be negligible compared to the contribution from unlabeled

glucose/lactate, see insert). In the first turn, unlabeled oxaloacetate condenses with either labeled or unlabeled acetyl-CoA. The scheme shows all the possible

M.L. Johansen et al. / Neurochemistry International 50 (2007) 1042–10511044

M.L. Johansen et al. / Neurochemistry International 50 (2007) 1042–1051 1045

were >98% enriched. The Phenomenex EZ:faast LC–MS kit was used for

amino acid analysis. All other chemicals used were of the purest grade available

from regular commercial sources.

2.2. Cell cultures

Cerebellar neurons were cultured from seven-day-old mice essentially as

described by Schousboe et al. (1989). After dissection of the cerebellum, the

tissue was exposed to a mild trypsinization (0.25 mg/ml trypsin, 15 min, 37 8C)

followed by trituration in a DNase solution (75 i.u./ml) containing a trypsin

inhibitor (0.53 mg/ml) from soybeans. The cells were suspended

(3.0 � 106 cells/ml) in a slightly modified Dulbecco’s medium (Hertz et al.,

1982) containing 24.5 mM KCl, 12 mM glucose, 7 mM p-aminobenzoate,

50 mM kainate and 10% fetal calf serum and cultured in poly-D-lysine coated

25 cm2 or 80 cm2 culture flasks (5 or 15 ml/flask, respectively). To prevent

astrocytic proliferation, cytosine arabinoside was added after 48 h in culture to a

concentration of 20 mM. The cells were cultured for 7–8 days, at which point

the cultures exhibit pronounced glutamatergic characteristics. This is due to the

treatment with cytosine arabinoside and kainate, which inhibit proliferation of

astrocytes and GABAergic cell function, respectively (Drejer and Schousboe,

1989; Sonnewald et al., 2004, 2006). The medium of the individual cultures was

supplemented twice with an aliquot of glucose to reach a minimum concentra-

tion of 12 mM. Cerebellar astrocytes were cultured from dissociated cerebella

from 7-day-old mice as detailed by Hertz et al. (1989) and Waagepetersen et al.

(2000). Briefly, the dissected cerebella were mechanically dissociated by

squeezing the tissue through 80 mm nylon sieves into the culture medium

and subsequently the cell suspension was seeded in 25 cm2 or 80 cm2 culture

flasks (5 or 15 ml/flask corresponding to 0.8 or 4 animals, respectively). The

cells were cultured in a slightly modified Dulbecco’s medium containing 6 mM

glucose and 2.5 mM glutamine (Hertz et al., 1982). The culture medium was

changed twice a week and maintained for three weeks before experiments were

performed. The last week of the culturing period, di-butyryl-cAMP (final

concentration 0.25 mM) was added to the culture medium, which induces

morphological and biochemical differentiation (Juurlink and Hertz, 1985; Hertz

et al., 1989).

2.3. Incubation experiments

Following the culture period, the medium was discarded and the cells were

rinsed twice in phosphate-buffered saline (PBS; 137 mM NaCl, 2.7 mM KCl,

7.3 mM Na2HPO4, 0.9 mM CaCl2, 0.5 mM MgCl2, pH 7.4, 37 8C) and sub-

sequently incubated for 5 h in serum-free culture medium (25 cm2 flasks: 5 ml;

80 cm2 flasks: 15 ml) containing ammonium chloride (0, 2 or 5 mM), 1 mM

lactate, 2.5 mM glucose and 1 mM [U-13C]isoleucine or 1 mM lactate, 2.5 mM

[U-13C]glucose and 1 mM isoleucine. In addition, the serum-free medium

contained 0.8 mM of unlabeled isoleucine. The incubation medium was

supplemented with ammonium chloride after 2.5 h to a minimum concentration

of the level for that experiment, i.e. 2 (astrocytes) or 5 mM (neurons and

astrocytes). For all experiments, a 30 min pre-incubation period was employed

by adding ammonium chloride to the culture medium. The incubation was

terminated by separating the medium from the cells which were rinsed twice in

PBS and extracted with 70% (v/v) ethanol. The cells were scraped off the flask

and the cell-ethanol mixture was centrifuged at 20,000 g (20 min) to separate

the soluble extract from the insoluble component. In parallel, the resulting

supernatant and the incubation medium (astrocytes only) were subsequently

lyophilized and reconstituted in water for biochemical analyses.

2.4. Biochemical analyses

Amino acids were separated and quantified by reversed-phase HPLC

employing pre-column, on-line o-phthaldialdehyde derivatization and fluores-

cence detection (excitation 350 nm; detection 450 nm) (Geddes and Wood,

1984). Mass spectrometric analysis was performed on an LC–MS system

combinations of label in glutamate and aspartate during two turns of the TCA cycle a

AT, aminotransferase; a-KG, a-ketoglutarate; GDH, glutamate dehydrogenase; LDH

pyruvate dehydrogenase.

consisting of a Shimadzu LCMS-2010 mass spectrometer coupled to a Shi-

madzu 10A VP HPLC system. The Phenomenex EZ:faast amino acid analysis

kit for LC–MS was used for analysis of labeling in relevant amino acids. Protein

content was determined according to Lowry et al. (1951) using bovine serum

albumin as the standard.

2.5. Calculation of percent molecular carbon labeling (MCL)

To obtain a measure of total incorporation of 13C label into intracellular

glutamate, the average percent of labeled carbon atoms was calculated, as

initially introduced by Bak et al. (2006). Glutamate may contain anywhere

between one and five 13C atoms. To provide a measure of the total labeling of

the glutamate pool, the percent of the individual isotopomers are first multiplied

by the number of carbons labeled. Subsequently these numbers are summed up

and expressed as a percent of the total number of carbon atoms. Hence, percent

MCL is a measure of the total labeling of glutamate.

2.6. Data analysis

All labeling data were corrected for natural abundance of 13C by subtracting

a sample of the relevant metabolite. Isotopic enrichment was calculated

according to Biemann (1962). Data analysis was performed using Microsoft

Excel 2003 and GraphPad Prism v4.01. All data are presented as averages � the

standard error of the mean (S.E.M.). Differences between groups were analyzed

using one-way ANOVA followed by Bonferroni post hoc test. A P-value <0.05

was considered statistically significant.

3. Results

In the present study, 13C-labeling from either [U-13C]iso-

leucine or [U-13C]glucose observed in alanine, glutamate,

glutamine (astrocytes only) and aspartate is employed to study

the effect of ammonium on TCA cycle metabolism of these two

substrates. TCA cycling may in any given microenvironment be

monitored using labeling of glutamate since aspartate

aminotransferase (AAT) activity is several-fold higher than

that of the TCA cycle (Drejer et al., 1985; Fitzpatrick et al.,

1990; Mason et al., 1992). When [U-13C]glucose is the

substrate, the incorporation of label in glutamate is a product of

the extent of label in the pyruvate/acetyl-CoA pool as well as

TCA cycling; see Bak et al. (2006) for a more extensive

discussion on this matter.

3.1. Incorporation of 13C-label into relevant amino acids

from [U-13C]glucose or [U-13C]isoleucine

Label from [U-13C]glucose (Fig. 2A) enters the TCA cycle

as [1,2-13C]acetyl-CoA formed from [U-13C]pyruvate gener-

ated in glycolysis. In addition, [U-13C]pyruvate can be

transaminated by alanine aminotransferase (ALAT) to

[U-13C]alanine or carboxylated in astrocytes by pyruvate

carboxylase (PC) to [1,2,3-13C]oxaloacetate, which will

equilibrate with the preceding reversible reactions of the

TCA cycle to form [2,3,4-13C]oxaloacetate as well (Fig. 2A,

insert). Glutamate and subsequently glutamine in astrocytes are

labeled from the TCA cycle intermediate a-ketoglutarate.

Likewise, aspartate is labeled from the TCA cycle intermediate

s indicated. AAT, aspartate aminotransferase; ALAT, alanine aminotransferase;

, lactate dehydrogenase; OAA, oxaloacetate; PC, pyruvate carboxylase; PDH,

Fig. 3. Cultured cerebellar neurons were incubated in serum-free culture

medium in the presence of [U-13C]glucose (2.5 mM) and isoleucine (1 mM)

for 5 h. The effect of 5 mM ammonium chloride was investigated (black bars;

see Section 2 for details). Cell extracts were analyzed by mass spectrometry for

percent 13C-labeling of glutamate (A), alanine (B) and aspartate (C). The results

are averages of 4–5 cultures � S.E.M. (*): significantly different from the

control condition (P < 0.05). n.q: not quantifiable.

M.L. Johansen et al. / Neurochemistry International 50 (2007) 1042–10511046

oxaloacetate. Fig. 2A shows all possible isotopomers of

glutamate and aspartate after three turns of the TCA cycle when

[1,2-13C]acetyl-CoA or unlabeled acetyl-CoA is metabolized,

as well as the isotopomers formed via PC activity in the first

turn.

Label from [U-13C]isoleucine (Fig. 2B) enters the TCA

cycle at the level of succinyl-CoA following several steps

(Yudkoff, 2006). The resulting succinyl-CoA and subsequently

oxaloacetate is triple labeled and the latter may be transami-

nated to triple labeled aspartate. In the first turn of the TCA

cycle, glutamate and glutamine (in astrocytes) will be double

and triple labeled, whereas aspartate will be double labeled.

Provided that an unlabeled acetyl-CoA condenses with labeled,

i.e. first-turn derived oxaloacetate, a second turn of the TCA

cycle results in mono and double labeling of glutamate/

glutamine.

3.2. Effects of ammonium on metabolism of

[U-13C]glucose or [U-13C]isoleucine in neurons

To study the effect of ammonium chloride (5 mM) on

neuronal TCA cycle metabolism of glucose or isoleucine,

cultured cerebellar neurons were incubated in the presence of

either [U-13C]isoleucine (1 mM) and glucose (2.5 mM) or

[U-13C]glucose and isoleucine. [U-13C]glucose was included to

study the effect of ammonium on glycolysis and TCA cycling

per se.

Labeling from [U-13C]glucose in glutamate was extensive,

showing a decrease in mono and double labeling whereas

quadruple and uniform label were increased when incubated in

the presence of ammonium chloride (5 mM; Fig. 3A). This

indicates that TCA cycle metabolism of [1,2-13C]acetyl CoA

was elevated, as quadruple and uniformly labeled glutamate

are formed in the second or third turn of the TCA cycle,

respectively. Uniform labeling of alanine was increased as an

effect of ammonium treatment (Fig. 3B), reflecting increased

formation of [U-13C]pyruvate which is transaminated into

[U-13C]alanine (see Fig. 2). Label in aspartate showed the same

pattern as that of glutamate, i.e. decreases in mono and double

plus increases in triple and uniform labeling, again indicating

that TCA cycling of 13C-label from glucose was increased

(Fig. 3C). The amounts (nmol/mg protein) of individual

isotopomers of glutamate (Table 1) showed a similar pattern of

significant differences with regard to the effect of ammonium as

the percent labeling except for double and triple labeling which

was unchanged and increased, respectively.

Labeling of glutamate and aspartate from [U-13C]isoleucine

was lower than that from [U-13C]glucose (Fig. 4); however, it

should be noted that the enrichment of [U-13C]isoleucine was

only 56% in the medium. Mono, double and triple labeling in

glutamate decreased significantly in the presence of ammonium

chloride (Fig. 4A). In line with this, mono and double labeled

aspartate decreased as well, probably as a consequence of

increased TCA cycle metabolism of unlabeled glucose or lactate.

However, triple labeled aspartate did not change, indicating that

the flow of label from [U-13C]isoleucine into succinyl-CoA was

not inhibited (Fig. 4B). In addition, the amount of triple labeled

aspartate actually increased significantly from 0.20 � 0.01 to

0.27 � 0.02 nmol/mg protein (P < 0.05), showing increased

synthesis of aspartate from [1,2,3-13C]/[2,3,4-13C]oxaloacetate

originating from [U-13C]isoleucine. Significant decreases were

observed in the amounts of mono and double labeling of

glutamate (Table 1).

The total amounts of glutamate, aspartate and alanine

increased after incubation in the presence of ammonium

chloride (Table 2), indicating that ammonium is fixed by the

neuronal GDH reaction forming glutamate that is used for

transamination to form alanine and aspartate.

3.3. Effects of ammonium on metabolism of

[U-13C]glucose or [U-13C]isoleucine in astrocytes

Cultures of astrocytes were incubated in the presence of

either [U-13C]isoleucine (1 mM) and glucose (2.5 mM) or

[U-13C]glucose and isoleucine with or without exposure to

ammonium chloride (2 or 5 mM).

Labeling from [U-13C]glucose in glutamate was extensive,

with double labeled glutamate being the most abundant

isotopomer (Fig. 5A). Mono labeling in glutamate decreased

Table 1

Amount of 13C-labeled glutamate isotopomers (nmol/mg protein) in cultured cerebellar neurons or astrocytes incubated for 5 h in the presence of either 2.5 mM

[U-13C]glucose and 1 mM isoleucine or [U-13C]isoleucine and glucose

Amount of 13C-labeled glutamate (nmol/mg of protein)

Precursor [U-13C]glucose

Cell type Condition Mono Double Triple Quadruple Uniform

Neurons Control 5.0 � 0.2 12.5 � 0.4 8.3 � 0.2 5.8 � 0.3 3.3 � 0.2

5 mM NH4+ 2.2 � 0.2 12.0 � 0.8 14.1 � 0.9* 18.4 � 1.2* 22.8 � 2.0*

Astrocytes Control 2.4 � 0.1 7.3 � 0.3 2.8 � 0.1 1.6 � 0.1 0.7 � 0.04

2 mM NH4+ 2.1 � 0.1 9.7 � 0.2* 3.9 � 0.1* 4.2 � 0.1* 1.7 � 0.05*

5 mM NH4+ 1.9 � 0.1 10.0 � 0.6* 4.2 � 0.2* 5.0 � 0.3* 2.1 � 0.1*

Precursor [U-13C]isoleucine

Cell type Condition Mono Double Triple Quadruple Uniform

Neurons Control 2.8 � 0.2 2.4 � 0.2 0.4 � 0.03 – –

5 mM NH4+ 1.3 � 0.1* 1.3 � 0.1* 0.1 � 0.01 – –

Astrocytes Control 1.4 � 0.1 3.1 � 0.3 0.5 � 0.04 – –

2 mM NH4+ 1.9 � 0.1 4.2 � 0.3* 0.6 � 0.1 – –

5 mM NH4+ 1.6 � 0.1 3.2 � 0.2� 0.6 � 0.03 – –

Cultured cerebellar neurons or astrocytes were incubated for 5 h in serum-free culture medium containing either 2.5 mM [U-13C]glucose and 1 mM isoleucine or

[U-13C]isoleucine and glucose, as detailed in Section 2. Cell extracts were analyzed by HPLC and LC–MS for amount of intracellular glutamate and 13C-labeling,

respectively (see Section 2). Results are amounts of 13C-labeled intracellular glutamate isotopomers (nmol/mg protein) expressed as the average � S.E.M. of 3–9

individual values. (*): significantly different from the control condition (P < 0.05); (�): significantly different from the values obtained in the presence of 2 mM

ammonium (P < 0.05).

M.L. Johansen et al. / Neurochemistry International 50 (2007) 1042–1051 1047

whereas quadruple and uniform labeling increased with

increasing ammonium concentrations (2 and 5 mM;

Fig. 5A), suggesting increased TCA cycling. In addition,

double and triple labeling in glutamate increased as a

consequence of ammonium treatment; however, no differences

were observed between 2 and 5 mM ammonium. Increased

double and triple labeling along with quadruple and uniform

labeling may indicate increased PC activity as a consequence of

Fig. 4. Cultured cerebellar neurons were incubated in serum-free culture

medium in the presence of [U-13C]isoleucine (1 mM) and glucose (2.5 mM)

for 5 h. The effect of 5 mM ammonium chloride was investigated (black bars;

see Section 2 for details). Cell extracts were analyzed by mass spectrometry for

percent 13C-labeling of glutamate (A) and aspartate (B). The results are

averages of 4–5 cultures � S.E.M. (*): significantly different from the control

condition (P < 0.05).

treatment with ammonium (see Fig. 2 and Section 4.2). In

contrast to what was observed in neurons, uniform labeling in

alanine did not increase upon ammonium treatment indicating

that glycolysis was not up-regulated in these cells (Fig. 5B).

Aspartate mimicked glutamate showing ammonium-dependent

increases in double and uniform labeling, although only

modestly in double labeling (Fig. 5C). Glutamine, which is

formed from glutamate, showed ammonium-dependent

increases in label for quadruple and uniformly labeled

isotopomers (Fig. 5D). Amounts of double, triple, quadruple

and uniformly labeled glutamate were increased in the presence

of ammonium (Table 1).

Table 2

Amount of amino acids in cultured cerebellar neurons incubated for 5 h in the

presence of either 2.5 mM [U-13C]glucose and 1 mM isoleucine or [U-13C]iso-

leucine and glucose

Metabolite Content (nmol/mg protein)

Control 5 mM NH4+

Glutamate 42.5 � 1.8 68.7 � 2.8*

Aspartate 10.7 � 0.5 15.4 � 0.7*

Alanine 17.2 � 0.8 20.9 � 1.0*

Leucine 7.0 � 0.7 7.8 � 0.2

Isoleucine 24.3 � 2.7 27.1 � 1.0

Valine 8.8 � 1.0 10.1 � 0.6

Cultured cerebellar neurons were incubated for 5 h in serum-free culture

medium containing either 2.5 mM [U-13C]glucose and 1 mM isoleucine or

[U-13C]isoleucine and glucose, as detailed in Section 2. Cell extracts were

analyzed by HPLC for amount of intracellular amino acids (See Section 2). As

the experimental conditions were the same regardless of whether glucose or

isoleucine was labeled, results are shown as the average � S.E.M. of all (9–14)

cultures contained in both groups. (*): significantly different from the control

condition (P < 0.05).

Fig. 5. Cultured cerebellar astrocytes were incubated in serum-free culture

medium in the presence of [U-13C]glucose (2.5 mM) and isoleucine (1 mM) for

5 h. The effect of 2 (white bars) or 5 mM ammonium chloride was investigated

(black bars; see Section 2 for details). Cell extracts were analyzed by mass

spectrometry for percent 13C-labeling of glutamate (A), alanine (B), aspartate

(C) and glutamine (D). The results are averages of 4–5 cultures � S.E.M. (*):

significantly different from the control condition (P < 0.05); (�): significantly

different from cultures incubated in the presence of 2 mM ammonium

(P < 0.05).

Fig. 6. Cultured cerebellar astrocytes were incubated in serum-free culture

medium in the presence of [U-13C]isoleucine (1 mM) and glucose (2.5 mM) for

5 h. The effect of 2 (white bars) or 5 mM ammonium chloride was investigated

(black bars; see Section 2 for details). Cell extracts were analyzed by mass

spectrometry for percent 13C-labeling of glutamate (A), aspartate (B) and

glutamine (C). The results are averages of 4–5 cultures � S.E.M. (*): signifi-

cantly different from the control condition (P < 0.05); (�): significantly

different from cultures incubated in the presence of 2 mM ammonium

(P < 0.05).

M.L. Johansen et al. / Neurochemistry International 50 (2007) 1042–10511048

Label from [U-13C]isoleucine in glutamate was most

pronounced for double labeling, with mono and triple labeled

glutamate being about one half and one fourth of that observed

for double labeling, respectively (Fig. 6A). Double labeling in

aspartate was most significant, compared to mono and triple

labeling (Fig. 6B). The double and triple labeled isotopomers of

aspartate decreased slightly upon ammonium treatment

(5 mM). Interestingly, both mono and double labeling in

glutamine increased as an effect of ammonium treatment

(Fig. 6C), which was in contrast to glutamate. This indicates

that glutamine is synthesized from a glutamate pool different

from the average glutamate pool, and that glutamine may be

produced from the isoleucine carbon skeleton in the presence of

ammonium. Compared to percent labeling, the amount of

double labeled glutamate actually increased in the presence of

2 mM ammonium compared to both the control and the

presence of 5 mM ammonium (Table 1).

The amount of glutamate, glutamine and alanine increased

as an effect of ammonium treatment at both 2 and 5 mM of

ammonium (Table 3), whereas aspartate increased at 2 mM

ammonium but decreased at 5 mM ammonium as compared to

the control value. The amount of extracellular glutamine was

doubled, from 107.7 � 5.3 to 226.0 � 17.8 nmol/mg protein

(P < 0.05), as a consequence of incubation in the presence of

2 mM ammonium.

3.4. Comparison of [U-13C]isoleucine or [U-13C]glucose

metabolism in neurons and astrocytes

To obtain a comparable measure of the extent of

[U-13C]isoleucine or [U-13C]glucose metabolism in neurons

Table 3

Amount of amino acids in cultured cerebellar astrocytes incubated for 5 h in the

presence of either 2.5 mM [U-13C]glucose and 1 mM isoleucine or [U-13C]iso-

leucine and glucose

Metabolite Content (nmol/mg protein)

Control 2 mM NH4+ 5 mM NH4

+

Glutamate 24.8 � 0.4 38.6 � 2.0* 32.3 � 1.2*

Glutamine 3.1 � 0.3 7.3 � 0.5* 8.0 � 0.8*

Aspartate 15.1 � 0.7 23.4 � 1.0* 6.3 � 0.2*,�

Alanine 4.5 � 0.2 5.6 � 0.3* n.d.

Leucine 3.6 � 0.2 4.4 � 0.5 n.d.

Isoleucine 11.8 � 0.6 14.2 � 1.7 n.d.

Valine 4.3 � 0.2 5.1 � 0.5 n.d.

Cultured cerebellar astrocytes were incubated for 5 h in serum-free culture

medium containing either 2.5 mM [U-13C]glucose and 1 mM isoleucine or

[U-13C]isoleucine and glucose, as detailed in Section 2. Cell extracts were

analyzed by HPLC for amount of intracellular amino acids (See Section 2). As

the experimental conditions were the same regardless of whether glucose or

isoleucine was labeled, results are shown as the average � S.E.M. of all (4–10)

cultures contained in both groups. (*), significantly different from the control

condition (P < 0.05); (�): significantly different from cultures incubated in the

presence of 2 mM ammonium (P < 0.05). n.d., not determined.

M.L. Johansen et al. / Neurochemistry International 50 (2007) 1042–1051 1049

and astrocytes, MCL values (see Section 2.5) were calculated

and presented in Fig. 7. [U-13C]glucose (Fig. 7A) was

metabolized to a much higher extent than [U-13C]isoleucine

(Fig. 7B). Furthermore, metabolism of [U-13C]isoleucine

tended to decrease in the presence of ammonium whereas

Fig. 7. Cultured cerebellar neurons or astrocytes were incubated in serum-free

culture medium in the presence of either [U-13C]isoleucine (1 mM) and glucose

(2.5 mM) (A) or [U-13C]glucose and isoleucine (B) for 5 h. For neurons, the

effect of 5 (white bars), and for astrocytes 2 and 5 mM ammonium was

investigated (white and black bars, respectively). Results are averages of

percent molecular carbon labeling (MCL) � S.E.M. for intracellular glutamate

(see Section 2 for details). (*): significantly different from the control condition

(P < 0.05); (�): significantly different from cultures incubated in the presence

of 2 mM ammonium (P < 0.05).

that of [U-13C]glucose increased in both neurons and

astrocytes, although the changes were most pronounced in

the neurons.

4. Discussion

Cerebral metabolic disturbances caused by hyperammone-

mia which may be related to the development of HE have so far

not been fully elucidated. It was shown more than 4 decades

ago, that ammonium inhibits oxygen consumption in brain

slices and isolated mitochondria (McKhann and Tower, 1961)

and, in addition, it has been suggested that the malate-aspartate

shuttle may be inhibited due to a lowered amount of cytosolic

glutamate (e.g. Hindfelt et al., 1977). Two hypotheses evolving

around ammonium-induced inhibition of enzymes have been

put forward, namely inhibition of a-KGDH (Lai and Cooper,

1986) or PDH (Zwingmann et al., 2003), resulting in inhibition

of TCA cycle flux or lactate production, respectively. It was

recently suggested by Zwingmann et al. (2003) that lactate,

rather than glutamine accumulation was the cause of the brain

edema observed in acute liver failure induced by initial

portacaval anastomosis followed by ligation of the hepatic

artery in rats. Accumulation of glutamine as the cause of brain

edema has been a longstanding dogma (Albrecht and Dolinska,

2001). However, inhibition of energy metabolism, i.e.

inhibition of the TCA cycle or the PDH reaction, may explain

some of the cerebral impairments observed during hyper-

ammonemia. Catabolism of isoleucine results in formation of

both succinyl-CoA (a TCA cycle intermediate) and acetyl-

CoA, thus supplying both an anaplerotic substrate down-stream

of the a-KGDH step as well as acetyl-CoA for synthesis of

citrate from oxaloacetate by citrate synthase (see Fig. 1). As

recently pointed out by Ott et al. (2005), this may serve a dual

purpose: (i) it bypasses the potential ammonium-dependent

inhibition of both a-KGDH and PDH; and (ii) it provides the

carbon skeleton for detoxification of ammonium via synthesis

of glutamate and glutamine.

4.1. Effects of ammonium on neuronal metabolism of

glucose or isoleucine

The present results suggest that neuronal glycolysis was

increased as a consequence of ammonium treatment due to an

increase in labeling of alanine. This may be consistent with a

direct effect of ammonium on glycolytic enzymes (Clarke and

Sokoloff, 1999). In contrast, a recent study in cultured cells

found no increase in glycolysis in cultured (GABAergic)

interneurons (Kala and Hertz, 2005); however, this may be due

to differences between cultured cerebellar (glutamatergic)

neurons and cortical GABAergic interneurons. One significant

difference between these two cell types is the activity of GDH,

the activity being highest in cerebellar neurons (Zaganas et al.,

2001). As the mitochondrial GDH reaction produces NAD(P)+

when operating in the direction of reductive amination

producing glutamate from a-ketoglutarate, this may increase

TCA cycle metabolism by supplying NAD+ substrates for the

TCA cycle dehydrogenases. In turn, this consumes pyruvate/

M.L. Johansen et al. / Neurochemistry International 50 (2007) 1042–10511050

acetyl-CoA generated from glucose contributing to the

observed increase in glycolysis. Alternatively, the increase

in labeled alanine may be interpreted as inhibition of PDH;

however, this is unlikely as inhibition of neither the PDH

reaction nor the TCA cycle at the a-KGDH step was apparent.

This is evidenced by the fact that both glutamate and aspartate,

respectively, exhibited ammonium-dependent increases in

labeling patterns from [U-13C]glucose associated with

subsequent turns of the TCA cycle. In support of this, a

previous study performed in cultured neurons found no

inhibition by acute ammonium (3 mM) treatment of 14CO2

production from [U-14C]glucose (Hertz et al., 1987). In

addition, the amount of glutamate was increased following

ammonium treatment, consistent with glutamate synthesis via

the GDH reaction, as was observed in a previous study by

Yudkoff et al. (1990). Hypothetically, the synthesis of

glutamate may to some extent be supported by the anaplerotic

action of isoleucine metabolism. Mono and double labeling

from [U-13C]isoleucine of both glutamate and aspartate

decreased as an effect of ammonium treatment. This is

probably an effect of simultaneous increased metabolism of

the unlabeled glucose or lactate present in the incubation

medium. Furthermore, percent triple labeling in aspartate,

most likely formed from [U-13C]isoleucine-derived triple

labeled oxaloacetate, did not decrease in the face of the

ammonium-challenge. However, the amount (in contrast to

percent) of triple labeled aspartate in fact increased, indicating

that catabolism of isoleucine supports formation of TCA cycle

intermediates during exposure to ammonium.

The incorporation of 13C from [U-13C]isoleucine and

[U-13C]glucose into glutamate may be compared using the

MCL values. The MCL of glutamate when [U-13C]glucose was

the precursor was 10 times higher than when [U-13C]isoleucine

was the precursor. Ammonium had an opposite effect on the

MCL of glutamate produced from the two labeled precursors,

i.e. decreasing and increasing for [U-13C]isoleucine and

[U-13C]glucose, respectively. The comparison is complicated

by the fact that the increase of glucose metabolism in the

presence of ammonium decreases the labeling from [U-13C]iso-

leucine which might be evident from the size of the relative

effect of ammonium on the incorporation from the two

precursors.

4.2. Effects of ammonium on astrocytic metabolism of

glucose or isoleucine

Interestingly, alanine labeling from [U-13C]glucose was

unaffected by ammonium treatment, showing that glycolysis

was not up-regulated in astrocytes in the face of an

ammonium challenge. This was in contrast to the observation

in neurons. Moreover, anaplerosis by formation of oxaloa-

cetate by the astrocytic PC reaction may be increased, as (i)

the intracellular amount of amino acids as well as the amount

of glutamine in the incubation medium increased, and (ii) the

relative amount of double, triple, quadruple and uniformly

labeled glutamate increased. Under the same conditions in

neurons in which pyruvate carboxylation is absent no

increase was observed in double and triple labeling of

glutamate. These observations might reflect that glutamate

generated in astrocytes down-stream of [1,2,3-13C]/

[2,3,4-13C]oxaloacetate is de novo synthesized from uni-

formly labeled pyruvate. Labeling in aspartate from

[U-13C]glucose increased as an effect of ammonium

treatment, albeit at a lower extent compared to glutamate.

This may indicate that the astrocytic TCA cycle is slightly

inhibited at the a-KGDH step by exposure to ammonium, as

labeling of aspartate takes place after this step. Isoleucine

was metabolized to a lower extent than glucose, with only

minor decreasing effects of ammonium treatment in

glutamate and aspartate labeling suggesting that isoleucine

is employed as an anaplerotic substrate during both control

conditions and hyperammonemia, cf. the reasoning in Section

4.1 that [U-13C]isoleucine sustains the ability to label

glutamate/aspartate in the face of increased metabolism of

unlabeled acetyl-CoA.

The labeling pattern of glutamine was similar to that of

glutamate from [U-13C]glucose and the effects of ammonium

were similar although more pronounced with regard to

glutamate. This indicates that the pools of glutamate and

glutamine labeled from glucose are not compartmentalized. In

contrast, the effect of ammonium on labeling of glutamine, i.e.

increased mono and double labeling, was different from that of

glutamate when isoleucine was the labeled precursor indicat-

ing compartmentation of glutamate and glutamine. In addition,

it might be tempting to suggest that isoleucine to some extent

particularly supports synthesis of glutamine during elevated

ammonium. Compartmentation of cellular glutamate is not

surprising, given the fact that astrocytes show a significant

degree of metabolic compartmentation (e.g. Schousboe et al.,

1993; Waagepetersen et al., 2001, 2006; Sickmann et al.,

2005).

4.3. Summary

The focus of this work was: (i) to detect any significant

inhibitory effect of hyperammonemia on TCA cycle metabo-

lism; and (ii) to determine the ability of isoleucine to support

TCA cycle metabolism and formation of glutamate/glutamine

in the face of an ammonium challenge. To conclude: (i) only

slight indications for inhibition of astrocytic TCA cycle

metabolism was obtained in the experimental paradigm

employed, whereas neuronal TCA cycling was increased;

and (ii) isoleucine was metabolized in both neurons and

astrocytes serving a role as anaplerotic substrate. Hence, the

carbon skeleton of isoleucine was employed for glutamate and

aspartate synthesis in neurons and glutamate, aspartate and

glutamine synthesis in astrocytes, thus aiding in detoxification

of the ammonium.

Based on these findings, the role of branched-chain amino

acids in the treatment and pathology of cerebral hyperammo-

nemia/HE seems to be worth further investigations. One

pertinent future goal is to substantiate the present findings in an

in vivo model of chronic or acute cerebral hyperammonemia,

and such studies are presently underway.

M.L. Johansen et al. / Neurochemistry International 50 (2007) 1042–1051 1051

Acknowledgements

Mr. Lars Evje, M.Sc. (Norwegian University of Science and

Technology) and Ms. Lene Vigh (Danish University of

Pharmaceutical Sciences) are cordially acknowledged for their

expert technical assistance. The Lundbeck, Alfred Benzons and

Hørslev Foundations as well as the Danish Medical Research

Council (grants 22-04-0314 and 22-03-0250) are cordially

acknowledged for generous funding.

References

Albrecht, J., Dolinska, M., 2001. Glutamine as a pathogenic factor in hepatic

encephalopathy. J. Neurosci. Res. 65, 1–5.

Bak, L.K., Schousboe, A., Sonnewald, U., Waagepetersen, H.S., 2006. Glucose

is necessary to maintain neurotransmitter homeostasis during synaptic

activity in cultured glutamatergic neurons. J. Cereb. Blood Flow Metab.

26, 1285–1297.

Biemann, K., 1962. The mass spectra of isotopically labeled molecules, in Mass

spectrometry; Organic chemical applications. McGraw-Hill, New York, pp.

223–227.

Clarke, D.D., Sokoloff, L., 1999. Circulation and energy metabolism of the

brain. In: Siegel, G.J., Agranoff, B.W., Fisher, S.K., Albers, R.W., Uhler,

M.D. (Eds.), Basic Neurochemistry. Lippincott-Raven Publishers, Phila-

delphia, USA, pp. 637–669.

Drejer, J., Larsson, O.M., Kvamme, E., Svenneby, G., Hertz, L., Schousboe, A.,

1985. Ontogenetic development of glutamate metabolizing enzymes in

cultured cerebellar granule cells and in cerebellum in vivo. Neurochem.

Res. 10, 49–62.

Drejer, J., Schousboe, A., 1989. Selection of a pure cerebellar granule cell

culture by kainate treatment. Neurochem. Res. 14, 751–754.

Fitzpatrick, S.M., Hetherington, H.P., Behar, K.L., Shulman, R.G., 1990. The

flux from glucose to glutamate in the rat brain in vivo as determined by 1H-

observed, 13C-edited NMR spectroscopy. J. Cereb. Blood Flow Metab. 10,

170–179.

Geddes, J.W., Wood, J.D., 1984. Changes in the amino acid content of nerve

endings (synaptosomes) induced by drugs that alter the metabolism of

glutamate and gamma-aminobutyric acid. J. Neurochem. 42, 16–24.

Hawkins, R.A., Peterson, D.R., Vina, J.R., 2002. The complementary mem-

branes forming the blood-brain barrier. IUBMB. Life 54, 101–107.

Hertz, L., Juurlink, B.H.J., Fosmark, H., Schousboe, A., 1982. Astrocytes in

primary cultures. In: Pfeiffer, S.E. (Ed.), Neuroscience Approached through

Cell Culture. CRC Press, Boca Raton, pp. 175–186.

Hertz, L., Murthy, C.R., Lai, J.C., Fitzpatrick, S.M., Cooper, A.J., 1987. Some

metabolic effects of ammonia on astrocytes and neurons in primary cultures.

Neurochem. Pathol. 6, 97–129.

Hertz, L., Juurlink, B.H.J., Hertz, E., Fosmark, H., Schousboe, A., 1989.

Preparation of primary cultures of mouse (rat) astrocytes. In: Shahar ,

A., de Vellis, J., Vernadakis, A., Haber, B. (Eds.), A Dissection and Tissue

Culture Manual for the Nervous System. Liss, New York, pp. 105–108.

Hindfelt, B., Plum, F., Duffy, T.E., 1977. Effect of acute ammonia intoxication

on cerebral metabolism in rats with portacaval shunts. J. Clin. Invest. 59,

386–396.

Juurlink, B.H., Hertz, L., 1985. Plasticity of astrocytes in primary cultures: an

experimental tool and a reason for methodological caution. Dev. Neurosci.

7, 263–277.

Kala, G., Hertz, L., 2005. Ammonia effects on pyruvate/lactate production in

astrocytes-interaction with glutamate. Neurochem. Int. 47, 4–12.

Lai, J.C., Cooper, A.J., 1986. Brain alpha-ketoglutarate dehydrogenase com-

plex: kinetic properties, regional distribution, and effects of inhibitors. J.

Neurochem. 47, 1376–1386.

Lowry, O.H., Rosebrough, N.J., Farr, A.L., Randall, R.J., 1951. Protein

measurement with the Folin phenol reagent. J. Biol. Chem. 193, 265–

275.

Mason, G.F., Rothman, D.L., Behar, K.L., Shulman, R.G., 1992. NMR deter-

mination of the TCA cycle rate and alpha-ketoglutarate/glutamate exchange

rate in rat brain. J. Cereb. Blood Flow Metab. 12, 434–447.

McKhann, G.M., Tower, D.B., 1961. Ammonia toxicity and cerebral oxidative

metabolism. Am. J. Physiol. 200, 420–424.

Norenberg, M.D., Martinez-Hernandez, A., 1979. Fine structural localization

of glutamine synthetase in astrocytes of rat brain. Brain Res. 161, 303–

310.

Ott, P., Clemmesen, O., Larsen, F.S., 2005. Cerebral metabolic disturbances in

the brain during acute liver failure: from hyperammonemia to energy failure

and proteolysis. Neurochem. Int. 47, 13–18.

Schousboe, A., Meier, E., Drejer, J., Hertz, L., 1989. Preparation of primary

cultures of mouse (rat) cerebellar granule cells. In: Shahar, A., de Vellis, J.,

Vernadakis, A., Haber, B. (Eds.), A Dissection and Tissue Culture Manual

for the Nervous System. Liss, New York, pp. 183–186.

Schousboe, A., Westergaard, N., Sonnewald, U., Petersen, S.B., Huang, R.,

Peng, L., Hertz, L., 1993. Glutamate and glutamine metabolism and

compartmentation in astrocytes. Dev. Neurosci. 15, 359–366.

Sickmann, H.M., Schousboe, A., Fosgerau, K., Waagepetersen, H.S., 2005.

Compartmentation of lactate originating from glycogen and glucose in

cultured astrocytes. Neurochem. Res. 30, 1295–1304.

Sonnewald, U., Kortner, T.M., Qu, H., Olstad, E., Sunol, C., Bak, L.K.,

Schousboe, A., Waagepetersen, H.S., 2006. Demonstration of extensive

GABA synthesis in the small population of GAD positive neurons in

cerebellar cultures by the use of pharmacological tools. Neurochem. Int.

48, 572–578.

Sonnewald, U., Olstad, E., Qu, H., Babot, Z., Cristofol, R., Sunol, C., Schous-

boe, A., Waagepetersen, H., 2004. First direct demonstration of extensive

GABA synthesis in mouse cerebellar neuronal cultures. J. Neurochem. 91,

796–803.

Waagepetersen, H.S., Hansen, G.H., Fenger, K., Lindsay, J.G., Gibson, G.,

Schousboe, A., 2006. Cellular mitochondrial heterogeneity in cultured

astrocytes as demonstrated by immunogold labeling of alpha-ketoglutarate

dehydrogenase. Glia 53, 225–231.

Waagepetersen, H.S., Sonnewald, U., Larsson, O.M., Schousboe, A., 2000. A

possible role of alanine for ammonia transfer between astrocytes and

glutamatergic neurons. J. Neurochem. 75, 471–479.

Waagepetersen, H.S., Sonnewald, U., Larsson, O.M., Schousboe, A., 2001.

Multiple compartments with different metabolic characteristics are involved

in biosynthesis of intracellular and released glutamine and citrate in

astrocytes. Glia 35, 246–252.

Yudkoff, M., 2006. Disorders of amino acid metabolism. In: Siegel, G.J.,

Albers, R.W., Brady, S.T., Price, D.L. (Eds.), Basic Neurochemistry. 7th

ed. Elsevier, Amsterdam, pp. 667–683.

Yudkoff, M., Nissim, I., Hertz, L., 1990. Precursors of glutamic acid nitrogen in

primary neuronal cultures: studies with 15N. Neurochem. Res. 15, 1191–

1196.

Zaganas, I., Waagepetersen, H.S., Georgopoulos, P., Sonnewald, U., Plaitakis,

A., Schousboe, A., 2001. Differential expression of glutamate dehydrogen-

ase in cultured neurons and astrocytes from mouse cerebellum and cerebral

cortex. J. Neurosci. Res. 66, 909–913.

Zwingmann, C., Chatauret, N., Leibfritz, D., Butterworth, R.F., 2003. Selective

increase of brain lactate synthesis in experimental acute liver failure: results

of a [H–C] nuclear magnetic resonance study. Hepatology 37, 420–428.