Neurochemistry International 60 (2012) 690–696

Contents lists available at SciVerse ScienceDirect

Neurochemistry International
journal homepage: www.elsevier.com/locate/nci

Review

Pathogenesis of hepatic encephalopathy and brain edema in acute liver failure:

Role of glutamine redefined
Paul Desjardins a, Ting Du b, Wenlei Jiang a, Liang Peng b, Roger F. Butterworth a,⇑
a
Neuroscience Research Unit, Saint-Luc Hospital (CHUM), University of Montreal, Montreal, Quebec, Canada
b
Department of Clinical Pharmacology, China Medical University, Shenyang, PR China

a r t i c l e i n f o a b s t r a c t

Article history: Acute liver failure (ALF) is characterized neuropathologically by cytotoxic brain edema and biochemically
Available online 21 February 2012 by increased brain ammonia and its detoxification product, glutamine. The osmotic actions of increased
glutamine synthesis in astrocytes are considered to be causally related to brain edema and its complications
Keywords: (intracranial hypertension, brain herniation) in ALF. However studies using multinuclear 1H- and 13C-NMR
Acute liver failure spectroscopy demonstrate that neither brain glutamine concentrations per se nor brain glutamine synthesis
Ammonia rates correlate with encephalopathy grade or the presence of brain edema in ALF. An alternative mechanism
Glutamine
is now proposed whereby the newly synthesized glutamine is trapped within the astrocyte as a conse-
Small neutral amino acid transporters
Brain edema
quence of down-regulation of its high affinity glutamine transporter SNAT5 in ALF. Restricted transfer
Hepatic encephalopathy out of the cell rather than increased synthesis within the cell could potentially explain the cell swelling/
In vivo cerebral microdialysis brain edema in ALF. Moreover, the restricted transfer of glutamine from the astrocyte to the adjacent gluta-
NMR spectroscopy matergic nerve terminal (where glutamine serves as immediate precursor for the releasable/transmitter
pool of glutamate) could result in decreased excitatory transmission and excessive neuroinhibition that
is characteristic of encephalopathy in ALF. Paradoxically, in spite of renewed interest in arterial ammonia
as a predictor of raised intracranial pressure and brain herniation in ALF, ammonia-lowering agents aimed
at reduction of ammonia production in the gut have so far been shown to be of limited value in the preven-
tion of these cerebral consequences. Mild hypothermia, shown to prevent brain edema and intracranial
hypertension in both experimental and human ALF, does so independent of effects on brain glutamine syn-
thesis; whether or not hypothermia restores expression levels of SNAT5 in ALF awaits further studies. While
inhibitors of brain glutamine synthesis such as methionine sulfoximine, have been proposed for the preven-
tion of brain edema in ALF, potential adverse effects have so far limited their applicability.
Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction been reported in experimental animal models of ALF at coma/ede-

Hepatic Encephalopathy (HE) and brain edema and its complica-
tions (intracranial hypertension, brain herniation) remain serious
cerebral complications of acute liver failure (ALF). The occurrence
of HE in ALF heralds the need for liver transplantation and compli-
cations of brain edema remain one of the major causes of mortality
in ALF patients. The precise pathophysiological mechanisms
responsible for these complications have not been completely eluci-
dated. However, studies related to the cytotoxic effects of ammonia
continue to be a focus of research in this area. Hyperammonemia
occurs in >90% of ALF patients (Ferenci et al., 1996; Brusilow et al.,
2010) and brain ammonia concentrations in the 1–5 mM range have
ma stages of encephalopathy (Swain et al., 1992; Mans et al., 1979).
The advent of modern techniques such as in vivo cerebral micro-
dialysis, multi-nuclear magnetic resonance (NMR) spectroscopy
and related molecular approaches have provided new impetus to
studies on pathophysiologic mechanisms in ALF, helping to resolve
existing controversies and to point the way to new diagnostic and
therapeutic strategies in this area of investigation. The current re-
view is both a critical appraisal of the role of intracellular accumu-
lation of glutamine in the pathogenesis of HE and brain edema in
ALF and also an attempt to identify new areas of investigation
and novel treatment opportunities for ALF.

2. Arterial ammonia and brain edema in acute liver failure

⇑ Corresponding author. Address: Neuroscience Research Unit, CHUM, Campus Although not reliable for diagnosis purposes due undoubtedly
Saint-Luc, University of Montreal, 1058 St-Denis Street, Montreal, Quebec, Canada to difficulties in measurement, arterial ammonia concentrations
H2X 3J4. Tel.: +1 514 890 8000x35759; fax: +1 514 412 7377. in excess of 300 lM are reliable predictors of brain herniation in
E-mail address: roger.butterworth@umontreal.ca (R.F. Butterworth). ALF patients (Clemmesen et al., 1999; Bernal et al., 2007). In a

0197-0186/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.neuint.2012.02.001
 P. Desjardins et al. / Neurochemistry International 60 (2012) 690–696
recent study using the in vivo microdialysis technique in 17 pa-
tients with ALF resulting primarily from acetaminophen overdose,
a significant positive correlation between arterial ammonia con-
centrations and intracranial pressure was observed with r = 0.73,
p < 0.05 (Fig. 1) (Tofteng et al., 2006) further suggestive of a role
for ammonia in the pathogenesis of this complication of brain ede-
ma in these patients.
3. Brain glutamine in acute liver failure
Two enzyme systems are theoretically capable of removing
ammonia in brain. These enzymes are glutamate dehydrogenase
(GDH) and glutamine synthetase (GS). GDH catalyses the conver-
sion of alpha-ketoglutarate to glutamate, a reaction shown to occur
in both neurons and astrocytes, according to reaction (1); GS catal-
yses the conversion of glutamate to glutamine, a reaction that oc-
curs primarily if not exclusively in astrocytes (Norenberg, 1979),
according to reaction (2):
Alpha  ketoglutarate þ NHþ4 þ NADðPÞH þ Hþ ¢ Glutamate
þ NADðPÞþ þ H2 O ð1Þ
NH3 þ Glutamate þ ATP ! Glutamine þ ADP þ Pi ð2Þ
Studies using 13N-ammonia have shown convincingly that
blood-borne ammonia is incorporated almost exclusively into glu-
tamine by reaction (2) in the brain under both normal and hyper-
ammonemic conditions (Cooper et al., 1985). Consequently it is the
astrocyte that is called upon to remove both blood-borne and lo-
cally synthesized ammonia in the brain in ALF. It is not surprising,
therefore, that brain glutamine concentrations have consistently
been found to be elevated up to 5-fold in experimental ALF using
classical biochemical procedures (Traber et al., 1989; Swain et al.,
1992; Mans et al., 1994) and subsequently confirmed by 1H-NMR
spectroscopy (Zwingmann et al., 2003). Both CSF and brain concen-
trations of glutamine are also increased in patients with ALF due to
acetaminophen hepatotoxicity (Record et al., 1976). More recently,
using the technique of in vivo cerebral microdialysis, glutamine
concentrations were found to be increased up to 5-fold in brain
extracellular fluid starting early (prior to the onset of HE and brain
edema) and remaining high until coma/edema stages of HE (Fig. 2)
Fig. 1. Correlation between arterial ammonia concentration (lM) and intracranial
pressure (mm Hg) in patients (n = 17) with fulminant hepatic failure. Regression
line and 95% confidence intervals are shown (p < 0.05 ; r = 0.73) (adapted from
Tofteng et al., 2006).
691
(Michalak et al., 1996). Although these findings (as well as those
reported by Zwingmann et al., 2003) support a role for glutamine
in the pathogenesis of early encephalopathy in ALF, measurement
of intracellular (astrocytic) concentrations of glutamine may be re-
quired in order to assess its role in the pathogenesis of severe
encephalopathy (coma) or brain edema in ALF. The issue was par-
tially addressed using in vivo cerebral microdialysis to measure
extracellular brain glutamine in patients with ALF (Tofteng et al.,
2006). Patients with high intracranial pressure (ICP) were found
to have significantly higher brain glutamine compared to patients
with normal ICP (with increased ICP: 6.54 lM, without increased
ICP: 3.52 lM; p < 0.05) (Fig. 3). The authors suggested that the rise
in glutamine was a prerequisite for the development of increased
cerebral blood flow and high ICP in ALF patients.
4. Brain ammonia removal (glutamine synthesis) in acute liver
failure
Although increased brain glutamine is a consistent feature of
both experimental and human ALF, activities of GS are either un-
changed or slightly reduced in acute hyperammonemia (Cooper
et al., 1985) and in experimental ALF in the rat resulting from he-
patic devascularisation where a small but significant 18% reduction
in GS activity was noted in cerebral cortex (Chatauret et al., 2006).
A limited capacity for brain glutamine synthesis via GS was con-
cluded by Deutz et al. (1988) who noted a faster rate of increase
of brain ammonia compared to brain glutamine in an ALF model
and by Bosman et al. (1990) who observed that increased gluta-
mine synthesis was not sustained in experimental ALF and, at
coma/edema stages of HE, brain glutamine synthesis was actually
decreased. Similar findings were reported by Michalak et al.
(1996). Finally, a lack of correlation between HE grade and
in vivo GS activities was reported by Kanamori et al. (1996) using
1
H-NMR spectroscopy in a hyperammonemia model. One possible
explanation for the observed loss of GS activity in brain in ALF may
relate to inactivation of the enzyme due to protein tyrosine nitra-
tion. Exposure of primary astrocyte cultures to millimolar concen-
trations of ammonia results in significant nitration of GS protein
leading to loss of enzyme activity (Schliess et al., 2002). However,
whether or not such mechanisms are applicable to GS loss in brain
in ALF remains a matter of conjecture. The findings of a lack of
induction of GS in brain underscores the vulnerability of the brain
to hyperammonemic conditions in general and to ALF, in
particular.
In contrast to brain, skeletal muscle of ALF animals manifested a
significant 27% increase of GS activity resulting from a post-trans-
lational induction of GS expression (Fig. 4). Increased de novo
synthesis of glutamine in muscle of ALF animals was confirmed
400
* *
Glutamine (µM)
300
*
200
100
0
SHAM ALF-6h PRECOMA COMA
Fig. 2. Extracellular concentrations of glutamine in the frontal cortex of rats with
ALF resulting from hepatic devascularization as a function of progression of
encephalopathy compared to sham-operated controls (⁄p < 0.05 vs. sham-operated
controls) (adapted from Michalak et al., 1996).
692 P. Desjardins et al. / Neurochemistry International 60 (2012) 690–696

been undertaken in brain in experimental ALF using the technique

Fig. 3. Correlation between extracellular glutamine concentration (lM) and
intracranial pressure (mm Hg) in patients (n = 17) with fulminant hepatic failure.
Regression line and 95% confidence intervals are shown (p < 0.05; r = 0.58) (adapted
from Tofteng et al., 2006).
by 13C-NMR spectroscopy in this study (Chatauret et al., 2006).
Skeletal muscle also plays an important role in the detoxification
of ammonia to glutamine in patients with liver disease (Lockwood
et al., 1979).
These findings provide a molecular explanation for the view
that, in ALF, skeletal muscle becomes the major organ responsible
for ammonia detoxification and suggests fundamentally distinct
mechanisms for GS regulation in muscle and brain.
Although measurement of activities of GS in vitro may be indic-
ative of metabolic capacity, measurement of metabolic fluxes
through the enzymic pathway are required in order to adequately
assess the dynamics of glutamine synthesis. Such studies have
GS expression (units)
of 13C-NMR. Using this approach, it was demonstrated that the
synthesis of glutamine from 13C-glucose via both the pyruvate
dehydrogenase and pyruvate carboxylase metabolic pathways
was increased at early time points following onset of ALF but that
this increase was not maintained at coma/edema time points again
suggesting a limit on the capacity of the astrocytes to remove
ammonia in ALF (Fig. 5) (Zwingmann et al., 2003).
As with the findings of a lack of correlation between brain glu-
tamine levels and HE severity or presence of brain edema, these
findings cast doubt on the role of glutamine synthesis in the path-
ogenesis of these complications.
On the other hand, it has been known for almost half a decade
since the pioneering work of Warren and Schenker (1964) who
showed that GS inhibition by methionine sulfoximine (MSO) pro-
tected against the lethal effects of acute ammonia toxicity. Since
that time MSO has been shown to be partially protective against
the CNS effects of hyperammonemia in a variety of experimental
paradigms. For example, MSO treatment results in decreased
swelling of cortical astrocytes in ammonium acetate-treated rats
(Willard-Mack et al., 1996) and MSO treatment is protective in
ammonia-precipitated brain edema in portacaval-shunted rats
(Blei et al., 1994). It was concluded that increased glutamine pro-
duction in the astrocyte contributes to the pathogenesis of ammo-
nia-induced brain edema in the context of liver failure.
5. Inter-cellular trafficking of glutamine in brain in acute liver
failure
Unequivocal evidence of the cellular localization of synthetic
and degradative enzymes together with transporters for substrates
and release of products implicated in glutamine synthesis and
metabolism is now available. This evidence has given rise to the
concept of the ‘‘Glutamate–Glutamine Cycle’’ in brain (Fig. 6)
whereby glutamine formed in the astrocyte by the action of GS is
released from the cell via the glutamine transporter SNAT5 (Cube-
los et al., 2005) into the extracellular space where a portion of it
is available for uptake into the surrounding neurons via SNAT1
and SNAT2 (Melone et al., 2004, 2006). Once inside the neuron, glu-

1 tamine is transformed by the action of glutaminase into glutamate.

0.8 * sham-operated
In this way, glutamine serves as the immediate precursor of the
0.6 controls
releasable pool of the principal excitatory neurotransmitter (gluta-
* ALF
mate). The glutamate is released into the synaptic space where it
0.4
stimulates post-synaptic receptors. Surplus glutamate is cleared
0.2

0
brain muscle

GS activity
(µmole/mg protein/h)
Tissue Sham-operated ALF
controls

Cortex 5.86+/-0.25 4.76+/-0.35 *

Muscle 0.81+/-0.03 1.12+/-0.11 *

Fig. 4. GS expression and activities in cerebral cortex and skeletal muscle of rats
with ALF resulting from hepatic devascularization. (A) Expression levels of GS (as
determined by Western blot analysis) (n = 8) in ALF rats and sham-operated
controls; (B) GS activities in cerebral cortex and skeletal muscle of ALF rats and
sham-operated controls (⁄p < 0.05 vs. Sham-operated controls) (adapted from
Chatauret et al., 2006).
Fig. 5. Incorporation of 13C-label into glutamine (lmole/g tissue) as calculated from
13
C-NMR spectra of brain extracts from sham-operated controls (white bars), ALF
rats resulting from hepatic devascularization at precoma (grey bars) and coma
(black bars) stages of encephalopathy. Values represent means of n = 4 animals +/
SD (⁄p < 0.001 vs. sham-operated controls;  p < 0.01 vs. precoma) (adapted from
Zwingmann et al., 2003).
 P. Desjardins et al. / Neurochemistry International 60 (2012) 690–696
glutamatergic nerve terminal
GLN
GLNase
NH3 NH3
GLU
GLU
GLU
Fig. 6. Schematic representation of a glutamate–glutamine cycle. The cartoon illustrates position of the astrocytic glutamine synthetase (GS), neuronal glutaminase (GLNase),
the astrocytic glutamate transporter EAAT-2 and the astrocytic and neuronal glutamine transporters SNAT-5 and SNAT-1.
from the synapse by the action of a high affinity, high capacity glu-
tamate transporter EAAT-2. Once inside the astrocyte, glutamate
serves as the substrate for GS and the cycle goes on. A similar inter-
cellular cycle exists for GABA with distinct enzymes and transport-
ers; in the case of the GABA system, the astrocyte glutamine
transporter involved is primarily SNAT3 rather than SNAT5
(Boulland et al., 2002).
Studies using molecular techniques reveal a reduction in
expression of the astrocytic glutamate transporter EAAT-2 in brain
in experimental ALF due to hepatic devascularization (Knecht et al.,
1997). Preliminary evidence shows that HE and brain edema in this
same animal model of ALF results in a selective loss of expression
of the gene coding for SNAT5 in brain (Fig. 7) (Du et al., 2010).
Loss of SNAT5 expression could effectively trap glutamine inside
the astrocyte and this action (glutamine ‘‘trapping’’) rather than in-
creased glutamine synthesis could itself result in cell swelling and
brain edema in ALF. Moreover it is important to bear in mind that
glutamine functions not only as an osmoactive substance in brain
but is also intimately involved in the cell–cell interactions that
are key to neurotransmitter regulation. In this regard, it is well
established from classical biochemical techniques (Rothstein and
Tabakoff, 1985) as well as immunogold histochemistry (Laake
et al., 1995) that glutamine synthesized in the astrocyte is essential
for the maintenance of the releasable (neurotransmitter) pool of
glutamate in the adjoining glutamatergic neuron (Fig. 6). Conse-
quently, trapping of glutamine in the astrocyte resulting from
down-regulation of SNAT5 in ALF has the potential to result in
impairment of glutamatergic (excitatory) neurotransmission in
brain leading to the excessive neuroinhibition that is characteristic
of encephalopathy in ALF. Although attractive, this hypothesis has
not yet been totally validated and other astrocytic and/or neuronal
glutamine transporters may also be implicated. These mechanisms
may also apply in a brain region-dependent manner.
693
astrocyte
GLN
(to CSF)
SNAT-1 SNAT-5 GLN
Glutamine
synthetase
NH3
EAAT2 GLU
NH3
(Blood-borne)
It is true that total brain glutamate is decreased in a range of
hyperammonemias including those associated with liver failure
(Fitzpatrick et al., 1989; Lavoie et al., 1987; Mans et al., 1994; Swain
et al., 1992) and, although increased concentrations of glutamate
have been reported in brain extracellular fluid in experimental
ALF (Michalak et al., 1996) using the technique of in vivo microdi-
alysis, it is important to bear in mind that microdialysates may
not accurately reflect true presynaptic, Ca2+-dependent, evoked re-
lease of glutamate from nerve terminals. Further studies are re-
quired in order to address this important issue.
6. Glutamine metabolism in brain in acute liver failure
Phosphate-activated glutaminase (GLNase) is a mitochondrial
enzyme responsible for the deamination of glutamine to glutamate
with the production of ammonia according to the reaction:
Glutamine þ H2 O ! Glutamate þ NH3
Primarily neuronal in localization, GLNase is an integral compo-
nent of the Glutamate–Glutamine cycle (Fig. 6). Thus one turn of
the cycle removes one molecule of ammonia in the astrocyte due
to GS and produces one molecule of ammonia in the neuron due
to GLNase. It should be borne in mind, however that the steps de-
picted in the cycle are not meant to imply that the steps are con-
certed; in fact intermediates enter and leave the cycle at various
steps dependent upon the nature of the intermediate. It is also
important to recall that brain GLNase is inhibited by NH3 (product
inhibition) (Matheson and van den Berg, 1975) and, in view of the
increased brain ammonia concentrations in ALF, it has been sug-
gested that the increased brain glutamine results, at least in part,
from decreased activity of GLNase (Felipo and Butterworth, 2002).
A novel hypothesis was recently proposed whereby glutamine,
acting as a ‘‘Trojan horse’’ was transported into the mitochondrion
694 P. Desjardins et al. / Neurochemistry International 60 (2012) 690–696

(A) STD
SHAM ALF
(bp)

310

SNAT1
234 (249bp)

194

SNAT5
(149bp)
118

603

β-actin
310 (360bp)

1 2 3 4 5 6

SNAT5
(B) 150

SNAT1
mRNA expression
(% beta actin)

100
*

50

0
SHAM ALF SHAM ALF

Fig. 7. Expression of the neuronal amino acid transporter SLC38A1 (SNAT1) and the astrocytic transporters SLC38A5 (SNAT5) in the brain of rats with ALF resulting from
hepatic devascularization. (A) Total RNA was extracted from the frontal cortex of ALF rats at coma stage of encephalopathy (ALF) (lanes 5–6) or from sham-operated controls
(SHAM) (lanes 3–4). SNAT1 (279 bp) and SNAT5 (149 bp) mRNAs were reverse-transcribed and amplified by PCR. Expression were normalized to beta-actin housekeeping
gene. Reverse transcriptase was omitted from the reaction as a negative control (lane 2). (B) Graph bar representation of data shown in panel A. Data represent mean ± SEM of
four animals in each group (⁄p < 0.001 vs. SHAM by Student t test).

of the astrocyte and deaminated to ammonia which, in turn,
caused cellular dysfunction and cell swelling/edema (Albrecht
and Norenberg, 2006). Evidence for this was based primarily on
the results of experiments performed in cultured astrocytes which
are known to express much higher concentrations of GLNase than
those observed in astrocytes in situ and on these grounds together
with other fundamental differences in astrocyte phenotype be-
tween the in vitro and in vivo preparations, the theory has been se-
verely criticised (Brusilow et al., 2010). A major contentious issue
was that of the existence and function of GLNase in the astrocyte.
It is well established that GLNase is present in mouse brain astro-
cytes in culture (Kvamme et al., 1982), a finding that may relate to
culture conditions. However, it was recently shown, using immu-
nohistochemical techniques, that GLNase is also expressed by a
population of astrocytes in human brain (Olalla et al., 2008). A
novel form of GLNase has been reported in cultured astrocytes
(Kvamme et al., 2008). However, in contrast to neuronal GLNase,
the astrocytic enzyme is not subject to inhibition by the product
of the reaction, ammonia (Kvamme et al., 1982). Precisely how
the ‘‘Trojan horse’’ (glutamine) would result in the production of
sufficient ammonia to cause cell swelling and HE rather than
undergoing uncontrolled futile cycling to and from glutamate with
production and release of ammonia is difficult to reconcile with
these observations and with the report from Kanamori and Ross
(1995) that, under hyperammonemic conditions, brain GLNase
operates at less than 1% of maximal levels.
On a related topic, transamination of glutamine (with a suitable
alpha-ketoacid co-substrate) in brain results in the formation of an
interesting neuroactive/neurotoxic molecule known as alpha-keto-
glutaramate (KGM). CSF concentrations of KGM are increased over
10-fold in patients with liver failure (Vergara et al., 1974) and CSF
levels correlate better with severity of HE than did CSF glutamine.
It was proposed that this compound may represent a putative
endogenous neurotoxin or, alternatively, act as a biomarker of
CNS damage in liver failure. Urinary KGM is also markedly in-
creased in urea cycle disorders and other primary hyperammone-
mic diseases (Kuhara et al., 2011).
7. Therapeutic implications
Given the impressive correlation between arterial ammonia and
raised ICP in patients with ALF (Fig. 1) coupled to the findings of a
significant value of raised arterial ammonia for the prediction of
brain herniation in ALF patients (Clemmesen et al., 1999; Bernal
et al., 2007), it would be anticipated that ammonia-lowering strat-
egies using non-absorbable disaccharides, antibiotics, branched-
chain amino acids or L-ornithine L-aspartate (LOLA) would be effec-
tive in the prevention of brain edema and its complications in ALF.
 P. Desjardins et al. / Neurochemistry International 60 (2012) 690–696
Fig. 8. Effect of hypothermia on synthesis of brain glutamine in rats with ALF
resulting from hepatic devascularization. Percentage of 13C enrichments in
individual carbon position via the pyruvate carboxylase (PC) and pyruvate
dehydrogenase (PDH) pathways calculated from 13C-NMR spectra of extracts from
frontal cortex of normothermic (37 °C; black bars) and hypothermic (35 °C; hatched
bars) ALF rats compared to sham-operated controls (white bars). Values represent
means of n = 6 animals +/ SD (⁄p < 0.001 vs. sham-operated controls;  p < 0.001 vs
normothermic by ANOVA) (adapted from Chatauret et al., 2003).
However, to date, clinical trials using lactulose (Stravitz et al.,
2007) or LOLA (Acharya et al., 2009) have failed to show a robust
beneficial effect. Further clinical trials with other agents are clearly
warranted.
Reduction of brain glutamine concentrations would theoreti-
cally reduce astrocyte swelling and, in this way, prevent brain ede-
ma and its complications in ALF. With this principal in mind,
several studies have demonstrated the effective use of MSO in both
in vivo (Willard-Mack et al., 1996) and in vitro (Norenberg and
Bender, 1994) preparations in the reduction of brain glutamine
and the concomitant prevention of swelling caused by ammonia.
This approach was pioneered by Warren and Schenker (1964)
almost 50 years ago and continues to garner interest today (Brusilow
et al., 2010). MSO treatment attenuates brain edema in portacaval
rats with ammonia-induced brain edema (Blei et al., 1994). MSO
decreases the size of the astrocytic glutamine pool (Laake et al.,
1995) and also restores the loss of control of cerebral blood flow
(Master et al., 1999) and cerebral metabolic rate for glucose
(Hawkins and Jessy, 1991) caused by acute hyperammonemia. Fur-
thermore, studies in mice demonstrate no impairment of memory
function or other behavioural markers (Blin et al., 2002) which was
taken to indicate lack of neurotoxicity. However, the concept that
GS inhibition may afford a new therapeutic strategy in ALF has
been challenged (Zielińska et al., 2004) on the basis of its action
as a convulsant which may or may not relate to inhibition of GS,
to the fact that MSO is metabolized to products of unknown toxic-
ity (Rao and Meister, 1972; Cooper et al., 1976) and to the known
species-selectivity of its actions. In addition, given the evidence
that astrocyte-derived glutamine is required for the replenishment
of the releasable (neurotransmitter) pool of glutamate, restriction
on availability of this precursor has the potential to result in im-
paired excitatory transmission and worsening of HE.
Mild hypothermia delays the onset of HE and prevents brain ede-
ma and intracranial hypertension in both humans (Jalan et al., 1999)
and experimental animals (Traber et al., 1989; Rose et al., 2000;
Chatauret et al., 2003) with ALF. However, 13C-NMR studies in ALF
animals demonstrate that, although mild hypothermia prevents
brain edema, this protective effect is not accompanied by significant
decreases of glutamine synthesis via either of the two metabolic
pathways involving either pyruvate dehydrogenase or pyruvate car-
boxylase (Fig. 8) (Chatauret et al., 2003).
Consequently, mild hypothermia does not appear to prevent
brain edema and its complications in ALF by prevention of the rise
in astrocytic glutamine. Other mechanisms have been identified
695
including effects on brain energy metabolism, brain lactate accu-
mulation (Chatauret et al., 2003), and anti-inflammatory proper-
ties (Jiang et al., 2009).
In summary, ALF, in common with other acute hyperammone-
mic disorders, is characterized by increased CSF and brain gluta-
mine. The osmotic consequences of increased glutamine synthesis
in astrocytes have traditionally been considered to be causally re-
lated to the cytotoxic brain edema and its complications (intracra-
nial hypertension, brain herniation) in ALF. However studies using
multinuclear 1H- and 13C-NMR spectroscopy demonstrate that nei-
ther brain glutamine concentrations nor brain glutamine synthesis
rates correlate with HE grade or the presence of brain edema in ALF.
An alternative mechanism is now proposed whereby glutamine is
trapped within the astrocyte as a consequence of down-regulation
of the glutamine transporter SNAT5 in ALF. Restricted transfer out
of the cell rather than increased synthesis within the cell could ex-
plain the cell swelling/brain edema and restricted transfer of gluta-
mine from the astrocyte to the adjacent glutamatergic nerve
terminal (where glutamine serves as immediate precursor for the
releasable/transmitter pool of glutamate) could result in decreased
excitatory transmission and excessive neuroinhibition that is
characteristic of HE in ALF.
In spite of renewed interest in arterial ammonia as a predictor
of raised intracranial pressure and brain herniation in ALF, ammo-
nia-lowering agents aimed at reduction of ammonia production in
the gut are of limited value in the prevention of these cerebral con-
sequences. Moreover, mild hypothermia, shown to prevent brain
edema and intracranial hypertension in both experimental and hu-
man ALF, does so independent of effects on brain glutamine syn-
thesis. While inhibitors of brain glutamine synthesis have been
proposed for the prevention of brain edema in ALF, potential ad-
verse effects have so far limited their applicability.
Acknowledgment
Funded by the Canadian Institutes of Health Research
(China–Canada Joint Health Research Initiative).
References
Acharya, S.K., Bhatia, V., Sreenivas, V., Khanal, S., Panda, S.K., 2009. Efficacy of L-
ornithine L-aspartate in acute liver failure: a double-blind, randomized,
placebo-controlled study. Gastroenterology 136, 2159–2168.
Albrecht, J., Norenberg, M.D., 2006. Glutamine: a Trojan horse in ammonia
neurotoxicity. Hepatology 44, 788–794.
Bernal, W., Hall, C., Karvellas, C.J., Auzinger, G., Sizer, E., Wendon, J., 2007. Arterial
ammonia and clinical risk factors for encephalopathy and intracranial
hypertension in acute liver failure. Hepatology 46, 1844–1852.
Blei, A.T., Olafsson, S., Therrien, G., Butterworth, R.F., 1994. Ammonia-induced brain
edema and intracranial hypertension in rats after portacaval anastomosis.
Hepatology 19, 1437–1444.
Blin, M., Crusio, W.E., Hévor, T., Cloix, J.F., 2002. Chronic inhibition of glutamine
synthetase is not associated with impairment of learning and memory in mice.
Brain Res. Bull. 57, 11–15.
Bosman, D.K., Deutz, N.E., De Graaf, A.A., vd Hulst, R.W., Van Eijk, H.M., Bovée, W.M.,
Maas, M.A., Jörning, G.G., Chamuleau, R.A., 1990. Changes in brain metabolism
during hyperammonemia and acute liver failure: results of a comparative 1H-
NMR spectroscopy and biochemical investigation. Hepatology (12) 281–290.
Boulland, J.L., Osen, K.K., Levy, L.M., Danbolt, N.C., Edwards, R.H., Storm-Mathisen, J.,
Chaudhry, F.A., 2002. Cell-specific expression of the glutamine transporter SN1
suggests differences in dependence on the glutamine cycle. Eur. J. Neurosci. 15,
1615–1631.
Brusilow, S.W., Koehler, R.C., Traystman, R.J., Cooper, A.J., 2010. Astrocyte glutamine
synthetase: importance in hyperammonemic syndromes and potential target
for therapy. Neurotherapeutics 7, 452–470.
Chatauret, N., Desjardins, P., Zwingmann, C., Rose, C., Rao, K.V., Butterworth, R.F.,
2006. Direct molecular and spectroscopic evidence for increased ammonia
removal capacity of skeletal muscle in acute liver failure. J. Hepatol. 44, 1083–
1088.
Chatauret, N., Zwingmann, C., Rose, C., Leibfritz, D., Butterworth, R.F., 2003. Effects
of hypothermia on brain glucose metabolism in acute liver failure: a H/C-
nuclear magnetic resonance study. Gastroenterology 125, 815–824.
696 P. Desjardins et al. / Neurochemistry International 60 (2012) 690–696
Clemmesen, J.O., Larsen, F.S., Kondrup, J., Hansen, B.A., Ott, P., 1999. Cerebral
herniation in patients with acute liver failure is correlated with arterial
ammonia concentration. Hepatology 29, 648–653.
Cooper, A.J., Mora, S.N., Cruz, N.F., Gelbard, A.S., 1985. Cerebral ammonia
metabolism in hyperammonemic rats. J. Neurochem. 44, 1716–1723.
Cooper, A.J., Stephani, R.A., Meister, A., 1976. Enzymatic reactions of methionine
sulfoximine. Conversion to the corresponding alpha-imino and alpha-keto acids
and to alpha-ketobutyrate and methane sulfinimide. J. Biol. Chem. 251, 6674–
6682.
Cubelos, B., González-González, I.M., Giménez, C., Zafra, F., 2005. Amino acid
transporter SNAT5 localizes to glial cells in the rat brain. Glia 49, 230–244.
Deutz, N.E., Chamuleau, R.A., de Graaf, A.A., Bovée, W.M., de Beer, R., 1988. In vivo
31
P NMR spectroscopy of the rat cerebral cortex during acute hepatic
encephalopathy. NMR Biomed. 1, 101–106.
Du, T., Grimm, A., Jiang, W., Desjardins, P., Butterworth, R.F., 2010. Pathogenesis of
Brain Edema in Acute Liver Failure: Key Role of the Astrocytic Glutamine
Transporter SNAT-3. Hepatology (52) 271A (abstract).
Felipo, V., Butterworth, R.F., 2002. Neurobiology of ammonia. Prog. Neurobiol. 67,
259–279.
Ferenci, P., Herneth, A., Steindl, P., 1996. Newer approaches to therapy of hepatic
encephalopathy. Semin. Liver Dis. 16, 329–338.
Fitzpatrick, S.M., Hetherington, H.P., Behar, K.L., Shulman, R.G., 1989. Effects of acute
hyperammonemia on cerebral amino acid metabolism and pHi in vivo,
measured by 1H and 31P nuclear magnetic resonance. J. Neurochem. 52, 741–
749.
Jalan, R., O Damink, S.W., Deutz, N.E., Lee, A., Hayes, P.C., 1999. Moderate
hypothermia for uncontrolled intracranial hypertension in acute liver failure.
Lancet (354) 1164–1168.
Jiang, W., Desjardins, P., Butterworth, R.F., 2009. Direct evidence for central
proinflammatory mechanisms in rats with experimental acute liver failure:
protective effect of hypothermia. J. Cerebl. Blood Flow Metab. 29, 944–952.
Hawkins, R.A., Jessy, J., 1991. Hyperammonaemia does not impair brain function in
the absence of net glutamine synthesis. Biochem. J. 277, 697–703.
Kanamori, K., Ross, B.D., Chung, J.C., Kuo, E.L., 1996. Severity of hyperammonemic
encephalopathy correlates with brain ammonia level and saturation of
glutamine synthetase in vivo. J. Neurochem. 67, 1584–1594.
Kanamori, K., Ross, B.D., 1995. In vivo activity of glutaminase in the brain of
hyperammonaemic rats measured by 15N nuclear magnetic resonance.
Biochem. J. 305, 329–336.
Knecht, K., Michalak, A., Rose, C., Rothstein, J.D., Butterworth, R.F., 1997. Decreased
glutamate transporter (GLT-1) expression in frontal cortex of rats with acute
liver failure. Neurosci. Lett. 229, 201–203.
Kuhara, T., Inoue, Y., Ohse, M., Krasnikov, B.F., Cooper, A.J., 2011. Urinary 2-hydroxy-
5-oxoproline, the lactam form of a-ketoglutaramate, is markedly increased in
urea cycle disorders. Anal. Bioanal. Chem. 400, 1843–1851.
Kvamme, E., Nissen-Meyer, L.S., Roberg, B.A., Torgner, I.A., 2008. Novel form of
phosphate activated glutaminase in cultured astrocytes and human
neuroblastoma cells, PAG in brain pathology and localization in the
mitochondria. Neurochem Res 33, 1341–1345.
Kvamme, E., Svenneby, G., Hertz, L., Schousboe, A., 1982. Properties of phosphate
activated glutaminase in astrocytes cultured from mouse brain. Neurochem.
Res. 7, 670–761.
Laake, J.H., Slyngstad, T.A., Haug, F.M., Ottersen, O.P., 1995. Glutamine from glial
cells is essential for the maintenance of the nerve terminal pool of glutamate:
immunogold evidence from hippocampal slice cultures. J. Neurochem. 65, 871–
881.
Lavoie, J., Giguère, J.F., Layrargues, G.P., Butterworth, R.F., 1987. Amino acid changes
in autopsied brain tissue from cirrhotic patients with hepatic encephalopathy. J.
Neurochem. 49, 692–697.
Lockwood, A.H., McDonald, J.M., Reiman, R.E., Gelbard, A.S., Laughlin, J.S., Duffy, T.E.,
Plum, F., 1979. The dynamics of ammonia metabolism in man. Effects of liver
disease and hyperammonemia. J. Clin. Invest. 63, 449–460.
Mans, A.M., DeJoseph, M.R., Hawkins, R.A., 1994. Metabolic abnormalities and grade
of encephalopathy in acute hepatic failure. J. Neurochem. 63, 1829–1838.
Mans, A.M., Saunders, S.J., Kirsch, R.E., Biebuyck, J.F., 1979. Correlation of plasma
and brain amino acid and putative neurotransmitter alterations during acute
hepatic coma in the rat. J. Neurochem. 32, 285–292.
Master, S., Gottstein, J., Blei, A.T., 1999. Cerebral blood flow and the development of
ammonia-induced brain edema in rats after portacaval anastomosis.
Hepatology 30, 876–880.
Matheson, D.F., van den Berg, C.J., 1975. Ammonia and brain glutamine: inhibition
of glutamine degradation by ammonia. Biochem. Soc. Trans. 3, 525–528.
Melone, M., Varoqui, H., Erickson, J.D., Conti, F., 2006. Localization of the Na(+)-
coupled neutral amino acid transporter 2 in the cerebral cortex. Neuroscience
140, 281–292.
Melone, M., Quagliano, F., Barbaresi, P., Varoqui, H., Erickson, J.D., Conti, F., 2004.
Localization of the glutamine transporter SNAT1 in rat cerebral cortex and
neighboring structures, with a note on its localization in human cortex. Cereb.
Cortex 14, 562–574.
Michalak, A., Rose, C., Butterworth, J., Butterworth, R.F., 1996. Neuroactive amino
acids and glutamate (NMDA) receptors in frontal cortex of rats with
experimental acute liver failure. Hepatology 24, 908–913.
Norenberg, M.D., Bender, A.S., 1994. Astrocyte swelling in liver failure: role of
glutamine and benzodiazepines. Acta Neurochir. Suppl. (Wien) 60, 24–27.
Norenberg, M.D., 1979. Distribution of glutamine synthetase in the rat central
nervous system. J. Histochem. Cytochem. 27, 756–762.
Olalla, L., Gutiérrez, A., Jiménez, A.J., López-Téllez, J.F., Khan, Z.U., Pérez, J., Alonso,
F.J., de la Rosa, V., Campos-Sandoval, J.A., Segura, J.A., Aledo, J.C., Márquez, J.,
2008. Expression of the scaffolding PDZ protein glutaminase-interacting protein
in mammalian brain. J. Neurosci. Res. 86, 281–292.
Rao, S.L., Meister, A., 1972. In vivo formation of methionine sulfoximine phosphate,
a protein-bound metabolite of methionine sulfoximine. Biochemistry 11, 1123–
1127.
Record, C.O., Buxton, B., Chase, R.A., Curzon, G., Murray-Lyon, I.M., Williams, R.,
1976. Plasma and brain amino acids in fulminant hepatic failure and their
relationship to hepatic encephalopathy. Eur. J. Clin. Invest. 6, 387–394.
Rose, C., Michalak, A., Pannunzio, M., Chatauret, N., Rambaldi, A., Butterworth, R.F.,
2000. Mild hypothermia delays the onset of coma and prevents brain edema
and extracellular brain glutamate accumulation in rats with acute liver failure.
Hepatology 31, 872–877.
Rothstein, J.D., Tabakoff, B., 1985. Glial and neuronal glutamate transport following
glutamine synthetase inhibition. Biochem. Pharmacol. 34, 73–79.
Schliess, F., Görg, B., Fischer, R., Desjardins, P., Bidmon, H.J., Herrmann, A.,
Butterworth, R.F., Zilles, K., Häussinger, D., 2002. Ammonia induces MK-801-
sensitive nitration and phosphorylation of protein tyrosine residues in rat
astrocytes. FASEB J. 16, 739–741.
Stravitz, R.T., Kramer, A.H., Davern, T., Shaikh, A.O., Caldwell, S.H., Mehta, R.L., Blei,
A.T., Fontana, R.J., McGuire, B.M., Rossaro, L., Smith, A.D., Lee, W.M; Acute Liver
Failure Study Group, 2007. Intensive care of patients with acute liver failure:
recommendations of the U.S. Acute Liver Failure Study Group. Crit Care Med
(35) 2498–2508.
Swain, M., Butterworth, R.F., Blei, A.T., 1992. Ammonia and related amino acids in
the pathogenesis of brain edema in acute ischemic liver failure in rats.
Hepatology 15, 449–453.
Tofteng, F., Hauerberg, J., Hansen, B.A., Pedersen, C.B., Jørgensen, L., Larsen, F.S.,
2006. Persistent arterial hyperammonemia increases the concentration of
glutamine and alanine in the brain and correlates with intracranial pressure in
patients with fulminant hepatic failure. J. Cerebl. Blood Flow Metab. 26, 21–27.
Traber, P., DalCanto, M., Ganger, D., Blei, A.T., 1989. Effect of body temperature on
brain edema and encephalopathy in the rat after hepatic devascularization.
Gastroenterology 96, 885–891.
Vergara, F., Plum, F., Duffy, T.E., 1974. Alpha-ketoglutaramate: increased
concentrations in the cerebrospinal fluid of patients in hepatic coma. Science
183, 81–83.
Warren, K.S., Schenker, S., 1964. Effect of an inhibitor of glutamine synthesis
(Methionine sulfoximine) on ammonia toxicity and metabolism. J. Lab. Clin.
Med. 64, 442–449.
Willard-Mack, C.L., Koehler, R.C., Hirata, T., Cork, L.C., Takahashi, H., Traystman, R.J.,
Brusilow, S.W., 1996. Inhibition of glutamine synthetase reduces ammonia-
induced astrocyte swelling in rat. Neuroscience 71, 589–599.
Zielińska, M., Stafiej, A., Law, R.O., Albrecht, J., 2004. Effects of methionine
sulfoximine on the glutamine and glutamate content and cell volume in rat
cerebral cortical slices: involvement of mechanisms not related to inhibition of
glutamine synthesis. Neurotoxicology 25, 443–449.
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.