ORIGINAL PAPER

Acute and Chronic Hyperammonemia Modulate AntioxidantEnzymes Differently in Cerebral Cortex and Cerebellum

Santosh Singh Æ Raj K. Koiri Æ Surendra Kumar Trigun

Accepted: 18 June 2007 / Published online: 4 August 2007

� Springer Science+Business Media, LLC 2007

Abstract Studies on acute hyperammonemic models

suggest a role of oxidative stress in neuropathology of

ammonia toxicity. Mostly, a low grade chronic type

hyperammonemia (HA) prevails in patients with liver

diseases and causes derangements mainly in cerebellum

associated functions. To understand whether cerebellum

responds differently than other brain regions to chronic type

HA with respect to oxidative stress, this article compares

active levels of all the antioxidant enzymes vis a vis extent

of oxidative damage in cerebral cortex and cerebellum of

rats with acute and chronic HA induced by intra-peritoneal

injection of ammonium acetate (successive doses of

10 · 103 & 8 · 103 lmol/kg b.w. at 30 min interval for

acute and 8 · 103 lmol/kg b.w. daily up to 3 days for

chronic HA). As compared to the respective control sets,

cerebral cortex of acute HA rats showed significant decline

(P < 0.01–0.001) in the levels of superoxide dismutase

(SOD), catalase and glutathione peroxidase (GPx) but with

no change in glutathione reductase (GR). In cerebellum of

acute HA rats, SOD, catalase and GR though declined

significantly, GPx level was found to be stable. Contrary to

this, during chronic HA, levels of SOD, catalase and GPx

increased significantly in cerebral cortex, however, with a

significant decline in the levels of SOD and GPx in cere-

bellum. The results suggest that most of the antioxidant

enzymes decline during acute HA in both the brain regions.

However, chronic HA induces adaptive changes, with

respect to the critical antioxidant enzymes, in cerebral cor-

tex and renders cerebellum susceptible to the oxidative

stress. This is supported by ~ 2- and 3-times increases in the

level of lipid peroxidation in cerebellum during chronic and

acute HA respectively, however, with no change in the

cortex due to chronic HA.

Keywords Hyperammonemia � Ammonia neurotoxicity �Antioxidant enzymes � Oxidative stress � Cerebral cortex �Cerebellum

Introduction

Hepatic encephalopathy (HE) is a serious nervous system

disorder developed due to increased ammonia level in brain

resulting from liver dysfunction. This is of great concern

because a number of liver disorders like viral hepatitis,

liver intoxication, alcoholism and inborn errors of urea

cycle are associated with different grades of hyperammo-

nemic conditions in the patients [1]. It has been reported

that acute ammonia exposure of brain cells causes

dysfunction of multiple neurotransmitter system [1, 2] and

glutamate & ammonia mediated excitotoxicity of neurons

[3]. At down stream level, defects in brain bioenergetics [4]

and mitochondrial dysfunction mediated oxidative stress

[5, 6] are considered to play important roles in patho-

physiology of HE. Moreover, most of the evidences for a

role of oxidative stress in ammonia neurotoxicity have

been derived either from cell culture studies [6, 7] and/or

from acute hyperammonemic animal models [8–10].

Nonetheless, low grade chronic hyperammonemic condi-

tion is more prevalent in the patients suffering from viral

hepatitis and liver dysfunction due to alcoholism and long

term drug abuse. Therefore, it is important to understand

how chronic HA affects cellular antioxidant defense

mechanisms in susceptible brain regions.

S. Singh � R. K. Koiri � S. K. Trigun (&)

Biochemistry & Molecular Biology Laboratory, Department of

Zoology, Banaras Hindu University, Varanasi 221005, India

e-mail: sktrigun@sify.com; sktrigun@bhu.ac.in

123

Neurochem Res (2008) 33:103–113

DOI 10.1007/s11064-007-9422-x

There are some reports on the role of oxidative stress in

chronic hyperammonemic models also; however, most of

them are focused to the hyperammonemia (HA) dependent

impairment of NMDA receptor activity [1] via alterations

in glutamate-NO-cGMP pathway [11, 12]. In addition,

chronic HA has been found to induce adaptive changes

in brain energy and ammonia metabolites, which are

altered otherwise during acute ammonia intoxication [13].

Increases in the levels of ammonia, glutamate and mito-

chondrial NAD/NADH ratio in chronic HA models [13]

hint for a mitochiondrial dysfunction and implication of

oxygen free radicals in the pathophysiology of chronic type

HE also. Nonetheless, information is scarce on implication

of antioxidant enzyme system during chronic HA in animal

models.

Primary level neuropathology of HE, like motor

disturbances, expressionless face, rigid muscle tone, tremor

etc, is common with the low-grade chronic type HE

patients [14, 15] and these functions are mainly associated

with the derangements in motor activities of cerebellum.

Thus, it is likely that cerebellum responds differently to

HA than the other brain regions. Differential susceptibility

of cerebellum and cerebral cortex with respect to the

activation of guanylate cyclase by NO in mild HA animal

models has been reported [12] and importantly, similar

changes were also observed in these brain regions of

chronic type HE patients [16]. Therefore, it is important to

ascertain whether and how different brain regions respond

to chronic HA with respect to O2– based oxidative stress.

Brain consumes more O2 than any other tissues and thus,

produces high level of reactive oxygen species (ROS) and

operates efficient antioxidant enzyme systems to counteract

the deleterious effects of oxidative stress [17, 18]. Super-

oxide dismutase (SOD) and catalase scavenge O2– to pro-

duce water & O2, whereas, interplay of SOD, glutathione

peroxidase (GPx) and glutathione reductase (GR) channels

O2– in a NADPH dependent pathway to maintain the ratio of

GSH/GSSG and to prevent lipid peroxidation during

oxidative stress. It has been reported that though catalase is

also found in brain cells, it is SOD-GPx-GR pathway that is

more important for antioxidant activities in brain [19, 20].

In view of a high degree of metabolic plasticity in brain

cells in general [18] and with respect to antioxidant

enzymes in particular [21–23], it may be speculated that as

compared to the acute conditions, chronic HA may produce

differential changes in the antioxidant enzyme system in

different brain regions. In the present report, we have

compared, in a concerted manner, the extent of oxidative

damage and levels of all the key antioxidant enzymes in rat

brain cortex (less affected due to mild HA) and cerebellum

(whose functions are affected the most in chronic HE

patients) in acute and chronic HA rat models.

Experimental procedure

Animal and chemicals

Male adult albino rats weighing 100–120 g were main-

tained in an animal house as per the recommendations from

institutional ethical committee for the care and use of

laboratory animals.

All chemicals used were of analytical grade or of the best

quality supplied by E-Merk, Glaxo and SRL (INDIA).

Acrylamide, N N-methylene bis acrylamide, Coomassie

Brilliant Blue R-250 (CBB), TEMED (N N N N-tetrameth-

ylethylene diamine) and Phenyl methyl sulphonyl fluoride

(PMSF) were purchased from Sigma Chemical Co., USA.

Experimental design

Acute and chronic HA in rats were induced by intraperi-

toneal injection of ammonium acetate prepared in physio-

logical saline (0.9% NaCl). As described earlier [24], for

acute HA, first 10 · 103 lmol/kg b.w. of ammonium

acetate was administered to the rats followed by a second

injection of 8 · 103 lmol/kg b.w. after 30 min interval.

Chronic HA group rats were injected daily up to 3 days

with 8 · 103 lmol/kg b.w. ammonium acetate. Control

group rats for each experimental set were simultaneously

given with equivalent volume of physiological saline.

About 80% of the rats with acute/ episodic treatment

survived up to 30 min after the last injection. In case of

chronically treated rats, 90% of them could survive after

the last injection. All animals were sacrificed by decapi-

tation after 30 min of the final injection and cerebral cortex

& cerebellum were dissected out, washed in ice cold saline

(0.9% NaCl) and stored frozen at –70�C for further studies.

Level of HA was ascertained by measuring ammonia

concentration in whole brain taking fresh tissues from 3

rats from each control as well as experimental groups.

Preparation of tissue extracts

Whole brain, cerebral cortex and cerebellum extracts were

prepared in 0.02 M Tris–Cl (pH 7.4) containing protease

inhibitors as described from our lab [25]. Extracts were

centrifuged at 35,000 g for 45 min at 4�C. The superna-

tants collected were used for the studies on antioxidant

enzymes and other biochemical assays. Protein content was

determined by the method of Lowery et al. [26].

Biochemical estimations

Ammonia concentration was measured using a kit sup-

plied by Sigma–Aldrich, USA. The brain extracts were

104 Neurochem Res (2008) 33:103–113

123

deproteinized in 1/5 volumes of ice-cold 100 g/l trichlo-

roacetic acid, and kept on ice for 15 min. After centrifu-

gation at 15,000 g for 15 min at 4�C, the supernatants were

neutralized with 2.0 M KHCO3, centrifuged again and used

for estimating ammonia. The method employed measuring

the rate of conversion of a-ketoglutarate to glutamate cat-

alyzed by glutamate dehydrogenase in the presence of

ammonia. The reaction mixture (1 ml) contained 50 ll of

sample, 3.4 mM a-ketoglutarate and 0.23 mM reduced

NADPH in 50 mM phosphate buffer (pH 7.4). The reaction

was started by the addition of suitably diluted glutamate

dehydrogenase. Initial and final (after 5 min) absorbance at

340 nm was used to calculate the concentration of

ammonia in terms of lmol/g wet wt of tissue.

Malondialdehyde (MDA), the product of lipid peroxi-

dation, was measured by the method reported earlier [27].

Briefly, 1 ml of Tris–Maleate buffer (0.2 M, pH 5.9) and

0.5 ml of the extract was incubated at 37�C for 30 min.

Thereafter, 1.5 ml of thiobarbituric acid (TBA) was added

and the mixture was incubated in boiling water bath for

10 min using tight condensers. After cooling, 3 ml of

pyridine: n-butanol mixture (3:1 v/v) and 1 ml of 1.0 N (w/

v) NaOH were added. The contents were thoroughly mixed

and allowed to stand for 10 min. The absorbance was read

at 548 nm and the levels of lipid peroxidation were ex-

pressed as nmole MDA/g wet wt.

Total thiol was estimated as described earlier [28].

Aliquots of 0.1 ml tissue extracts were mixed with 1.5 ml

of 0.2 M Tris buffer, pH 8.2 and 0.1 ml 0.01 M 5,5’-Di-

thio-bis (2-nitrobenzoic acid) (DTNB) . The mixture was

made up to 10 ml with methanol and was incubated for

30 min. The mixture was then centrifuged at 3,000 rpm for

15 min. and absorbance of the supernatant was read at

412 nm. The molar extinction coefficient of 13,100 was

used to calculate GSH (reduced glutathione) and values

were presented as nmol/mg protein.

Studies on antioxidant enzymes

Assay of SOD and catalase

The activity of superoxide dismutase (SOD; EC: 1.15.1.1)

was measured following an earlier described method [29].

The reaction mixture consisted of 0.02 M sodium

pyrophosphate buffer (pH 8.3), 6.2 lM phenazine metho-

sulphate (PMS), 30 lM nitroblue tetrazolium (NBT), and

0.1 ml suitably diluted tissue extracts. The reaction was

started by the addition of 50 lM NADH at 30�C and

stopped after 90 s by the addition of 2.0 ml glacial acetic

acid. A control set without tissue extract was run simul-

taneously. The reaction mixture was stirred, shaken with

4 ml of n-butanol, allowed to stand for 10 min and

centrifuged to separate butanol layer containing the

chromogen. Absorbance was measured at 560 nm using

butanol as blank. Unit of the enzyme was defined as the

amount of enzyme that produced 50% inhibition of NBT

reduction per min. and the activity was expressed as units/

mg protein.

Catalase (EC: 1.11.1.6) was assayed following an earlier

reported procedure [30] with some modifications. Briefly,

in a reaction mixture containing 0.01 M Potassium phos-

phate buffer (pH 7.0) and 0.1 ml of tissue extract, reaction

was started by the addition of 0.8 M hydrogen peroxide

(H2O2) and stopped after 60 s by 2.0 ml dichromate acetic

acid reagent. All the tubes were heated in a boiling water

bath for 10 min., cooled and absorbance was read at

570 nm. After comparing with a standard plot constructed

using a range of 10–160 lmoles of H2O2, the activity of

catalase was expressed as lmoles of H2O2 consumed/min/

mg protein.

Analysis of SOD and catalase by non-denaturing PAGE

Non-denaturing PAGE of the tissue extracts were per-

formed as reported from this laboratory [31]. For SOD, the

extract containing 60 lg protein was loaded in each lane of

12% non-denaturing PAGE. After electrophoresis, the gels

were subjected to substrate specific staining of SOD bands

as described earlier [32]. The staining mixture consisted of

2.5 mM NBT, 28 lM riboflavin, and 28 mM TEMED.

After 20 min incubation in the dark, gels were exposed to

a fluorescent light to develop achromatic bands against

dark blue background corresponding to SOD protein in

the gel.

For catalase, tissue extracts containing 60 lg proteins

were electrophoressed on 8% non-denaturing PAGE.

Catalase specific bands were developed according to Sun

et al. [33]. Briefly, gels were soaked for 10 min in 0.003%

H2O2 and then incubated in a staining mixture consisted of 2%

potassium ferricyanide and 2% ferric chloride. Achromatic

catalase bands appeared against a blue–green background.

The intensity of bands was quantitated by gel densitometry

using alphaimager 2200 gel documentation software.

Active level of glutathione peroxidase

Glutathione peroxidase (GPx; EC:1.11.1.9) level was

determined by in gel detection method as described earlier

[34]. After 10% non-denaturing PAGE of the extracts

containing 30 lg protein in each lane, the gels were

incubated in a GPx specific staining mixture composed of

50 mM Tris–Cl buffer (pH 7.9), 3 mM GSH, 0.004%

H2O2, 1.2 mM NBT and 1.6 mM PMS. Achromatic bands

corresponding to GPx activity appeared against a violet–

blue background. The level of GPx was quantified by gel

densitometry as described earlier.

Neurochem Res (2008) 33:103–113 105

123

During PAGE based detection of all the three antioxi-

dant enzymes, SOD, catalase and GPx, development of

enzyme specific bands were confirmed by comparing the

results of similarly run gels stained in the presence and

absence of the enzyme specific substrates. In each case,

PAGE was performed 3–4 times and mean ± SD of

densitometric values of the bands as % of control lane from

all the gels run were presented with a representative gel

photograph.

Glutathione reductase assay

Activity of glutathione reductase (GR; EC: 1.6.4.2) was

determined following the method of Carlberg and

Mannervik [35]. In brief, the reaction mixture (1 ml)

consisted of 0.2 M sodium phosphate buffer (pH 7.0),

0.2 mM EDTA, 1 mM oxidized glutathione (GSSG) and

0.2 mM NADPH. The reaction was initiated by the

addition of the tissue extract and oxidation of NADPH was

recorded as decrease in absorbance at 340 nm for 5 min.

Nonspecific oxidation of NADPH was corrected by the

absorbance obtained in the absence of GSSG. Unit of the

enzyme was defined as lmole NADP/min/ at 30�C and

the enzyme activity was expressed as units/mg protein.

Statistical analysis of the data was done as reported

earlier [25] and the student ‘t’ test was performed to find

the level of significance between control and experimental

groups.

Results

As compared to the respective control groups, ~ 5–7 fold

increases in brain ammonia level was observed in rats with

episodic treatment of ammonium acetate and ~1.5–1.8 fold

increase with those treated once daily up to 3 days. As

reported earlier [24], these groups were referred to as acute

and chronic HA groups respectively.

Comparison of oxidative damage due to acute and

chronic HA

Measuring MDA level, as a stable product of lipid peroxi-

dation, is a reliable tool to assess the extent of oxidative

damage at cellular level. According to Table 1, as compared

to the control rats, there was a significant increase (1.3 fold)

in MDA level in cerebral cortex of the rats with acute HA,

but with no change during chronic HA. In cerebellum,

however, MDA level was 3- and 2-fold higher in acute and

chronic HA rats respectively than the corresponding control

groups. When compared between cortex and cerebellum,

there was ~ 2 times higher MDA level in cerebellum than

the cortex in both the HA group rats. The level of total GSH,

as a measure of reducing equivalents in the brain cells,

was observed to be unaltered in both cerebral cortex and

cerebellum under acute and chronic HA.

Degree of HA & the level of antioxidant enzymes

In general, activity of the enzymes measured in cell free

extracts is correlated with the metabolic efficiency of the

cells under a variety of pathophysiological conditions.

However, measuring enzyme activity in cell free extracts

may not reflect actual levels of the enzymatic proteins in

the cells. Therefore, to monitor active levels of the anti-

oxidant enzymes, in the present study, cell extracts were

subjected to non-denaturing PAGE followed by activity

staining based detection of enzymatic proteins in the gel.

This method is relatively less sensitive than to detecting

proteins by Western blotting. However, it is more relevant

for physiological interpretations, as in this method detec-

tion is based on specificity of the enzyme for its substrate

and thus, activity based intensity of bands in gel reflects

only active level of the enzyme (native protein). In com-

parison, antibody based detection can not differentiate

between the active and inactive structures of the proteins.

A difference between Western blot detected enzymatic

protein level and that with the intensity of activity bands in

gel has been reported in case of most of the antioxidant

enzymes in a tumor cell line [36]. Thus, in the present

article, results from spectrophotometric measurements

have been interpreted as activity level of the enzyme and

PAGE bands as the level of active fraction of the enzymatic

protein in brain tissues.

Effect of acute HA on antioxidant enzymes

The first step of neutralization of O2– is completed by

synchronized activities of SOD & catalse and/or by SOD &

GPx in mammalian cells. As compared to the control group

rats, activities of SOD and catalase were observed to be

declined significantly (P < 0.01–0.001) in both, cerebral

cortex and cerebellum of rats with acute HA (Figs. 1A,

2A). The intensity of SOD band in gel also followed the

declining pattern in the cortex, however, with a significant

(P < 0.05) increase in cerebellum of acute HA rats

(Fig. 1B, C). And in case of catalase, intensity of PAGE

bands were found unchanged in both the brain regions

during acute HA (Fig. 2B, C). Such a non-correlative

pattern between the activity data and PAGE results of SOD

and catalase could be attributed to some inhibitory mech-

anisms for these enzymes in brain during acute HA.

Four isoforms of GPx have been reported in mammalian

tissues [20]. Though, brain contains pre-dominantly phos-

pholipids hydrogen peroxide GPx (pHGPx), the other three

isoforms have also been reported in brain but in less

106 Neurochem Res (2008) 33:103–113

123

amount [20]. In the absence of a literature on classification

of GPx isoforms based on their migration in non-denatur-

ing PAGE, in this article, GPx bands have been referred to

as GPx1–GPx 4 based on their relative migration in non-

denaturing PAGE starting from top to bottom (Figs. 3, 7).

According to Fig. 3A and B, as compared to the control

lanes, all the four GPx isoforms though declined slightly

(P < 0.05) in the cerebral cortex, but with an insignificant

change in cerebellum of rats with acute HA. A similar

pattern was also observed when GPx activity was measured

in vitro in the cell extracts from the respective brain re-

gions (unpublished results). Contrary to this, in comparison

to the samples from control rats, though there was a small

decline (P < 0.05) in the activity of GR in the cerebellum,

GR activity in the cerebral cortex remained unchanged

during acute HA.

Effect of chronic HA on antioxidant enzymes

Figures 5–7 illustrate that in cerebral cortex of chronic HA

rats, as compared to the control group, activities as well as

levels of active fractions of SOD, catalase and all isoforms

of GPx increased significantly (P < 0.05–0.001). However,

in cerebellum, though the activity and active levels of SOD

(P < 0.001) & all the GPx bands including GPx 2 (pHGPx)

declined significantly (P < 0.05), there were no significant

change observed in the activity and the level of catalase

during chronic HA. Moreover, as compared to the control

group rats, rats with chronic HA showed significant decline

(P < 0.01) in the activity of GR in cerebral cortex but with

no change in cerebellum (Fig. 8).

Discussion

In the present article, we intended to address two aspects of

ammonia neurotoxicity, one the relationship between the

degree of HA and oxidative stress in brain & secondly,

since cerebellum associated functions are affected the most

during chronic HA, is it that cerebellum is more susceptible

to ammonia toxicity than other brain regions with respect

to oxidative stress. For such comparative studies on

ammonia toxicity in brain, pure hyperammonemic animal

models, induced by administration of ammonium salt, with

normal liver function is recommended over other HA

models with acute liver failure [37]. This is because the

findings from pure HA models are assumed to be devoid of

the interferences from other pathological factors associated

to liver dysfunction. Additionally, ammonia diffuses in

brain with a faster rate during HE than the normal condi-

tion [38] and thus, brain ammonia level, than the concen-

tration of ammonia in blood, is considered more relevant

for interpreting the data obtained using HA animal models

[24]. In the present report, we have used hyperammonemic

rats induced by administration of ammonium acetate

wherein, as reported earlier [24], ~5–7 and 1.5–1.8 fold

increases in brain ammonia level was considered as acute

and chronic HA groups respectively.

Brain processes ~20% of O2 consumed by the whole

body for generating ATP via oxidative phosphorylation in

mitochondria and therefore, brain cells are consistently

exposed to high ROS. Abundance of myelinated nerve

fibers makes brain enriched with phospholipids containing

poly unsaturated fatty acids, and thus, brain cells become

highly prone to ROS dependent derangements in mem-

brane structure and functions [39]. The level of lipid

peroxidation is a good indicator to assess the extent of

oxidative damage produced by ROS in the brain. The over

activation of NMDA receptors [1] and ammonia induced

mitochondrial dysfunction [4, 5] could be the main source

of excess of ROS in brain during HA. The rate of free

radical production and the level of lipid peroxidation have

been reported to be significantly high in the whole brain of

acute HE rats [40]. According to Table 1, however, when

compared between the cerebral cortex and the cerebellum

in pure HA rats, significantly increased level of lipid per-

oxidation (~ 2 times higher) in cerebellum than the cortex

under both acute and chronic conditions clearly suggest

that cerebellum is more susceptible for oxidative damage

due to ammonia toxicity than the cortex. Furthermore, ~ 3

and 2 fold increases in the MDA level in cerebellum of

acute and chronic HA rats respectively suggest for a cause

and effect relationship between the degree of HA in brain

and the oxidative damage in cerebellum. Nonetheless,

Table 1 Effect of acute and chronic hyperammonemia on the level of lipid peroxidation and total thiol (GSH) in cerebral cortex and cerebellum

Tissues Biochemical parameter Control Acute Control Chronic

Cerebral cortex Lipid peroxidation (MDA nmol/g wet wt) 58.65 ± 6.85 78.30 ± 5.8* 55.4 ± 5.41 55.09 ± 4.0

Total thiol (GSH) (nmol/mg protein) 1.35 ± 0.075 1.26 ± 0.125 1.37 ± 0.193 1.28 ± 0.13

Cerebellum Lipid peroxidation (MDA nmol/g wet wt) 54.50 ± 5.36 163.07 ± 7.12*** 54.72 ± 5.82 108.02 ± 8.48**

Total thiol (GSH) (nmol/mg protein) 1.28 ± 0.166 1.04 ± 0.080 1.06 ± 0.114 1.17 ± 0.075

Values are mean ± SD where n = 4 and each experiment done in duplicates* P < 0.05, **P < 0.01, ***P < 0. 001 (Control versus experimental group)

Neurochem Res (2008) 33:103–113 107

123

cortex showed resistance to HA dependent oxidative stress,

as there was no change in MDA level in the cortex of

chronic but with a mild (1.3 fold) increase in that from

acute HA rats.

The level of reduced glutathione (GSH), a tripeptide

responsible to maintain reducing equivalents under oxida-

tive stress, is another critical factor to assess the level of

oxidative stress in mammalian cells. Interestingly, there

was no significant change in the level of GSH in both the

brain regions under acute as well as chronic HA conditions

(Table 1). In the cellular antioxidant pathway, the turnover

of GSH/GSSG is regulated by synchronized activities of

GPx and GR in mammalian cells. Both these enzymes did

not show much alternation, except a moderate decrease in

GPx and GR in cortex and cerebellum respectively

(Figs. 3, 4), due to acute HA, and thus, could be correlated

with the unchanged level of GSH in both the brain regions

during acute HA. However, significantly opposite trends of

GPx and GR in the cerebral cortex of chronic HA rats

(Figs. 7, 8) did not correlate with the unchanged level of

GSH in the cortex of rats with chronic HA. It is suggested

0

2

4

6

8

01

21

Cerebral cortex Cerebellum

Cerebral cortex Cerebellum

SO

D (

U/m

g p

rote

in)

control

HA

DOS

***

**

lortnoC HA C lortno AH

A

B

0

02

04

06

08

01 0

21 0

41 0

% o

f co

ntr

ol

Control

HA

**

*C

Fig. 1 Effect of acute hyperammonemia on activity (A) and level of

active SOD protein (B & C) in cerebral cortex and cerebellum of rats.

The values in A represent mean ± SD where n = 4 and each

experiment done in duplicate. In case of B, pooled tissue extracts

from 4 rats containing 60 lg protein in each lane was electropho-

ressed on 12% non- denaturing PAGE followed by substrate specific

development of SOD bands. The gel photograph in B is a

representative out of the 4 PAGE repeats. In panel C, relative

densitometric values of SOD bands from experimental group as % of

the control lane have been presented as mean ± SD from the 4 PAGE

repeat experiments *P < 0.05, **P < 0.01, ***P < 0.001 (control

versus experimental groups)

0

.0 5

1

.1 5

2

.2 5

3

.3 5

4

Cerebr c la ort xe Cerebe mull

Cat

alas

e (U

/mg

pro

tein

)

oC ntrol

AH

esalataC

***

***

lortnoC HA C lortno AH

A

B

0

02

04

06

08

01 0

21 0

Cer rbe c la ort xe Cer mullebe

% o

f co

ntr

ol

C

Fig. 2 Effect of acute hyperammonemia on activity (A) and level of

active catalase protein (B & C) in cerebral cortex and cerebellum of

rats. The values in A represent mean ± SD where n = 4 and each

experiment done in duplicates. In case of B, pooled tissue extracts

from 4 rats containing 60 lg protein in each lane was electropho-

ressed on 8% non- denaturing PAGE followed by substrate specific

development of catalase bands. The gel photograph in B is a

representative out of the 3 PAGE repeats. In panel C, relative

densitometric values of catalase bands from experimental group as %

of the control lane have been presented as mean ± SD from the 3

PAGE repeat experiments. ***P < 0.001 (control versus experimental

groups)

108 Neurochem Res (2008) 33:103–113

123

that a highly adaptive metabolic coupling operates between

astrocytes and neurons to maintain the normal level of this

tripeptide under unphysiological conditions in brain [41,

42]. When neuron’s GSH gets depleted due to acute

ammonia intoxication, the precursors for GSH synthesis

are supplied from astrocytes which are supposed to be less

susceptible to ROS insult [41]. Furthermore, gamma glut-

amyl-cystein synthetase is also responsible to produce GSH

in the cells, and this enzyme has been reported to be in-

creased in the astrocytes under acute HA condition [43].

Thus, it is likely that these additional routes could con-

tribute for maintaining GSH level in the cortex of chronic

HA rats even when GR activity declined significantly

(P < 0.01). Similar argument may be given for the unal-

tered level of GSH in cerebellum of chronic HA rats where,

GPx showed significant decline (P < 0.05) but with a little

change in GR activity (Figs. 7, 8).

The changes in the levels of antioxidant enzymes during

oxidative stress are the most critical factors in determining

the extent of oxidative damage produced by ROS during

neuropathology [17, 44]. All parts of brain contain SOD,

catalase, GPx and GR in high concentration to counter

balance the deleterious effects of ROS [44, 45]. Excess

of superoxide anion (O2–), the major ROS produced in

mitochondria, is converted to H2O2 by SOD. Simultaneous

removal of H2O2 by either catalse and/or by GPx is crucial

for preventing membrane damage due to oxidative stress.

In brain, SOD-GPx-GR pathway is considered to play

major role of antioxidant activities [19, 20]. With the

increased production of ROS, most of these enzymes were

found to be declined in whole brain of rat with acute HE

[8]. However, according to the results presented here, when

the levels of all these enzymes were compared in concerted

manner in two different brain regions (cerebral cortex and

cerebellum) under acute and chronic HA, changes in all

these enzymes were found to differ as a function of degree

of HA but with a regional specificity. In cerebellum,

though GPx showed resistance against acute HA, there was

a significant decline in the levels of SOD, catalase and GR

under acute HA and thus, suggested for acute HA depen-

dent oxidative stress in rat cerebellum. It was also corre-

lated well with a significant increase in the level of lipid

peroxidation in cerebellum of acute HA rats (Table 1).

Cerebral cortex also showed significant decline in SOD and

catalase, however, with a moderate decrease in GPx and no

change in the level of GR under acute HA (Figs. 3, 4). In

view of relatively less increase in the level of lipid

peroxidation due to acute HA in the cortex (~ 2 times less

than cerebellum), it may be assumed that resistance of GPx

and GR to acute HA might be accountable to prevent

oxidative damage in cerebral cortex even at the face of

significant decline in SOD and catalase.

During chronic HA, a significant decline in the level of

SOD (both by activity and PAGE results) and GPx (Figs. 5,

7) coincided with the significant increase in the level of

lipid peroxidation in cerebellum (Table 1), however, with

no change in the level of catalase and GR (Figs. 6, 8).

This suggests that decline in the level of SOD and GPx

are mainly accountable to allow oxidative damage in

cerebellum and unaffected GR plays a permissive role in

1 xPG

2 xPG

3 xPG

4 xPG

lortnoC HA C lortno AHA

B

0

02

04

06

08

01 0

21 0

41 0

Cer rbe c la ort xe Cer mullebe

% o

f co

ntr

ol *

Fig. 3 Effect of acute hyperammonemia on level of active GPx

protein in cerebral cortex and cerebellum of rats. In case of A, pooled

tissue extracts from 4 rats containing 30 lg protein in each lane was

electrophoressed on 10% non-denaturing PAGE followed by substrate

specific development of GPx bands. The gel photograph in A is a

representative out of the 4 PAGE repeats. In panel B, relative

densitometric values of GPx bands from experimental group as % of

the control lane have been presented as mean ± SD from the 4 PAGE

repeat experiments. *P < 0.05 (control versus experimental group)

0

5

10

15

20

25

30

35

40

Cerebral cortex Cerebellum

GR

(U

/mg

pro

tein

)

Control

HA

*

Fig. 4 Effect of acute hyperammonemia on activity of GR in cerebral

cortex and cerebellum of rats. The values represent mean ± SD where

n = 4 and each experiment done in duplicates. *P < 0.05 (control

versus experimental groups)

Neurochem Res (2008) 33:103–113 109

123

maintaining the normal level of GSH (Table 1) during

chronic HA in this brain region. This again supports the

view that SOD and GPx are the most critical antioxidant

enzymes in brain [19, 20]. Nonetheless, since, both these

enzymes declined specifically in cerebellum (as compared

to the cortex) and that cerebellum associated functions are

affected the most in HE patients [14, 15], it may be argued

that relatively greater susceptibility of cerebellum for

ammonia toxicity dependent antioxidant defense could be

accountable for pathogenesis of low grade chronic HA.

In case of cortex, contrary to the effect of acute

ammonia exposure, chronic HA produced significant in-

creases in the levels of SOD, catalase and GPx (Figs. 5–7)

and thus, suggested positive adaptation in brain cortex

against a low grade chronic HA with respect to these

antioxidant enzymes. Brain is considered to be a highly

plastic tissue so far metabolic adaptations are concerned

[18]. The whole brain of rats pre-exposed to chronic HA

have been found to resist the changes in the level of crucial

metabolites which are normally produced otherwise during

acute HA [8]. At the face of significant decline in the

activity of most of the antioxidant enzymes, SOD activity

was reported to be increased significantly in all the brain

regions of rats with fulminate liver type acute HE [10]. As

per the results presented here, however, it is evident that

chronic HA produces adaptive changes only in cerebral

cortex with respect to the SOD-GPx pathway in particular.

This could contribute for relatively less effect of chronic

HA on the cortex associated function than the cerebellum

(Fig. 8).

It has been suggested that each antioxidant enzyme

has a functionally distinct role, or cooperates with other

enzymes to protect the cell under a variety of pathophysi-

ological conditions [46] and thus, HA dependent differen-

tial changes in the set of antioxidant enzymes e.g. up

regulation of SOD-GPx in cortex and their down regulation

0

2

4

6

8

01

21

41

61

Cer rbe c la ort xe Cer mullebe

SO

D (U

/mg

pro

tein

)

oC ntrolAH

DOS

**

***

oC lortn AH oC lortn AH

A

B

0

02

04

06

08

01 0

21 0

41 0

61 0

81 0

Cer rbe c la ort xe Cer mullebe

% o

f co

ntr

ol

C**

**

Fig. 5 Effect of chronic hyperammonemia on activity (A) and level

of active SOD protein (B & C) in cerebral cortex and cerebellum of

rats. The values in A represent mean ± SD where n = 4 and each

experiment done in duplicate. In case of B, pooled tissue extracts

from 4 rats containing 60 lg protein in each lane was electropho-

ressed on 12% non-denaturing PAGE followed by substrate specific

development of SOD bands. The gel photograph in B is a

representative out of the 4 PAGE repeats. In panel C, relative

densitometric values of SOD bands from experimental group as % of

the control lane have been presented as mean ± SD from the 4 PAGE

repeat experiments. **P < 0.01, ***P < 0.001 (control versus exper-

imental groups)

0

1

2

3

4

5

6

7

Ce er bral c o tr ex Ce er bellum

Ce er bellum

Cat

alas

e (U

/mg

pro

tein

)

oC ntrolAH

taC al esa

***

lortnoC HA C lortno AH

A

B

0

20

40

60

80

100

120

140

Cerebral cortex

% o

f co

ntr

ol

C*

Fig. 6 Effect of chronic hyperammonemia on activity (A) and level

of active catalase protein (B & C) in cerebral cortex and cerebellum

of rats. The values in A represent mean ± SD where n = 4 and each

experiment done in duplicates. In case of B, pooled tissue extracts

from 4 rats containing 60 lg protein in each lane was electropho-

ressed on 8% non-denaturing PAGE followed by substrate specific

development of catalase bands. The gel photograph in B is a

representative out of the 3 PAGE repeats. In panel C, relative

densitometric values of catalase bands from experimental group as %

of the control lane have been presented as mean ± SD from the 3

PAGE repeat experiments. *P < 0.05, ***P < 0.001 (control versus

experimental groups)

110 Neurochem Res (2008) 33:103–113

123

in cerebellum during chronic HA could be the result of

differential sensitivity of cortex and cerebellum to chronic

HA. Opposite responses of cortex and cerebellum to NO

dependent signaling pathway during HA in rats [12] and

also in HE patients [16] provide support to this argument.

Such a pattern has been shown in other neurological

disorders also. Different antioxidant enzymes showed

differential alterations in the brain of patients with

Alzheimers type dementia [21] and also in D-amphitamine

induced neurotoxicity [22]. Increase in the level of SOD

and catalase in different brain regions of rats with mala-

thion-induced oxidative stress is another example of

adaptive changes in antioxidant enzymes [23].

With a view to have a molecular rationale behind

significant changes in the activities of SOD, catalase and

GPx during HA, these enzymes were further analyzed on

PAGE. It was interesting to note that while level of SOD

protein increased in cerebellum of rats with acute HA

(Figs. 1B, C), activity of this enzyme (when measured

in vitro) showed significant decline (Fig. 1A). Similar

pattern was observed with catalase in both the brain regions

of acute HA rats (Fig. 2A–C). Such a mismatch could be

resulted due to SOD and catalase inhibitory conditions

induced in brain during acute HA. H2O2 is a known

physiological inhibitor of SOD [47, 48] and has been

demonstrated recently to inhibit specific isoforms of this

enzyme in brain [49]. Increased accumulation of Mn2+ in

brain is associated with Alzheimers type II astrocytosis

[50], a hall mark of acute HA [1] and as reviewed by

Takeda [51], increased level of Mn2+ inhibits catalase and

also induces a burst of H2O2 in brain cells. Also, as per the

results presented here (Fig. 2A), a drastic decrease in the

activity of catalase in cerebellum of acute HA rats may also

contribute for an unusual increase in H2O2 and thus, can

further potentiate inhibition of SOD in this brain region

during acute HA. A two times higher level of MDA in

cerebellum than the cortex of acute HA rats (Table 1)

provide additional support to this argument. Furthermore, it

has been demonstrated that inhibition of SOD at cellular

level induces increase in the mRNA level of this enzyme

[52], and SOD proteins are highly resistant to denaturation

& oxidative damage even at a high concentration of H2O2

[48]. Therefore, it is likely that inactivation of SOD

observed in cell free extracts due to increased oxidative

burst in cerebellum of acute HA rats might not be reflected

at protein level (Fig. 1). Accordingly, since inactivation

of SOD would be expected to be minimal during mild

oxidative stress, in vitro activity data and level of SOD

protein should be mutually correlative. And indeed, a

similar pattern of SOD profile was observed in the cerebral

cortex of acute HA rats (Fig. 1) with ~ 2 times less

oxidative stress than cerebellum (Table 1, MDA data). As

Mn2+ also inhibits catalase in the brain [51] and such

transitory metal-protein interaction is likely to get disso-

ciated during electrophoresis, a similar argument may be

given for significant decreases in the activity of catalase

in vitro but with insignificant change in its level on PAGE

analysis in both the brain regions of acute HA rats. These

1 xPG

2 xPG

3 xPG

4 xPG

oC lortn AH oC lortn HAA

.0 00

02 .00

04 .00

06 .00

08 .00

01 .0 00

21 .0 00

41 .0 00

Cer rbe c la ort xe Cer mullebe

% o

f co

ntr

ol

*

*

B

Fig. 7 Effect of chronic hyperammonemia on level of active GPx

protein in cerebral cortex and cerebellum of rats. In case of A, pooled

tissue extracts from 4 rats containing 30 lg protein in each lane was

electrophoressed on 10% non-denaturing PAGE followed by substrate

specific development of GPx bands. The gel photograph in A is a

representative out of the 4 PAGE repeats. In panel B, relative

densitometric values of GPx bands from experimental group as % of

the control lane have been presented as mean ± SD from the 4 PAGE

repeat experiments. *P < 0.05 (control versus experimental group)

0

5

10

15

20

25

30

35

Cerebral cortex Cerebellum

GR

(U

/mg

pro

tein

)

Control

HA

**

Fig. 8 Effect of chronic hyperammonemia on activity of GR in

cerebral cortex and cerebellum of rats. The values represent

mean ± SD where n = 4 and each experiment done in duplicates.**P < 0.01 (control versus experimental groups)

Neurochem Res (2008) 33:103–113 111

123

arguments get further support from a uniform correlative

pattern observed between in vitro data and PAGE patterns

of SOD and catalase in both the brain regions of chronic

HA rats (Figs. 5, 6) showing significantly less oxidative

stress as compared to the acute HA rats (Table 1, MDA

data). Thus, it is evident that the extent of oxidative stress

induced during acute HA acts as an additional factor in

modulating the activities of SOD and catalase irrespective

of the actual levels of these proteins in both the brain

regions.

In conclusion, active levels of all the antioxidant enzymes

were found altered differently in cerebral cortex and cere-

bellum as a function of degree of HA. As compared to a

uniform decline in the activities of most of the antioxidant

enzymes due to acute HA, chronic HA was found to induce

brain region specific changes which are likely to render

cerebellum susceptible and cerebral cortex resistant to the

oxidative stress during chronic HA. Since, cerebellum asso-

ciated functions are mainly affected during low grade chronic

HA, such changes in antioxidant enzymes might be impli-

cated in the encephalopathy of chronic HA.

Acknowledgments This work was financially supported by a DAE:

BRNS grant (P-29/64) to SKT. The instrumental facilities provided

by the DST FIST and UGC-CAS Program to the department of

Zoology are also acknowledged.

References

1. Felipo V, Butterworth RF (2002) Neurobiology of ammonia.

Prog Neurobiol 67:259–279

2. Butterworth RF (2000) Hepatic encephalopathy: a neuropsychi-

atric disorder involving multiple neurotransmitter systems. Curr

Opin Neurol 13:721–727

3. Marcaida G, Felipo V, Hermengildo C et al (1992) Acute

ammonia toxicity is mediated by the NMDA type of glutamate

receptor. FEBS Lett 296:67–68

4. Rama Rao KV, Norenberg MD (2001) Cerebral energy metabo-

lism in hepatic encephalopathy and hyperammonemia. Metab

Brain Dis 16:67–78

5. Kosenko E, Felipo V, Montoliu C et al (1996) Effects of acute

hyperammonemia in vivo on oxidative metabolism in nonsy-

naptic rat brain mitochondria. Metab Brain Dis 12:69–82

6. Rama Rao KV, Jayakumar AR, Norenberg MD (2005) Role of

oxidative stress in the ammonia-induced mitochondrial perme-

ability transition in cultered astrocytes. Neurochem Int 47:31–38

7. Murthy CRK, Rama Rao KV, Bai G et al (2001) Ammonia in-

duced production of free radicals in primary culture of rat as-

trocytes. J Neurosci Res 66:282–288

8. Kosenko E, Kaminski Y, Lopata O et al (1999) Blocking of

NMDA receptors prevents the oxidative stress induced by acute

ammonia intoxication. Free Radical Res 26:1369–1374

9. Norenberg MD, Jayakumar AR, Rama Rao KV (2004) Oxidative

stress in the pathogenesis of hepatic encephalopathy. Metab Brain

Dis 19:313–329

10. Sathyasaikumar KV, Swapna I, Reddy PVB et al (2007)

Fulminant hepatic failure in rats induces oxidative stress differ-

entially in cerebral cortex, cerebellum and pons medula. Neuro-

chem Res 32:517–524

11. Hermenegildo C, Montoliu C, Llansola M et al (1998) Chronic

hyperammonemia impairs the glutamate-nitric oxide-cyclicGMP

pathway in cerebellar neurons in culture and in the rat in vivo.

Eur J Neurosci 10:3201–3209

12. Rodrigo R, Felipo V (2006) Brain regional alternations in the

modulation of the glutamate-nitric oxide-cGMP pathway in liver

cirrhosis: role of hyperammonemia and cell types involved.

Neorochem Int 48:472–477

13. Kosenko E, Kaminsky YG, Felipo V et al (1993) Chronic

hyperammonemia prevents changes in brain energy and ammonia

metabolism induced by acute ammonia intoxication. Biochim

Biophys Acta 1180:321–326

14. Gilberstadt SJ, Gilberstadt H, Zieve L et al (1980) Psychomotor

performance defects in cirrhotic patients without overt encepha-

lopathy. Arch Int Med 140:519–521

15. Tarter RE, Hegedus AM, Van Thiel DH et al (1984) Non-

alcoholic cirrhosis associated with neuropsychological dysfunc-

tion in the absence of overt evidence of hepatic encephalopathy.

Gastroenterology 86:1421–1427

16. Corbalan R, Chaturet N, Behrends S et al (2002) Region selective

alternation of soluble guanylate cyclase content and modulation

in brain of cirrhotic patients. Hepatology 6:1155–1162

17. Mates JM (2000) Effects of antioxidant enzymes in the molecular

control of reactive oxygen species toxicology. Toxicology

153:83–104

18. Magistretti PJ (2003) Brain energy metabolism. In: Fundamental

neuroscience, 2nd edn. Elsevier Science, New York, pp339–360

19. Tureen JF (2003) Mitochondrial formation of reactive oxygen

species. J Physiol 552:335–344

20. Flohe RB (1999) Tissue-specific function of individual glutathi-

one peroxidases. Free Radical Biol Med 27:951–965

21. Gsell W, Conrad R, Hickethier M et al (1995) Decreased catalase

activity but unchanged superoxide dismutase activity in brain of

patients of Alzheimer type. J Neurochem 64:1216–1233

22. Frey BN, Valvassori SS, Reus GS et al (2006) Changes in antioxidant

defense enzymes after D-amphetamine exposure: implications as

an animal model of mania. Neurochem Res 31:699–703

23. Furtunato JJ, Feier G, Vitali AM (2006) Malathione-induced

oxidative stress in rat brain regions. Neurochem Res 31:671–678

24. Hilgier W, Albrecht J, Lisy V et al (1990) The effect of acute and

chronic hyperammonemia on y-glutamyl transpeptidase in vari-

ous brain regions of young and adult rats. Mol Chem Neuropathol

13:47–56

25. Pandey P, Singh SK, Trigun SK (2005) Developing brain of

moderately hypothyroid mice shows adaptive changes in the key

enzymes of glucose metabolism. Neurol Psych Brain Res 12:159–164

26. Lowry OH, Rosebrough NJ, Farr AL et al (1951) Protein mea-

surement with the Folin phenol reagent. J Biol Chem 193:

265–275

27. Placer ZA, Cushman LL, Johnson BC (1966) Estimation of

product of lipid peroxidation (malonyl dialdehyde) in biochemi-

cal systems. Anal Biochem 16: 359–364

28. Sedlak J, Raymond HL (1968) Estimation of total thiol, protein

bound, and non protein sulfhydryl groups in tissue with Ellman’s

reagent. Anal Biochem 25:192–205

29. Kakkar P, Das B, Viswanathan PN (1984) A modified spectro-

photometric assay of superoxide dismutase ( SOD). Ind J

Biochem Biophys 21:130–132

30. Sinha KA (1972) Colorimetric assay of catalase. Anal Biochem

47:389–394

31. Trigun SK, Singh AP, Asthana RK et al (2006) Assessment of

bioactivity of a Fischerella species colonizing Azadirachta Indica(neem) Bark. Appl Ecol Environ Res 4:119–128

32. Beauchamp C, Fridovich I (1971) Superoxide dismu-

tase:improved assays and an assay applicable to acrylamide gels.

Anal Biochem 44:276–287

112 Neurochem Res (2008) 33:103–113

123

33. Sun Y, Elwell JH, Oberley LW (1988) A simultaneous visuali-

zation of the antioxidant enzymes glutathione peroxidase and

catalase on polyacrylamide gels. Free Radical Res Commun

5:67–75

34. Lin CL, Chen HJ, Hou WC (2002) Activity staining of gluta-

thione peroxidase after electrophoresis on native and sodium

dodecylsulfate polyacrylamide gels. Electrophoresis 23:513–516

35. Carlberg I, Mannervik B (1975) Purification and characterization

of the flavoenzyme glutathione reductase from rat liver. J Biol

Chem 250:5475–5480

36. Wang H (2000) Over expression of L-PhGPx in MCF-7 cells.

In: The role of mitochondrial phospholipids hydroperoxide glu-

tathione peroxide in cancer therapy. Ph.D. thesis. The University

of Iowa, Iowa, pp16–56

37. Kanamori K, Ross BD, Chung JC et al (1996) Severity of hy-

perammonemic encephalopathy correlates with brain ammonia

level and saturation of glutmine synthetase in vivo. J Neurochem

67:1584–1594

38. Lockwood AH (2004) Blood ammonia levels and hepatic

encephalopathy. Metab Brain Dis 19:345–349

39. Halliwell B, Gutteridge JMC (1985) Oxygen radicals and nervous

system. Trends Neurosci 8:22–26

40. Kosenko E, Venediktova N, Kamanisky Y et al (2003) Sources of

oxygen radical in brain in acute ammonia intoxication in vivo.

Brain Res 981:193–200

41. Dringen R, Pawlowski PG, Hirrlinger J (2005) Peroxide detoxi-

fication by brain cells. J Neurosci Res 79:157–165

42. Hovatta I, Tennant RS, Helton R et al (2005) Glyoxalase 1 and

glutathione reductase 1 regulate anxiety in mice. Nature 438:

662–666

43. Murthy Ch RK, Bender AS, Dombro RS et al (2000) Elevation of

glutathione level by ammonium ions in primary cultures of rat

astrocytes. Neurochem Int 37:255–268

44. Halliwell B (2001) Role of free radicals in the neurodegenerative

diseases: therapeutic implications for antioxidant treatment. Drug

Aging 18:685–716

45. Patenaude A, Murthy MR, Mirault ME (2005) Emerging roles of

thioredoxin cycle enzymes in the central nervous system. Cell

Mol Life Sci 62:1063–1080

46. Imai H, Nakagawa Y (2003) Biological significance of phos-

pholipid hydroperoxide glutathione peroxidase (PHGPx, GPx4)

in mammalian cells. Free Radical Biol Med 34:145–169

47. Bray RC, Cockel SA, Fielden EM et al (1974) Reduction and

inactivation of superoxide dismutase by hydrogen peroxide.

Biochem J 139:43–48

48. Salo DC, Pacifici RE, Lin SW (1990) Superoxide dismutase

undergoes proteolysis and fragmentation following oxidative

modification and inactivation. J Biol Chem 265:11919–11927

49. Mokni M, Elkahoui S, Limam F et al (2007) Effect of resveratrol

on antioxidant enzyme activities in the brain of healthy rat.

Neurochem Res 32:981–987

50. Hazell AS, Normandin L, Norenberg MD et al (2006) Alzheimer

type II astrocytic changes following sub-acute exposure to

manganese and its prevention by antioxidant treatment. Neurosci

Lett 396:167–171

51. Takeda A (2003) Manganese action in brain function. Brain Res

Rev 41:79–87

52. Maitre B, Jornot L, Jonod AF (1993) Effects of inhibition of

catalase and superoxide dismutase activity on antioxidant enzyme

mRNA levels. Am J Physiol 265:636–643

Neurochem Res (2008) 33:103–113 113

123