TOXICOLOGIC PATHOLOGY, vol 30, no 3, pp 365– 372, 2002

Copyright C 2002 by the Society of Toxicologic Pathology

Alpha-Glutathione S-Transferase in the Assessment

of Hepatotoxicity—Its Diagnostic Utility in Comparison
with Other Recognized Markers in the Wistar Han Rat
PAUL S. GIFFEN,1 CHRIS R. PICK,2 MARK A. PRICE,3 ANN WILLIAMS,2 AND MALCOLM J. YORK1
1
Clinical Pathology, Cellular and Biochemical Toxicology, Safety Assessment, GlaxoSmithKline Research and Development,
Ware, Hertfordshire, United Kingdom
2
Department of Pathology, Preclinical Safety Sciences, GlaxoSmithKline Research and Development, Ware, Hertfordshire,
United Kingdom
3
Department of Toxicology, Preclinical Safety Sciences, GlaxoSmithKline Research and Development, Ware, Hertfordshire,
United Kingdom

ABSTRACT
The diagnostic utility of alpha-glutathion e S-transferase (aGST) in the assessment of acute hepatotoxicity was compared with a range of markers
including alanine aminotransferas e (ALT) and aspartate aminotransferas e (AST). Rats were given a single oral dose of either a-naphthylisothiocynate
(ANIT), bromobenzen e (BrB), or thioacetamide (TAM) at concentrations previously shown to induce marked hepatotoxicity. The progression of each
hepatic lesion was monitored by the measurement of a battery of markers, including aGST, in plasma collected at time points ranging from 3 h to
7 days after dosing. aGST was seen to increase signiŽcantly at 24 h (ANIT/BrB) and 3 h (TAM) postdosing, correspondin g with histopathologica l
Žndings. For each compound, when the degree of insult was most severe, fold increases in aGST were greater than those seen with ALT and AST,
yet lower than those seen with glutamate dehydrogenas e (BrB and ANIT), sorbitol dehydrogenas e (TAM), or total bilirubin and bile acids (ANIT).
Elevations in aGST were also detected no earlier than any other marker. aGST in the rat was shown to be a valid marker of hepatotoxicity ; however,
its measuremen t offered no additional information in detecting either the time of onset/recovery or the severity of each type of hepatic injury induced.
Keywords. Bromobenzene (BrB); a-glutathione S-transferase (aGST); hepatic markers; liver enzymes; a-naphthylisothiocynate (ANIT);
thioacetamide (TAM).

INTRODUCTION midzonal hepatotoxicity may be questioned. AST has a ubiq-

The glutathione S-transferases (EC 2.5.1.18 ) form a com-
plex family of multifunctional proteins displaying various bi-
ological functions (3), their main purpose being the phase II
detoxiŽcation of xenobiotics. This family of enzymes con-
sists of cytosolic isoforms (alpha, mu, pi, and theta) as well
as a more recently isolated microsomal form (6). In humans,
80% of all alpha-glutathione S-transferase (aGST) is found
within the liver (6) where it comprises 5 – 10% of the solu-
ble hepatic protein (7). This has resulted in much interest in
the measurement of aGST, via the use of species speciŽc
enzyme immunoassays (EIA) (8, 12), as a superior marker
of hepatotoxicity. Further advantages, such as its cytosolic
localization, small size ( Mr 50,000 ), and short half-life (ap-
proximately 90 min), potentially offer more information re-
garding acute changes to hepatocellular integrity than other
markers such as alanine aminotransferase (ALT) (EC 2.6.1.2 )
and aspartate aminotransferase (AST) (EC 2.6.1.1 ).
Traditionally, ALT and AST are the enzymes most com-
monly used in the assessment of hepatocellular status. Nev-
ertheless, it is recognized that both ALT and AST are more
abundant within the periportal region of the liver lobule (6),
and as such, their utility in the assessment of centrilobular or
Address correspondenc e to: P. S. Giffen, Cellular and Biochemical Tox-
icology, Safety Assessment, GlaxoSmithKline Research and Development ,
Park Road, Ware Hertfordshire , UK, SG12 0DP; e-mail: psg21280@
gsk.com.
365
Downloaded from tpx.sagepub.com by guest on July 15, 2015
uitous distribution (5) with signiŽcant activities in the heart,
liver, kidney, and skeletal muscle, respectively. It should
therefore not solely be used as an indicator of liver dam-
age unless other supporting enzymes are measured. Investi-
gations using the rat have been performed by several groups,
comparing the utility of aGST with the transaminases, using
carbon tetrachloride (1, 2, 6, 7) and bromobenzene (2). Both
of these compounds are recognized to induce centrilobular
necrosis; the region of the liver known to have lower levels
of ALT and AST. Little published data is available compar-
ing the diagnostic utility of aGST in comparison to markers
other than ALT or AST.
It was therefore decided to compare the diagnostic util-
ity of aGST in the assessment of hepatic injury induced by
different toxins in comparison with a wide range of liver in-
jury markers including ALT, AST, glutamate dehydrogenase
(GLDH) (EC 1.4.1.3 ), sorbitol dehydrogenase (SDH) (EC
1.1.1.14 ), and alkaline phosphatase (ALP) (EC 3.1.3.1 ). In
addition, markers of liver dysfunction were measured, in-
cluding total bile acids (BA ) and total bilirubin (T-bil), as
well as correlation to histopathological Žndings.
Preliminary work using allyl alcohol, N-
nitrosodiethylamine (NDEA), and N-nitrosodimethylamine
(NDMA) as hepatotoxins generated progressive liver lesions
of varying severity in a dose-dependent manner. Despite
the varying degrees of insult seen, aGST was not seen to
increase any more signiŽcantly than alternate parameters
measured at any time point. It was recognized, however,
0192-6233/02$3.00 $0.00
366 GIFFEN ET AL
that the time points at which samples were collected on
these studies (either 24 or 48 h and 168 h postdose) did not
address the early onset of the hepatic injury. Based on these
data, it was decided to investigate the potential of aGST as
a marker of acute hepatotoxicity further, by monitoring the
progression of differing hepatic lesions over a greater range
of time points.
The model hepatotoxins a-naphthylisothiocynate (ANIT)
(10, 11), bromobenzene (BrB ) (9), and thioacetamide (TAM)
(13) were chosen to induce varying zonal patterns of hepatic
insult. Previous studies with these compounds have shown
that they elicit hepatotoxicity within 24 – 48 h, indicated by
elevations in ALT and AST.
As BrB and TAM have also been described as potent
nephrotoxic agents (4), renal function was monitored by the
measurement of plasma creatinine and urea.
MATERIALS AND METHODS
Animals
Male Wistar Han rats [crl:WI(Glx/BRL/Han )BR] aged
10 – 11 weeks (BrB), 8 – 9 weeks (ANIT), and 9 – 10 weeks
(TAM) were used. Each animal was housed singly through-
out the study with food and water available ad libitum. All
animals were checked twice daily for signs of ill health
and body weights, and clinical observations were performed
daily. All studies were performed in accordance with Home
OfŽce guidelines speciŽc for GlaxoSmithKline Research and
Development.
Bromobenzene, a-Naphthylisothiocynate,
and Thioacetamide Studies
Rats were administered a single oral dose of BrB (1.5 g/kg ),
ANIT (150 mg/kg), or TAM (150 mg/kg) and autopsied at 3,
7, 24, 31, and 168 h postdose (n 5 rats/toxin/time point).
Each compound was obtained from Sigma Aldrich, Poole,
Dorset, and formulated in corn oil to give a Žnal dose vol-
ume of 10 ml/kg. All control animals were dosed with corn oil
only. Final autopsies were performed at 54 and 72 h postdose,
rather than 168 h, for animals treated with BrB and TAM, re-
spectively, due to animal ill health (progressive weight loss
and persistently hunched posture). Vehicle control animals,
dosed with corn oil only, were autopsied at 24 and 31 h
(n 5 rats/time point).
Sample Collection
At each time point, 0.6 ml blood was taken from each an-
imal, via aortic exsanguination under isourane anesthesia,
into heparinized microtainers (Becton Dickinson, Plymouth,
UK). Samples were centrifuged at room temperature
(15 – 25 C) within 1 h of blood collection and plasma was
separated for analysis. Plasma for aGST analysis was frozen
at 15 to 25 C until required. All other parameters were
measured using fresh plasma, stored at 2– 8 C, within 24 h of
collection.
Tissue Collection and Slide Preparation
At each time point postadministration of BrB, ANIT, or
TAM, the following tissues were Žxed in 10% buffered for-
malin, conventionally processed, and sectioned before stain-
ing with hematoxylin and eosin: heart, kidney (right), liver,
Downloaded from tpx.sagepub.com by guest on July 15, 2015
TOXICOLOGIC PATHOLOGY
esophagus, spleen, stomach, testes (right), bladder, and cere-
bral cortex.
Tissue Enzyme Distribution and Freeze – Thaw
Study—Tissue Homogenate Preparation
Samples of liver, heart, skeletal muscle, kidney, testes,
spleen, and small intestine were obtained from anesthetized
male rats (n 1). These were taken into 10 mM phosphate-
buffered saline (PBS) and frozen at 75 to 85 C until re-
quired. Tissues were homogenized in a ratio of 1 g tissue per
10 ml PBS (chilled) using an Ultra-Turrax T8 homogenizer.
The resultant suspension was centrifuged at 4 C at 6000 rpm
for 15 min. The supernatant was removed and stored at 2 – 8 C
prior to analysis. The samples were diluted in sample speciŽc
diluent (aGST) and 10 mM PBS (ALT, AST, aGST, GLDH,
SDH, and ALP).
Subsequently, liver was collected from anesthetized male
rats (n 3) into 10 mM PBS. Homogenates of each liver
were prepared, as described earlier, using sections of each
liver, pre- and postfreezing. Homogenates prepared from
fresh (n 3) and frozen liver (n 3) were each analyzed
for ALP, ALT, AST, GLDH, and SDH. aGST was assayed
in homogenate prepared from fresh (n 1) and frozen liver
(n 1).
aGST Analysis
aGST was measured using the Hepkit-Rt enzyme im-
munoassay (Biotrin International, Dublin, Ireland). The assay
procedure is based upon the sequential addition of sam-
ple, antibody – enzyme conjugate, and substrate to microw-
ells coated with anti-rat aGST immunoglobulin G (IgG).
The resultant color formation was proportional to the aGST
concentration within the sample. The assay range was
0 – 100 l g/L. Plate washing was performed using a Wave
100 automated plate washer (Dynex Technologies, Ashford,
Middlesex ). Curve Žtting and data handling were performed
automatically using an MRX automated plate reader (Dynex
Technologies).
Spectrophotometric Analysis
All parameters were measured at 37 C using a Hitachi 917
automated clinical chemistry analyzer (Roche Diagnostics
Limited, Lewes, UK). Reagents for ALT and AST (with-
out pyridoxal -5 -phosphate ), GLDH, ALP (diethanolamine
buffer), T-bil, creatinine (Jaffe method), and urea were sup-
plied by Roche Diagnostics, while SDH and BA reagents
were supplied by Sigma Aldrich (Poole, Dorset).
Study Data Handling
Data are expressed in terms of fold increase from control
mean values (n 10) at each time point (n 5). Statistical
signiŽcance of changes in any parameter was determined
using Student’s unpaired t-test.
RESULTS
Tissue Distribution
Tissue distribution studies (Figures 1 – 3) showed the
abundance of aGST within the liver and kidney. Measur-
able amounts of aGST were detectable in testes and spleen,
although to a much lesser degree. The tissue distribution of
Vol. 30, No. 3, 2002 aGST IN ASSESSING HEPATOTOXICITY 367

FIGURE 1.—Tissue distribution of aGST in selected tissues of the Wistar Han
rat (n 1). All tissues have gone through a single freeze/thaw cycle at 75 to
85 C.
FIGURE 3.—Comparison of aGST, ALP, ALT, AST, GLDH, and SDH levels
in male Wistar Han rat liver. Homogenate s prepared from both fresh and frozen
whole livers (n 3). aGST (n 1) results reect its concentration per gram of
liver protein (l g/g protein 10 1 ). ALP, ALT, AST, GLDH, and SDH (n 3)
results reect activity per gram of liver protein (U/g protein).

GLDH and SDH is comparable to aGST, with liver and kid-

ney being the primary sources. AST activity, as expected,
was found in a majority of tissues examined, most notably
the heart, liver, kidney, and skeletal muscle.
Enzyme analysis was performed on homogenates made
from fresh and frozen livers for comparison. aGST and ALT
were relatively unaffected by the freeze– thaw process; how-
ever, SDH showed a marked loss of activity. AST and GLDH
were both signiŽcantly increased after the freezing process.
Bromobenzene Study
Clinical Chemistry. No signiŽcant increases were seen in
any parameter at 3 and 7 h after dosing (Figure 4). At 24 h
postdose, elevations in AST, SDH, GLDH, and aGST were
detected. In addition, signiŽcant increases in ALT, BA, and T-
bil were noted at 31 h. Of the selected time points, the highest
measured levels of all parameters were seen at 54 h postdose.
Most notable of these were GLDH and aGST (increased 339-
and 134-fold over control values, respectively). No changes
in plasma creatinine or urea were seen throughout the study.
Histopathology. BrB resulted in multifocal liver necrosis
within 24 to 31 h. By 54 h all animals showed marked or
very marked centrilobular necrosis. In the kidney, slight to
moderate numbers of hyaline droplets were present in the
proximal tubules of the majority of animals killed at 24 to

FIGURE 2.—Distribution of ALP, ALT, AST, GLDH, and SDH in selected tissues of the Wistar Han rat (n 1). All tissues have gone through a single freeze/thaw
cycle at 75 to 85 C. SDH result for liver from fresh liver sample due to loss of activity when frozen.

Downloaded from tpx.sagepub.com by guest on July 15, 2015

368 GIFFEN ET AL
FIGURE 4.—Timed proŽle of plasma markers of hepatotoxicity following a single dose of BrB. Data are presented as fold increase from control mean. Number of
animals used at each time point 5. SigniŽcant differences: p < .05, p < .01,
ALT, AST, GLDH, SDH, BA. At 54 h, T-bil, aGST, ALT, AST, GLDH, SDH, BA.
54 h, but no degenerative tubular changes were seen. No
treatment-related effects were seen in the heart, testes, spleen,
esophagus, stomach, bladder, or cerebral cortex.
a-Naphthylisothiocynat e Study
Clinical Chemistry. No signiŽcant increases were seen in
liver enzymes at 3 and 7 h after dosing (Figure 5). T-Bil was
observed to increase at 3 h and remained elevated throughout
the study. Of the selected time points, the highest measured
levels of all parameters were seen at 31 h postdose. Most
notable of these were BA and T-bil, which were increased
38- and 35-fold over control values, respectively. No changes
in plasma creatinine or urea were seen throughout the study.
Histopathology. Following a single oral dose of ANIT,
progressive biliary and periportal necrosis was observed in
the liver of animals killed at 3, 7, 24, and 31 h postdose.
Evidence of partial recovery of these changes was seen at
168 h postdose. Centrilobular hepatocytes remained unaf-
fected throughout the study. In the testes, treatment-related
changes (spermatid retention, tubular epithelial vacuolation)
were seen in a single male killed at 3 h postdose. No changes
were seen in other males killed at this time point or in males
killed at 7, 24, or 31 h postdose. At 168 h postdose, sperm
retention was apparent in all animals (to be reported sub-
sequently ). In the spleen, a slight to moderate increase in
the degree of extramedullary hematopoiesis was seen in all
animals killed 168 h postdose. No treatment-related effects
were seen in the heart, kidney, esophagus, stomach, bladder,
or cerebral cortex.
Thioacetamide Study
Clinical Chemistry. SigniŽcant elevations in SDH,
GLDH, aGST, and BA were seen at 3 h (Figure 6).
Downloaded from tpx.sagepub.com by guest on July 15, 2015
TOXICOLOGIC PATHOLOGY
p < .001. At 24 h, AST, aGST, SDH, GLDH. At 31 h, T-bil, aGST,
Elevations were seen in all parameters by 7 h, with each re-
maining elevated throughout the study. Of the selected time
points, aGST, SDH, and BA peaked at 24 h with fold in-
creases of 123, 1213, and 58, respectively. AST, ALT, and
GLDH peaked later, at 31 h, with fold increases of 140, 60,
and 69, respectively. All parameters showed partial recovery
by 72 h, however, all remained elevated. Between 7 and 72 h
postdose, plasma creatinine and urea levels increased two-
and threefold over control values, respectively.
Histopathology. Following a single dose of TAM, very
slight periportal inammatory inŽltration of the liver was
present at 3 h (Figure 7). This increased in severity by 7 h,
and was accompanied by biliary hyperplasia, centrilobular
inammation, and multifocal single-cell necrosis. At 24 and
31 h, marked centrilobular necrosis with inammation, had
developed, the severity of which had declined by 72 h. In
the kidney, slight tubular foaminess of the medullary rays
was present at 7 h. By 24 and 31 h this change had pro-
gressed such that vacuolar degeneration of the medullary rays
was seen. This persisted until 72 h, by which time regenera-
tive changes were also present. Within the spleen, very slight
to slight lymphocytolysis was present in most animals. No
treatment-related Žndings were observed in the heart, testes,
esophagus, stomach, bladder, or cerebral cortex.
DISCUSSION
Preliminary studies using the hepatotoxic agents allyl alco-
hol, NDEA, and NDMA (data not shown) were inconclusive
regarding the diagnostic utility of aGST in the assessment
of acute hepatotoxicity. As a result, subsequent studies using
BrB, ANIT, and TAM were performed.
The results of these studies show that aGST is a viable
marker of hepatotoxicity in the Wistar Han rat, correlating
Vol. 30, No. 3, 2002 aGST IN ASSESSING HEPATOTOXICITY 369

FIGURE 5.—Timed proŽle of plasma markers of hepatotoxicity following a single dose of ANIT. Data are presented as fold increase from control mean. Number
of animals used at each time point 5. SigniŽcant differences: p < .05, p < .01, p < .001. At 3 h, T-bil. At 7 h, T-bil. At 24 h ALT, AST, GLDH,
aGST, SDH, BA, T-bil. At 31 h, aGST, ALT, AST, GLDH, aGST, SDH, BA, T-bil. At 168 h, T-bil.

FIGURE 6.—Timed proŽle of plasma markers of hepatotoxicity following a single dose of TAM. Data are presented as fold increase from control mean. Number
of animals used at each time point 5. Note: Actual SDH fold increase stated for each time point. SigniŽcant differences: p < .05, p < .01, p < .001. At
3 h, GLDH, aGST, SDH, AST, BA. At 7 h, ALT, T-bil, ALP, all other parameters. At 24 h, ALP, all other parameters. At 31 h, all parameters.
at 72 h, ALP, ALT, all other parameters.

Downloaded from tpx.sagepub.com by guest on July 15, 2015

370 GIFFEN ET AL TOXICOLOGIC PATHOLOGY

FIGURE 7.—Timed proŽle of liver histopatholog y induced by a single oral dose of TAM (staining with hematoxylin and eosin). Control: Periportal region ( 20).
3 hours: Slight periportal inammatory cell inŽltration ( ) ( 20). 7 hours: Slight centrilobular inammatory cell inŽltration ( ) ( 20). 24 hours: Severe centrilobular
necrosis ( ) contrasted to apparently normal periportal hepatocytes ( ) ( 2). 31 hours: Severe centrilobular necrosis with increased inammatory cell inŽltration
( ) ( 2). 72 hours: Localized, marked centrilobular necrosis ( ) contrasted to normal periportal hepatocyte s ( ) ( 2).

well with changes in other indicators of hepatic injury. aGST
has been reported as a marker of acute hepatotoxicity (2, 6,
12). Nevertheless, given each of the compounds studied, sig-
niŽcant elevations in plasma aGST were seen no earlier than
elevations in any other marker. This is in contrast to studies
performed using carbon tetrachloride (CCl4 ), which show
aGST to be signiŽcantly increased several hours prior to
those seen in plasma AST levels (6). Furthermore, increases
in aGST were lower than with GLDH (BrB and ANIT),
SDH (TAM), or T-bil (ANIT) when compared to control

Downloaded from tpx.sagepub.com by guest on July 15, 2015

Vol. 30, No. 3, 2002 aGST IN ASSESSING HEPATOTOXICITY
levels. However, for each hepatotoxin, aGST was typically
increased more signiŽcantly than either AST or ALT, corre-
lating well with previous reports (6). The timing of enzyme
release into the plasma and their subsequent decrease was
consistent with the histopathological changes seen with each
compound. This time-dependent increase in plasma enzymes,
including aGST, is consistent with previously reported Žnd-
ings in which CCl4 (1, 2, 6, 7) and BrB (2), have been used
as model hepatotoxins.
It has been reported that aGST ( Mr 50,000 ) may increase
within the circulation at an earlier time point than other en-
zyme markers (6) such as AST ( Mr 92,000 ), GLDH ( Mr
300,000 ), SDH ( Mr 140,000 ), and ALT ( Mr 110,000 ), given
its relative smaller size. However, there was no obvious cor-
relation between enzyme molecular weight and magnitude or
time of appearance in the plasma seen in any of the studies
performed.
Tissue enzyme distribution studies show that to accurately
assess hepatic injury, a panel of markers should be used
rather than a single marker in isolation. Although a single
enzyme may exist more abundantly in a particular tissue, it
is typically distributed among varying organs in the body.
Therefore, its elevation within the circulation may reect
damage to this main organ but may additionally reect dam-
age to tissues that express the same enzyme to a lesser de-
gree, as can sometimes be observed with the measurement
of AST. In this investigation the distribution of aGST cor-
related with the respective locations of GLDH, SDH, and
ALT in terms of their relative abundance in the liver. Addi-
tionally, with reference to all other tissues, the comparable
abundance of aGST, GLDH and SDH, within the kidney was
identiŽed.
From the freeze– thaw experiment (Figure 3), it was seen
that both AST and GLDH increased signiŽcantly in ho-
mogenate preparations using previously frozen liver sam-
ples. The loss of SDH activity postfreezing was also an
important Žnding, illustrating its greater utility if measured
on fresh sample material. Levels of aGST and ALT remained
relatively unaffected in the homogenate prepared from pre-
viously frozen liver tissue.
A single oral dose of BrB induced a marked degree of
centrilobular necrosis. aGST, like other plasma markers was
not increased before 24 h after dosing (Figure 4). Although at
54 h aGST was increased more signiŽcantly than both ALT
and AST, it was elevated to a lesser extent than GLDH. The
more pronounced increases in aGST in comparison to ALT
and AST at 24, 31, and 54 h may reect the hepatic lobular
localization of each enzyme, with aGST being recognized as
more ubiquitous throughout the liver than the transaminases
and their reported periportal prevalence (6). Although slight
changes in renal tubules were seen in the majority of animals
at 24 and 54 h, there was no degeneration of these cells, and
hence release of intracellular contents from these cells into
the blood is unlikely.
Administration of ANIT induced periportal necrosis
and inammatory changes consistent with that expected
(10). This compound was recognized to cause intrahepatic
cholestasis within 16 to 24 h of administration, and therefore
its use to induce acute hepatobiliary toxicity and furthermore
to assess the utility of aGST in the detection of this lesion was
not ideal. However, ANIT did induce changes within the por-
Downloaded from tpx.sagepub.com by guest on July 15, 2015
371
tal region and, as such, allowed direct comparison of aGST
with ALT and AST in the assessment of this type of injury.
While aGST was elevated to a greater extent than ALT and
AST at 24 and 31 h after dosing, it did not increase or de-
crease more rapidly than any other parameter. In comparison
to aGST, GLDH was the most signiŽcantly increased liver
enzyme at 24 and 31 h postdose. However, the hepatobiliary
lesion induced by ANIT administration was best indicated
by elevations in total plasma BA and most notably T-bil. An
unexpected testicular pathology was noted in a single animal
at 3 h, although no concomitant elevations in aGST, ALP, or
AST were detected.
TAM induced a biphasic pattern of hepatotoxicity,
with periportal inammation (3 – 7 h) preceding centrilob-
ular necrosis and apparent periportal recovery (24 – 72 h)
(Figure 7). In contrast with BrB and ANIT treatment, ele-
vations in plasma enzymes were detected at 3 h following
TAM administration. As well as SDH, GLDH, and BA, this
included aGST, which was elevated to a greater degree than
both ALT and AST at this time point. By 7 h each parame-
ter was signiŽcantly increased with respect to control values,
most notably SDH, which remained the most sensitive param-
eter throughout the study. With TAM-induced hepatic insult,
concordant elevations in plasma enzymes were not observed,
as seen with both BrB and ANIT treatment. aGST reached
its highest concentration in the plasma at 24 h, decreasing
by 72 h. GLDH, ALT, and AST continued to increase un-
til 31 h before decreasing. aGST, SDH, GLDH, ALT, and
AST still remained elevated at 72 h postdose. Nephrotoxi-
city was apparent histopathologically between 24 and 72 h,
supported by increases in plasma creatinine of 33.9 – 148%
and urea of 66.9 – 183% on these occasions. This severity of
renal toxicity may therefore be contributory to the apparent
increases in plasma aGST, ALP, AST, GLDH, and SDH. It
is typically recognized that markers of renal injury are ex-
creted into the urine; however, in the presence of signiŽcant
renal damage, their reabsorption and thus appearance in the
circulation should not be discounted.
The rat speciŽc aGST enzyme immunoassay (EIA) is sim-
ple to use and performs well. Intra- and interassay preci-
sion, sensitivity, and linearity were monitored throughout
each study, and the data produced were consistent with those
quoted by the manufacturer (data not shown). However, the
assay, including sample preparation, takes approximately 4 h.
The cost per test of aGST ($18.20 ) is also considerably higher
than other markers such as ALT ($0.08) and GLDH ($1.20 ).
As such, it is a labor- and cost-intensive assay in comparison
to the automated spectrophotometric assays discussed.
At no time point in any of the three studies did aGST offer
any additional information regarding hepatocellular integrity
compared to alternate parameters measured. aGST was gen-
erally shown to be a more sensitive marker of hepatic injury
than ALT and AST, corresponding to previously published
reports (6), while other markers such as GLDH, SDH, or T-bil
were shown to more effectively indicate the differing types
of damage induced.
In conclusion, aGST in the Wistar Han rat was shown to
be a valid marker of these types of induced hepatotoxicity.
However, the measurement of aGST offered no additional
information in detecting either the time of onset/recovery
or the severity of each type of hepatic injury induced, in
372 GIFFEN ET AL
comparison to the panel of markers already established within
this laboratory.
ACKNOWLEDGMENTS
The authors thank GlaxoSmithKline for the sponsorship
of these studies and are grateful for the cooperation and tech-
nical assistance supplied by the department of Preclinical
Safety Sciences.
REFERENCES
1. Adachi Y, Horii K, Suwa M, Tanihata M, Ohba Y, Yamamoto T (1981).
Serum glutathione S-transferase in experimental liver damage in rats. Gastro
Jpn 16: 129 – 133.
2. Aniya Y, Anders MW (1985). Alteration of hepatic glutathione
S-transferases and release into serum after treatment with bromobenzene ,
carbon tetrachloride, or N-nitrosodimethylamine. Biochem Pharmacol 34:
4239 – 4244.
3. Beckett GJ, Hayes JD (1987). Plasma glutathione S-transferase measure-
ments and liver disease in man. J Clin Biochem Nutr 2: 1– 24.
4. Berndt WO (1987). Renal toxicology. In: Handboo k of Toxicology. Haley
TJ, Berndt WO (eds). Hemisphere, London, pp 157 – 191.
5. Boyd JW (1962). The comparative activity of some enzymes in sheep, cattle
and rats—Normal serum and tissue levels and changes during experimental
liver necrosis. Res Vet Sci 3: 419 – 433.
Downloaded from tpx.sagepub.com by guest on July 15, 2015
TOXICOLOGIC PATHOLOGY
6. Clarke H, Egan DA, Heffernan M, Doyle S, Byrne C, Kilty C, Ryan MP
(1997). a-glutathione S-transferase (a-GST) release, an early indicator of
carbon tetrachloride hepatotoxicity in the rat. Hum Exp Toxicol 16: 154 – 157.
7. Igarashi T, Muramatsu H, Ohmori S, Ueno K, Kitagawa H, Satoh T
(1988). Plasma glutathione S-transferase in carbon tetrachloride treated
rats in association to hepatic cytosolic isozymes. Jpn. J Pharmacol 46:
211 – 216.
8. Kilty C, Doyle S, Hassett B, Manning F (1998). Glutathione S-transferases
as biomarkers of organ damage: Applications of rodent and canine GST
enzyme immunoassays . Chem Biol Interact 111 – 112: 123 – 135.
9. Mehendale HM (1987). Hepatotoxicity. In: Handbook of Toxicology. Haley
TJ, Berndt WO (eds). Hemisphere, London, pp 74 – 111.
10. Neghab M, Stacey NH (1996). Alpha-naphthylisothiocyanate-induce d ele-
vation of serum bile acids: Lack of causative effect on bile acid transport.
Chem Biol Interact 99: 179 – 192.
11. Orsler DJ, Ahmed-Choudhurr y J, Chipman K, Hammond T, Coleman R
(1999). ANIT-disruption of biliary function in rat hepatocyte couplets. Tox-
icol Sci 47: 203 – 210.
12. Rees GW, Trull AK, Doyle S (1995). Evaluation of an enzyme-
immunometric assay for serum a-glutathione S-transferase. Ann Clin
Biochem 32: 575 – 583.
13. Sun F, Shoko H, Yukako O, Sakiko H, Kyoko T, Yasuko Y, Sadako T,
Shosuke K (2000). Evaluation of oxidative stress based on lipid hydroper-
oxide, vitamin C and vitamin E during apoptosi s and necrosis caused by
thioacetamide in rat liver. Biochim Biophys Acta 1500: 181 – 185.