Endotoxin Induces Structure-Function Alterations of Rat Liver
Peroxisomes: Kupffer Cells Released Factors as Possible Modulators
MIGUEL A. CONTRERAS,1MUSHFIQUDDIN KHAN,1BRIAN T. SMITH,1ANNA M. CIMINI,1ANNE G. GILG,1JOHN ORAK,1
INDERJIT SINGH,1AND AVTAR K. SINGH2
We report that endotoxin treatment results in decreased
amounts of peroxisomes as well as changes instructureand
function of peroxisomal membranes. Peroxisomes isolated
from the liver of control and treated animals showed a
marked decrease in total protein, but no significant alter-
ation in the sodium dodecyl sulfate polyacrylamide gel
electrophoresis (SDS-PAGE) protein profile. However, the
Westernblot study oftheperoxisomal ?-oxidationenzymes
and catalase showed an increase in those enzymes in the
peroxisomal peak of normal density in endotoxin-treated
rats. Disintegration of peroxisomal membranes by carbon-
ate treatment from endotoxin-treated liver and change in
thefluidity of peroxisomal membranes suggests alterations
in peroxisomal membrane structure. No such alterations
were found in mitochondrial or microsomal membranes of
endotoxin-treated livers. The lipid analysis of these organ-
elles showed that the only organelle affected was the
peroxisome, withasignificant decreaseinthephospholipid
and cholesterol concentrations. To understand the mecha-
nismof endotoxin-mediated alterations in peroxisomes, we
studied the possible role of Kupffer cell secreted soluble
factors (tumor necrosis factor ? [TNF-?]) on the peroxi-
cells by gadolinium chloride before endotoxin treatment
did not normalize the overall peroxisomal protein amount
and the lipid composition of isolated peroxisomes. How-
ever, the levels of individual protein amount in remaining
peroxisomes were normalized. Endotoxin also decreased
peroxisomal ?-oxidation, and this was partially restored
with gadoliniumtreatment. Theseresults clearly show that
peroxisomes are severely affected by endotoxin treatment
and suggest that the damage to this organelle may contrib-
ute, at least inpart, toendotoxin-inducedhepatic cytotoxic-
ity. (HEPATOLOGY 2000;31:446-455.)
Endotoxemiaisan important clinical problem. Endotoxin,
a lipopolysaccharide (LPS) present in the cell wall of all
Gram-negative bacteria, consists of a lipid moiety, lipid A,
covalently linkedtoapolysaccharide.1Almost all thebiologi-
cal activities of LPS are elicited by lipid A. This compound
deregulates the production and release of cytokines; mostly
tumor necrosisfactor ? (TNF-?) and interleukins(IL)-1 and
6 and interferon-?.2The possible role of TNF-? as mediator
in the LPS shock has been shown by the detection of
increased circulating levels of TNF-?, following administra-
tion of LPS to animals,3the triggering of shock and tissue
injury in ratsinjected with recombinant human TNF-?,4and
by attenuation of lethal effects of LPS by passive immuniza-
tion with antibodiesagainst TNF-?.5,6
The liver is the main organ responsive to LPS during
bacterial sepsis7and the monocyte/macrophage lineage cells
aretheprincipal sourceof TNF-? in LPS-treated animals.8-10
LPS can serve as an effective signal for hepatic macrophages
(fixed Kupffer cells [KC]) that may in turn respond by
generating thecytokines (e.g., IL-1 and TNF-?) necessary to
activate hepatocytes.11The hepatocytes in turn respond to
IL-1 and TNF-? by generating IL-8, a potent chemotactic
factor for therecruitment of neutrophils into theliver.11The
activation of monocytes/macrophages leads to the chain of
eventsimplicated in theinflammatory reaction. Theseevents
are associated with the production of cytokines,2bioactive
lipids,12and oxidants including both reactiveoxygen species
(superoxide, hydrogen peroxide, and hydroxyl radical13) as
well as reactive nitrogen intermediates, in particular nitric
oxide.14Ultimately, the systemic inflammatory response can
lead to the characteristic pattern of hemodynamic and
metabolic derangements, which result in the septic shock
syndrome, multiple organ failure, and death. The possible
ment in liver is supported by theobserved protection in rats
treatedwithgadoliniumchloride(GdCl3) todepleteKC from
Endotoxemiais accompanied by marked cellular derange-
proteins.19Furthermore, serum levels of triglycerides and
cholesterol are affected by LPS and cytokines.18TNF-? also
down-regulatestheactivity ofacyl-CoA synthase, theenzyme
responsible for the first reaction in the utilization of fatty
acids for their degradation and their utilization for complex
Abbreviations: LPS, lipopolysaccharide; TNF-?, tumor necrosisfactor ?; IL, interleu-
kin; KC, Kupffer cells; GdCl3,gadoliniumchloride; BSA, bovineserumalbumin; NaCl,
sodium chloride; ER, endoplasmic reticulum; SDS-PAGE, sodium dodecyl sulfate-
polyacrylamidegel electrophoresis; ANSA, 8-anilino-1-naphthalenesulfonic acid; DAB,
3,3? diaminobenzidine; TNT, Trissodiumbuffer with 0.05%Tween.
From the1Department of Pediatrics, Medical University of South Carolina, Charles-
ton, SC;and2RalphH.JohnsonVeteran’sAdministrationMedical Center, Departmentof
Laboratory Medicine, Charleston, SC.
ReceivedApril 26, 1999; acceptedNovember 30, 1999.
Supported in part by grants from Dialysis Clinics Inc. and National Institutes of
Health (NS-22576, NS-34741, andNS-37766).
M.A.C.’s current address is Section in Brain Physiology and Metabolism, NIA, National
A.M.C.’scurrent addressisDipartimentodi Biologiadi Baseed Applicata, Universita’
degli Studi di L’Aquila, ViaVetoio, Loc. Coppito, 67010 Coppito(AQ), Italy.
Address reprint requests to: Avtar K. Singh, M.D., Ralph H. Johnson Veteran’s
Administration Medical Center, Department of Laboratory Medicine, 109 Bee Street,
Charleston, SC 29401. E-mail: email@example.com; fax: 843-792-7130.
Copyright?2000 by theAmerican Association for theStudy of Liver Diseases.
lipid biosynthesis in adipocytes.20TNF-? and LPS also
modulate the activities of antioxidant enzymes and the
enzyme system for fatty acid oxidation in peroxisomes in
liver.21-24This down-regulation of peroxisomal ?-oxidation
may contribute to the increased pool of fatty acids for the
synthesis of triglycerides. The increased levels of cholesterol
aretheresult of theobserved increasein levels of HMG-CoA
The role of peroxisomes in cellular metabolism and in
particular in the metabolism of fatty acids and lipids is well
established.26Therefore, in the present study we examined
the effect of LPS on the structure/function of peroxisomes.
The studies reported in this article show that administration
of a sublethal dose of LPS alters the amount of peroxisomal
proteins and the structure/function of peroxisomes. These
alterations, atleastinpart, aremediatedby effectormolecules
MATERIALS AND METHODS
Reagents. All reagents wereof analytical gradeor thepurest form
available. Bovine serum albumin (BSA), imidazole, dithiothreitol,
antipain, leupeptin, pepstatin, N-(?-L-Rhamnopyranosyl-oxyhy-
phenylmethylsulfonyl fluoride, LPS (from Salmonella typhimurium)
and ammonium persulfate were purchased from Sigma Chemical
Co. (St. Louis, MO). GdCl3hexahydratewasfromAldrichChemical
Company, Inc. (Milwaukee, WI). 5-(N-2,3-Dihydroxypropyl-
amide (Nycodenz) was obtained from Gibco BRL (Grand Island,
NY). [125I] radiolabeled anti-rabbit IgG antisera were from ICN
Pharmaceuticals, Inc. (Irvine, CA). Molecular weight wide-range
protein standard (Mark 12) was from Novex (San Diego, CA).
Methylated molecular weight marker was from Amersham (Arling-
ton Heights, IL). Antibodies against peroxisomal proteins (Acyl-
CoA oxidase, multifunctional enzymeandthiolase) wereagenerous
gift of Professor Takashi Hashimoto of Sinshu University, Japan.
Treatment of Rats With LPS and/or GdCl3. Theuseof animalswasin
accordance with the ‘‘Guide for the Care and Use of Laboratory
Animals’’ (National Institutes of Health, Pub. No. 86-23). Rats
(Sprague-Dawley) weighing approximately 200 to 250 g were
injected intraperitoneally with 1 mg/kg of body weight of LPS
dissolvedinsodiumchloride(NaCl), 0.9%(wt/vol).Thecontrol rats
wereinjected with NaCl, 0.9% (wt/vol) and killed after 18 hours. A
second group of ratswereinjected twice(24-hour interval) intrave-
nously with 10 mg/kg body of weight of GdCl327before an
intraperitoneal injection of 1 mg/kgof body weight of LPSdissolved
in NaCl, 0.9% (wt/vol). The control rats were preinjected twice
(24-hour interval) with GdCl3beforean intraperitoneal injection of
NaCl, 0.9% (wt/vol). After 18 hours postinjection (42 hours after
first GdCl3injection), theratswerekilled.
Preparation of Rat Liver Subcellular Organelles by Differential and
tal rats(7 g each) werehomogenizedin 10 volumesof homogeniza-
tion buffer (sucrose, 0.25 mol/L; ethylenediaminetetraacetic acid, 1
mmol/L; antipain, 1 mg/mL; aprotinin, 2 mg/mL; leupeptin, 2
mg/mL; pepstatin A, 0.7 mg/mL; phosphoramidon, 0.1 mmol/L;
phenylmethylsulfonyl fluoride, 0.2mmol/L; ethanol, 0.1%(vol/vol);
imidazole, 3 mmol/L; pH 7.4) at 4°C. The homogenates were first
fractionated by differential centrifugation to prepare the lambda
fraction (light mitochondrial fraction enriched with peroxisomes
and lysosomes28). The lambda fraction was further subjected to
isopycnic equilibrium centrifugation in a continuous (0-50% [wt/
vol]) Nycodenz gradient overlying a cushion of 55% (wt/vol) of
Nycodenz asdescribed previously.28Thegradientswerecollected in
1.5-mL fractions from the bottom of the tubes. Each fraction was
analyzedfor marker enzymeactivities.
Assay for Marker Enzymes. The location of the subcellular organ-
ellesin thegradient wasdeterminedusingspecific marker enzymes:
catalase for peroxisomes, cytochrome c oxidase for mitochondria,
and NADPH cytochrome c reductase for endoplasmic reticulum
(ER, microsomes), asdescribedelsewhere.28Theprotein concentra-
tions were determined by the procedure of Bradford,29using
Protein Analysis by SodiumDodecyl Sulfate-Polyacrylamide Gel Electro-
phoresisandWesternBlot Analysis. Theproteinanalysiswasperformed
by sodiumdodecyl sulfate-polyacrylamidegel electrophoresis(SDS-
PAGE) (7.5%) as described previously.30The electrophoresis was
carriedat constant voltage(36 V) for 4 to5 hours. For Western blot
analysis, proteins resolved by SDS-PAGE were transferred from the
gels to polyvinylidene fluoride (PVDF) nylon membranes (Milli-
pore) at 230 mA (constant current) for 2 hours. Thetransfer buffer
contained 0.02% of SDS. Immunoblot analysis was performed as
described previously.30Briefly, the nylon membranes were blocked
with BSA by incubation with asolution containing 3%(wt/vol) BSA
in phosphate-buffered saline(KH2PO4/Na2HPO4, 10 mmol/L; NaCl,
150mmol/L;pH 7.4) inawaterbathat40°C for1hour.Theprimary
BSA (3% [wt/vol]), and normal goat serum (10% [vol/vol]) were
incubated with BSA-blocked membranes for 90 minutes. After
extensive washes in solution containing 1% (wt/vol) BSA; 0.05%
(vol/vol) Tween 20 in phosphatebuffer saline, themembranes were
treatedwithsecondary antibody ([125I] labeledIgG [2µCi/membrane])
for 90 minutes. The membrane was washed, dried, and exposed to
radiograph films(X-OMAT; Kodak, Rochester, MN) at ?70°C.
Analysis of Lipids by High Performance Thin-Layer Chromatography.
Total lipids were extracted from peroxisomes isolated from control
and experimental livers according to theprocedureof Folch et al.31
Lipids were separated into polar and neutral lipid classes by high
performance thin-layer chromatography as described by Ganser et
by densitometric scanning using Imaging Densitometer (model
GS-670; Bio-Rad). Total phospholipidswerequantitatedby colorim-
Preparation of Integral Membranes. Integral membranes of different
organelles (peroxisomes, mitochondria, and microsomes) were
prepared by diluting the subcellular organelle fractions with a
ice-cold solution of sodiumcarbonate, 0.1 mol/L; iodoacetamide30
mmol/L; pH 11.5 as described previously.34After 30 minutes of
incubation at 4°C, the membranes were sedimented by centrifuga-
tion at 100,000g for 1 hour. The sedimented membranes were
washedoncewith coldwater, lyophilizedandstoredat ?70°C.
Fluidity Measurements. The membrane fluidity studies of peroxi-
somes, mitochondria, andmicrosomespurifiedfromliver of control
and treated animals were performed by using 8-anilino-1-naphtha-
lenesulfonic acid(ANSA).35Purifiedperoxisomes, mitochondria, or
microsomes (0.75 mg protein) in 1 mL of buffer containing
Tris-HCl, 20 mmol/L; sucrose, 250 mmol/L (pH 7.4); were mixed
with 2 mL of an ANSA 22.5 mmol/L, solution containing Tris-HCl,
20 mmol/L; sucrose250 mmol/L pH 7.4; at 4°C for 5 minutes. The
mixturewas then diluted to 25 mL and centrifuged at 200,000g for
15 minutes. The pellet was washed once with buffer (Tris-HCl, 20
mmol/L; sucrose, 250 mmol/L) pH 7.4 and then suspended in 1 mL
of the buffer for fluidity measurements. The P values of ANSA-
labeled subcellular organelles were measured with a SIM-Aminco
spectrofluorometer, T-format (SIM 8000 C) (SLM Instruments Inc,
Rochester, NY). The excitation was at 380 nm with a polarized
beam, and the emitted light was detected simultaneously in two
independent cross-polarized channels equipped with cut-off filter
for wavelengthsbelow 418 nm. Thesampleswereread 3 timeseach
during15 minutes, at 30°C.
Morphological Examinationof Peroxisomes. Isolatedcatalasecontain-
ing fractions (400 µg) from control and endotoxin gradients were
pelleted by centrifugation and the pellets were fixed in 2.5%
glutaraldehyde in 0.1 mol/L cacodylate buffer, pH 7.4, for 3 hours.
HEPATOLOGY Vol. 31, No. 2, 2000CONTRERAS ET AL.
The fixed peroxisomes were washed with cacodylate buffer, pH 7.4,
containing 0.25 mol/L sucrose and then incubated in alkaline 3,3?
diaminobenzidine (DAB) solution for the cytochemical localization of
peroxisomal catalase.36Samples were then infiltrated with Embed 812
and polymerized at 60°C for 36 hours. Thin sections were counter-
stained with uranyl acetate and lead citrate, and the ultrastructural
examinationwasperformedunderaHitachi 12-A electronmicroscope.
Immunohisochemistry. Formaldehyde-fixed tissues of 3 endotoxin-
treated, 3endotoxinplusGdCl3–treated, and3normal controlswere
paraffin-embeddedandsectionedat 6 microns. Tissuesectionswere
deparaffinized and sequentially rehydrated in graded alcohols and
immersed in Tris sodium buffer for 15 minutes. Slides were then
microwaved for 10 minutesin antigen unmasking fluid fromVector
Laboratories (Burlingame, CA). After allowing slides to cool, they
werewashed3timesfor 5minutesinTrissodiumbuffer with0.05%
Tween(TNT). Slideswereincubatedfor 10minutesin3%hydrogen
peroxide in methanol to eliminate endogenous peroxidase activity.
Then sections were blocked in Tris sodium buffer with 0.5%
blockingreagent (suppliedwithRenaissanceTSA-Indirect Kit; NEN
Life Sciences) for 30 minutes to reduce nonspecific staining.
Sectionswereincubatedovernight with apolyclonal anti-rat TNF-?
antibody (1:100) obtained from R&D Systems, Inc. (Minneapolis,
MN). In addition, control rabbit IgG (Vector) was incubated on
separate slides. Slides were rinsed 3 times for 5 minutes in TNT.
Sections were incubated with biotinylated anti-rabbit IgG antibody
followed by an avidin-biotin horseradish peroxidasecomplex (Vec-
tastain ABC kit; Vector) with biotinylated tyramide (Renaissance TSA-
Indirect Kit; NEN LifeSciences) as substrate. After thorough rinsing in
TNT slides were incubated with streptavidin-conjugated Texas Red
(NEN LifeSciences). Sectionswererinsed in TNT followed by distilled
Image Analysis. The sections were examined at ?400 magnifica-
tion under aZeiss fluorescent microscope(Thornwood, NY) with a
camera attachment using a dual band pass filter, and images were
captured under red and green filters and processed using Adobe
Photoshop4.0. (AdobeSystemInc, San Jose, CA).
Peroxisomal ?-Oxidation. Peroxisomal ?-oxidation was measured
in rat liver homogenatesusing [1-14C]–lignoceric acid essentially as
describedin Dhaunsi et al.23
Statistical Analysis. All data represent the mean ? SD for each
number of experiments. Statistical analysis was evaluated by two-
tailed t test for unpaired observations. Differences at P ? .05 were
taken asstatistically significant.
Effect of LPS on the Distribution of Peroxisomes in the Nycodenz
Gradient. The effects of a sublethal dose (1 mg/kg body
weight) of LPS on thestructure/function of peroxisomes was
examined in rat liver peroxisomes18 hoursafter intraperito-
neal injection of LPS. Control animalswereinjected with the
Nycodenz gradient from control and LPS-treated rat livers.
with marker enzyme activities for peroxisomes (catalase),
mitochondria (cytochrome c oxidase), and ER (NADPH
cytochrome c reductase). Catalase, the peroxisomal marker,
showed a bimodal distribution in the control gradient; one
peak at high density (1.21 g/cm3) in fractions 4 to 7
corresponding to mature peroxisomes and a second peak at
lighter density at the top of the gradient (fractions), corre-
spondingtofree(soluble) catalase.Incontrast, athirdpeak of
g/cm3) in the gradient from LPS-treated animals. This peak
was localized between the mitochondrial and ER markers in
thegradient (Fig. 1) indicating that LPS treatment appearsto
affect the amount and distribution of catalase in rat liver.
There was a shift in the marker for mitochondria towards a
higher density, in the gradient from liver of LPS-treated
animals, compared with control animals. The marker for ER
in agradient fromLPS-treatedliver presentedamorefocused
diffuse distribution in the control gradient. The shift in
mitochondrial density in LPS-treated liver may represent the
observedmorphological alterationsproducedby TNF-?.37
Effect of LPS Treatment on the Recovery of Peroxisomes. The
protein distribution showed variation in the protein content
mainly in fractions containing peroxisomes equilibrated at
normal density (1.21 g/cm3) from LPS-treated animals (Fig.
1). Table 1, shows that the protein content in the pooled
fractions of peroxisomes decreased by 56% in the gradient
from LPS-treated rats. However, the specific activity of
lular organelles from control and
subcellular organelles were isolated
fromcontrol (black bars) and endo-
toxin-treated (white bars) animals
by Nycodenz density gradient cen-
trifugation. The distribution of the
lase for peroxisomes, (B) cyto-
chromec oxidasefor mitochondria,
and (C) NADPH cytochrome c re-
tion in thegradients was also deter-
mined (D). The data represent the
mean ? SD of the values obtained
forn? 3to9experimental animals.
The thick line on (A) indicates the
localization of the third peak of
Fractionation of subcel-
CONTRERAS ET AL.HEPATOLOGY February 2000
catalase in the normal density peroxisomal fractions (1.21
g/cm3), as well as in fractions containing the new catalase
peak (fractions 11-13, d ? 1.14 g/cm3), was approximately
2-fold higher compared with control gradients. Theseresults
The selective loss of peroxisomes in LPS-treated liver was
studied by morphological examination by electron micros-
copy of catalasecontaining fractions after DABstaining (Fig.
2). A significant number of peroxisomes isolated from
untreated animals (d ? 1.14 g/cm3) wereintact with asingle
limiting membrane surrounding a homogenous granular
matrix (Fig. 2A). The peroxisomes from the corresponding
fractions (d ? 1.14 g/cm3) isolated from endotoxin-treated
animals were identified as catalase-positive organelles in
lysosomes (Fig. 2B through D). These results indicate that
autophagy of peroxisomes may be responsible for the ob-
served decrease in peroxisomes of normal density in LPS-
treatedliver (Fig. 2).
The effect of LPS treatment on the peroxisomal proteins
was studied by Western blot analysis. Proteins from the
peroxisomal fractions from control and LPS-treated liver
gradients wereanalyzed with antibodies against catalaseand
peroxisomal ?-oxidation enzymes (multifunctional enzyme,
and 3-ketoacyl-CoA thiolase). Western blot of proteins of
pooled peroxisomes of normal density (fractions 4-7) is
shown in Fig. 3. The increased signal for peroxisomal
enzymes found only in the fractions from LPS-treated liver
gradients indicate that peroxisomes from LPS-treated livers
contain a relatively higher amount of catalase and fatty acid
?-oxidation enzyme proteins. This increase in catalase pro-
tein is consistent with the observed high specific activity of
catalase in peroxisomes from LPS-treated liver. As shown by
the electrophoretic protein pattern (Fig. 4A), no appreciable
differences in the overall protein patterns of peroxisomes
were found between the control gradient and the gradient
from LPS-treated animals. These results indicate that the
decrease in peroxisomes from LPS-treated livers was not
TABLE 1. Protein Concentration in Subcellular Organelles Peak From
Liver of Control and Endotoxin-Treated Rats
2.23 ? 0.30
14.05 ? 2.43
20.70 ? 4.04
0.98 ? 0.20*
9.70 ? 1.44†
23.22 ? 4.19
NOTE. Seven grams wet weight of liver from control and endotoxin-
treated rats were used for these experiments. Subcellular organelles were
isolated from Nycodens density gradient as described in the Materials and
Methods and were identified using the enzymatic markers: catalase for
peroxisomes, cytochrome c oxidase for mitochondria, and NADPH cyto-
chromec reductasefor microsomes.
*P? .001 (n ? 9), for comparison between peroxisomesfromendotoxin-
†P?.001(n ? 9), forcomparisonbetweenmitochondriafromendotoxin-
control and endotoxin gradient. Fraction 11 from control gradient (A) and
the fraction containing additional peak of catalase from the endotoxin
gradient (B-D) stained with DAB for cytochemical localization of peroxi-
somal catalase.Protein(400µg) offraction11(secondpeak ofcatalase) from
control and endotoxin-treated animals were fixed and stained with DAB,
uranyl acetate, and lead cytrate, as described in the Materials and Methods
(original magnification?91,000). P , peroxisomes.
Electron micrographs of catalase stained peroxisomes from
endotoxin-treatedanimals. Eachsamplecontained50µgofprotein. Western
multifunctional enzyme, (b) catalase, and(c) 3-ketoacyl-CoA thiolase. Lane
1 is molecular weight markers-[14C] methylated proteins. Lanes 2 and 3
represent peroxisomal proteins from control and LPS-treated animals,
Western blot analysis of peroxisomal fractions from control and
HEPATOLOGY Vol. 31, No. 2, 2000CONTRERAS ET AL.
caused by enhanced proteolysis of individual proteins. It is
consistent with the observations in Fig. 2 showing turnover
ofperoxisomesby autophagy in LPS-treatedliver.
To analyze the effect of LPS treatment on membrane
elleswereprepared by carbonatetreatment.34Thisprocedure
released associated proteins from the membranes and pro-
duced unsealed sheet fragments of trilaminar appearance
retaining only integral membraneproteins.34Electrophoretic
analysis of the integral membrane proteins from subcellular
organelles is shown in Fig. 4B. Therewerevery few proteins
present in themembranessedimented fromLPS peroxisomal
fractionscomparedwith controls(Fig. 4B). In contrast, there
were no such visible alterations in the protein pattern of
membranes prepared from other subcellular organelles (e.g.,
mitochondria and ER). The significant alteration in peroxi-
somal membraneprotein pattern (Fig. 4B), ascomparedwith
no change in the protein pattern of intact peroxisomes (Fig.
4A), indicate that peroxisomal membranes from LPS-treated
livers may be unstable and that they disintegrate during this
experimental protocol of carbonate treatment.34Inability to
prepare membrane sheets that sediment by centrifugation at
100,000g from peroxisomes from LPS-treated livers after
carbonate treatment suggests that LPS induces alterations in
membrane structure and that these alterations might affect
LPS Treatment Alters the Peroxisomal Membrane Fluidity. The
possibility of alterations of theperoxisomal membraneprop-
erties by LPS was further investigated by measuring the
membrane fluidity using fluorescence depolarization. Fluores-
cence depolarization (P values) of ANSA-labeled peroxi-
somes, mitochondria, and ER was measured as described in
the Materials and Methods (Table 2). P values obtained for
peroxisomal membranes from LPS-treated animals were sig-
nificantly lower than values obtained for control animals
(P ? .05). This indicates that peroxisomal membranes from
LPS-treated animalshaveincreased fluidity. Nosuch changes
were observed in the fluidity of mitochondria or ER from
LPS-treated liver (Table 2). These results indicate that LPS
may induce major changes in the membrane lipid composi-
tion of peroxisomes. The analysis of lipid composition of
isolated subcellular fractions showed a marked decrease in
total phospholipids and cholesterol in peroxisomes (50.8%;
P ? .0001 and 28.7%; P ? .004, respectively) but not in
mitochondriaor ER (Fig. 5). Theseobservedchangesin lipid
protein and (B) integral membrane
proteins in gradient fractions from
control and endotoxin-treated rats.
each gradient fraction was resolved
CoomasieBlue. (B) Membranespre-
resolvedby 7.5%SDS-PAGE. (A and
B) Lanes1, 10, and23aremolecular
weight marker. Lanes 2, 4, 6, 8 and
3, 5, 7, 9 areproteinscorresponding
to mature peroxisome containing
fractions (d ? 1.21 g/cm3) from
control andendotoxingradients, re-
spectively. Lanes 13, 15, 17, 19, 21,
24, 26 and14, 16, 18, 20, 22, 25, 27
are proteins corresponding to mito-
chondria and ER containing frac-
tions from control and endotoxin
Distribution of (A) total
TABLE 2. MembraneFluidity of Subcellular Organelles FromControl and
Endotoxin Rat Livers Assayed by FluorescenceDepolarization
0.284 ? 0.006
0.275 ? 0.008
0.280 ? 0.010
0.254 ? 0.016*
0.277 ? 0.001
0.275 ? 0.009
NOTE. Peroxisomes were prepared from 7 g of control and endotoxin-
identifiedby respectivemarker enzymes.
*P ? .05 (n ? 4), for comparison between peroxisomes from endotoxin-
CONTRERAS ET AL.HEPATOLOGY February 2000
composition are most likely responsible for the variation in
In summary, the results from these studies show that
treatment of rats with a sublethal dose of LPS decreases the
number of peroxisomes of normal density. Furthermore, the
peroxisomes of normal density present in those livers have
alterations in the lipid composition that may be responsible
for theobservedchangesin themembranefluidity andhence
Effect of GdCl3on LPS-Induced Alterations in Peroxisomes. The
observed alterations in liver peroxisomes of LPS-treated
animalscouldbemediatedby solublefactors(e.g., cytokines)
secreted by LPS-activated KC.8,9To test this hypothesis, we
liver ofKC,38prior toLPStreatment.
the lambda fraction from GdCl3- and GdCl3/LPS-treated
liversin aNycodensdensity gradient. Asreportedpreviously,
the subcellular organelles in the gradient. The GdCl3treat-
ment did not normalize the LPS-induced decrease in total
peroxisomal protein present in theperoxisomal peak fraction
of normal density (d ? 1.21 g/cm3) (Fig. 6, Table 3) under
these experimental conditions. Although, there was a ten-
dency towards normalization of thedistribution of subcellu-
lar organelles, especially a decrease in the catalase peak of
1.14 g/cm3density, in the gradients from rat livers treated
The amount of peroxisomal ?-oxidation enzymes (multi-
functional enzymes and thiolase) and catalase of normal
peroxisomal peak fractions shows that the levels of these
enzymeswerenormalized in theliver of animalstreated with
GdCl3before treatment with LPS (Fig. 7). These results
contrast significantly with theresultsobtainedin theabsence
of treatment with GdCl3, in which we found a marked
induction of these enzymes (Fig. 3). However, the peroxi-
somes from livers of animals treated with GdCl3before LPS
in peroxisomes from livers of only LPS-treated animals (Fig.
8). These results suggest that the expression of peroxisomal
?-oxidation enzymes may be regulated by the mediators
secreted by LPS-activated KC, whereas mediators may not
have any role in the induction of alterations in the lipid
composition and the reduction in peroxisomes of normal
density in LPS-treatedlivers.
Endotoxin Increases TNF-? Production in KC and Decreases
Peroxisomal ?-Oxidation. To confirm that LPS was increasing
try on rat livers from animals treated with LPS and animals
pretreated with GdCl3followed by endotoxin. The results,
showninFig. 9, clearly showthatLPSincreasesthesignal for
TNF-? (red fluorescence) whereas pretreatment with GdCl3
substantially decreases the TNF-? signal. In addition, the
effects of endotoxin and GdCl3pretreatment on peroxisomal
?-oxidation wasdetermined. Theobserveddecreasein ligno-
ceric acid oxidation (Fig. 10A) in endotoxin-treated animals
and theapparent normalization with GdCl3treatment corre-
lates well with the TNF-? production. Endotoxin also de-
creased the amount of peroxisomal multifunctional enzyme
as judged by Western blot analysis of total cell homogenates
(Fig. 10B). However, pretreatment with GdCl3helped to
restored the decrease in multifunctional enzyme to normal
Recent reports have shown that one of the major peroxi-
somal functions(i.e., ?-oxidation) seemstobesusceptibleto
modulation by TNF-? and LPS.22,23Moreover, modulation of
peroxisomal enzymes(catalase, glutathioneperoxidase, CuZn
superoxidedismutase, andMnsuperoxidedismutase) related
with oxidative stress,24suggests a role for this organelle in
LPS-associated acute phase response.7,39Therefore, under-
standing the effects of LPS/cytokines on the structure/
function of peroxisomes is important to delineatetheroleof
peroxisomes in the pathophysiology of LPS-induced organ
failure and other peroxisomal disorders associated with
increased levels of TNF-? and other proinflammatory cyto-
kines. Thestudies reported in this manuscript clearly shown
that the LPS treatment leads to alterations in the structure/
function of peroxisomes. These conclusions are drawn from
the following observations: a sublethal dose of LPS reduced
the amount of peroxisomes and peroxisomal ?-oxidation
activity, increased the peroxisomal membrane fluidity, and
reduced the amount of phospholipid and cholesterol in
Peroxisomes are known to be associated with metabolism
of lipids including oxidation of fatty acids (very long chain
fatty acids, arachidonic acid metabolites, dicarboxylic acids,
branched chain fatty acid) and biosynthesis of cholesterol,
plasmalogen, and bile acids.26Therefore, the decrease in the
peroxisomal proteins in liver of LPS-treated animals may
result in alterations in the lipid metabolism. The previously
reported decreased activity of the peroxisomal ?-oxidation
enzyme system is consistent with the observed decrease in
theamountoftheperoxisomal protein(peroxisomes) thatwe
found in LPS-treated livers.23A decrease in peroxisomal
?-oxidation enzymes and peroxisomes was also observed in
the liver of TNF-? treated rats or during bacterial sep-
endotoxin-treated rat livers. Total lipids were extracted by Folch partition
and wereseparated into polar and neutral lipid classes by high performance
thin-layer chromatography as described in the Materials and Methods.
Phospholipid and cholesterol were quantitated by densitometric scanning.
Theresultsareexpressedasapercent changefromcontrol values(100%) for
eachofthesubcellularfractions. Dark barsrepresent phospholipidandwhite
bars represent cholesterol. Total phospholipid in control fraction was
150.6 ? 15.5 (peroxisomes); 141.3 ? 16.3 (mitochondria); 200.2 ? 18.2
(ER), expressedasnmol/mgprotein. Total cholesterol incontrol fractionwas
12.2 ? 2.1 (peroxisomes); 15.9 ? 3.8 (mitochondria); 17.0 ? 2.9 (ER),
expressed as nmol/mg protein. Thedatarepresent themean ? SD for n ? 7
(peroxisomes) and n ? 4 (mitochondria and ER each) experiments.
Statistical significance, in phospholipid and cholesterol levels(P ? .001 and
P ? .004, respectively), were found only in peroxisomes of control and
Lipid concentration of subcellular organelles from control and
HEPATOLOGY Vol. 31, No. 2, 2000CONTRERAS ET AL.
sis.21,22,40The molecular basis for a selective decrease in the
peroxisomal population, without affecting other organellesis
not known at the present time. A specific decrease in the
synthesis of peroxisomal proteins and/or higher turnover of
peroxisomes may besomeof thecontributing factors for the
The possibility of an increased turnover/autophagocytosis
of peroxisomes in endotoxin-treated liver was examined by
morphological studies. The identification of a dark/brown
peroxisomal structure41in lysosomesin fractionsfromendo-
toxin-treated liver suggests that autophagocytosis of peroxi-
somes may be one of the factors contributing to the lower
number of normal density peroxisomes in endotoxin-treated
liver. Similarly, autophagy of peroxisomes by lysosomes, was
observed in liver following withdrawal of hypolipidemic
drugs.42Thismechanismisbelieved tobeageneral phenom-
enon in theregulation ofperoxisomal population.43,44
In addition to the loss of peroxisomes by autophagy,
endotoxin treatment would seem to introduce alterations in
the protein biosynthesis of specific peroxisomal proteins.
Although endotoxin treatment decreases theamount of total
peroxisomes, residual peroxisomes have relatively higher
amounts of catalase and fatty acid ?-oxidation enzymes as
shown in Fig. 3. The decrease in peroxisomal protein in
endotoxin-treated livers directly correlates with thedecrease
in peroxisomal function (e.g., fatty acid ?-oxidation) in total
cellular homogenates (Fig. 10). Elimination/inactivation of
KC by GdCl3results in the decrease of soluble factors (e.g.,
TNF-? as seen in Fig. 9). This partially restores the peroxi-
somal ?-oxidation in total cellular homogenatesandnormal-
izes therelativeamount of catalaseand ?-oxidation proteins
in isolatedperoxisomes(Fig. 7).
While studying the protein pattern of peroxisomal mem-
branes prepared by carbonate treatment we observed that
peroxisomal membranes from liver of LPS-treated animals
disintegrate, becausethey cannot besedimented by centrifu-
gation at 100,000g. These observations show that LPS treat-
lular organelles from GdCl3- and
GdCl3/endotoxin-treated rat liver.
Rat liver subcellular organelles are
isolated from GdCl3- (black bars)
and GdCl3/endotoxin- (white bars)
treated animals by Nycodenz den-
tribution of the organelles in the
enzymes: (A) catalase for peroxi-
somes, (B) cytochromecoxidasefor
mitochondria,and(C) NADPH cyto-
chrome c reductase for ER. The
protein distribution in thegradients
was also determined (D). The data
represent the mean ? SD of the
values obtained for n ? 4 experi-
Fractionation of subcel-
GdCl3/endotoxin-treated animals. Fifty micrograms of protein was used for
SDS-PAGE. Western blot and visualization of immunoreactiveproteins with
antibodies against (A) multifunctional enzyme, (B) catalase, and (C) 3-keto
acyl-CoA thiolase.Lanes1and2representperoxisomal proteinsfromGdCl3-
Western blot analysis of peroxisomal fractions from GdCl3- and
TABLE 3. Protein Concentration in Peroxisomal Peak Fractions FromLiver
of Control, GdCl3-, and GdCl3/Endotoxin-Treated Rats
2.23 ? 0.30
2.55 ? 0.32
1.19 ? 0.18*†
NOTE. Seven gramsof wet rat liver fromeach of theexperimental groups
wereused for theseexperiments. Peroxisomal peak was identified using the
enzymatic marker catalase.
*P ? .001 (n ? 9), for comparison between peroxisomes from GdCl3/
†P ? .001 (n ? 4), for comparison between peroxisomes from GdCl3/
CONTRERAS ET AL.HEPATOLOGY February 2000
ment induced changes in the structure of the peroxisomal
membranes. However, no major difference was observed in
the protein pattern of intact peroxisomes (matrix and mem-
branes). On theother hand, wefound asignificant changein
the lipid concentration and composition (phospholipid and
cholesterol) of peroxisomal membranes from LPS-treated
livers.45Thesealterations in lipids and lipid-to-protein ratios
may account for the observed increase in the membrane
fluidity ofperoxisomesfromLPS-treatedlivers. Alterationsin
the membrane lipid composition of liver peroxisomes from
shown tobeassociatedwith increasein membranefluidity of
these organelles.35,46,47Lipid peroxidation of membranes, a
result of oxidative stress, is also known to cause changes in
membranefluidity.48However, lipid peroxidation produces a
decrease in membrane fluidity caused by a decrease in
unsaturated fatty acids49and/or formation of cross-linking
between membrane components.48Although there was an
increase in peroxisomal membrane fluidity, no such change
wasobserved in other subcellular organelles(e.g., mitochon-
driaorER-Golgi) afterLPStreatment.Therefore, theselective
changesin peroxisomal membranelipid composition may be
thebasis for theselectiveautophagy of peroxisomes in livers
of animals treated with LPS. Similarly, changes in the lipid
composition induced by peroxisomal proliferator were sug-
gested to be a signal for selective autophagy of peroxisomes
during thewithdrawal of peroxisomal proliferator drugs.50,51
The observation of lighter density peroxisomes (1.14 g/cm3)
wasacharacteristic oftheLPS-treatedliver gradients. Peroxi-
somes of different densities have been described in normal52
and regenerating rat liver.53Furthermore, peroxisomes are
induced by conditions such as hypolipidemic drugs treat-
ment,50cold shock,54and ischemiaand reperfusion.55Under
theseconditions, an additional catalasepeak ispresent in the
lighter part of the gradient. Therefore, lighter peroxisomes
may represent peroxisomes in theprocess of maturation,50,53
or peroxisomal particles generated due to fragmentation of
peroxisomal reticulum.52Theadditional peak ofperoxisomes
in LPS-treated liver represents peroxisomes undergoing au-
tophagy by lysosomes.
The observed decrease in number of peroxisomes in
LPS-treated liver indicates that some of the metabolic func-
tionmay becompromisedduringendotoxemia. Forexample,
theloss of peroxisomal ?-oxidation activity during endotox-
emia may result in higher in situ levels of bioactive arachi-
donic acid metabolites (e.g., eicosanoids), which are known
to bedegraded by theperoxisomal ?-oxidation system.26,56,57
Furthermore, fatty acids areknown to regulatesomecellular
functions (e.g., activation transcription factors58) and the
decrease in peroxisome number and function may effect the
cellular response to endotoxin. Moreover, at low concentra-
tions TNF-? released by KC was found to inducehepatocyte
proliferation, whereas as at high concentrations it had an
inflammatory response.59These observations indicate that
the observed alterations in the peroxisomal function may
GdCl3/endotoxin-treated rat liver. The procedure for lipid quantification is
similar to that described in the legend of Fig. 5. Total phospholipid and
cholesterol in peroxisomes from control fractions were 143.8 ? 16.3 and
10.7 ? 1.7 (nmol/mg protein), respectively. The results are expressed as a
percent changeof thecontrol values(100%). Thedatarepresent themean ?
SD (n ? 7). The dark bar represents phospholipid and the white bar
represents cholesterol. Differences in phospholipid and cholesterol were
statistically significant (both P ? .001), in the peroxisomal fractions of
Lipid composition of subcellular organelles from GdCl3- and
and endotoxin plus GdCl3(Gad) werefixed and probed for thepresenceof TNF-? using anti–TNF-? antibodies and visualized with Texas Red conjugated
streptavidin asdescribedin theMaterialsandMethods.
Immunocytochemistry ofcontrol, endotoxin, andendotoxin/GdCl3ratliveronTNF-? production. (A) Ratliversfromcontrol, endotoxin-treated,
HEPATOLOGY Vol. 31, No. 2, 2000CONTRERAS ET AL.
assistance with the animal injections. They also thank Dr.
RichardSiarey for readingthemanuscript.
The authors thank Dr. James Cook for
1. Sonesson HRA. Bacterial endotoxin. Chemical structure and biological
activity. In: Brigham KL, ed. Lung Biology in Health and Disease:
Endotoxin and the Lungs. Volume 77. New York: Marcel Dekker, Inc.,
2. Baumannn H, Gauldie J. The acute phase response. Immunol Today
3. MichieHR, ManogueKR, SpriggsDR, Revhaug A, O’Dwyer S, Dinarello
CA, Cerami A, et al. Detection of circulating tumor necrosis factor after
endotoxin administration. N Engl J Med1988;318:1481-1486.
4. Tracey KJ, Beutler B, Lowry SF, Merryweather J, Wolpe S, Milsark IW,
Hariri RJ, et al. Shock and tissueinjury induced by recombinant human
5. BeutlerB, Milsark IW, Cerami A.Passiveimmunizationagainstcachectin/
tumor necrosis factor protects mice from lethal effect of endotoxin.
6. Tracey KJ, Fong Y, HesseDG, ManogueKR, LeeAT, Kuo GC, Lowry SF,
et al. Anti-cachectin/TNF monoclonal antibodies prevent septic shock
duringlethal bacteraemia. Nature1987;330:662-664.
7. Darlington GJ, Wilson DR, Revel M, Kelly JH. Responseof liver genesto
acutephasemediator. Ann N Y AcadSci 1989;557:310-315.
8. Decker T, Lohmann-Matthes M-L, Karck U, Peters T, Decker K.
Comparativestudyofcytotoxicity, tumornecrosisfactor, andprostaglan-
din release after stimulation of rat Kupffer cells, murine Kupffer cells,
and murine inflammatory liver macrophages. J Leuk Biol 1989;45:139-
9. Dubravec DB, SpriggsDB, Mannick JA, Rodrick ML. Circulating human
peripheral blood granulocytes synthesize and secrete tumor necrosis
factor ?. Proc Natl AcadSci U SA 1990;87:6758-6761.
10. Beutler B, Brown T. A CAT reporter construct allows ultrasensitive
estimation of TNF synthesis, and suggests that the TNF gene has been
silencedinnon-macrophagecell lines. J ClinInvest 1991;87:1336-1344.
11. Kunkel SL, Chensue SW, Standiford TJ, Strieter RM. Endotoxin-
dependent cytokine networks. In: Brigham KL, ed. Lung Biology in
Health and Disease: Endotoxin and the Lungs. Volume 77. New York:
Marcel Dekker, Inc., 1994:305-320.
12. Hagmann W, Denzlinger C, Keppler D. Production of peptide leukotri-
enesin endotoxin shock. FEBSLett 1985;180:309-313.
13. Bautista AP , Schuler A, Spolarics Z, Spitzer JJ. Tumor necrosis factor-?
cells. AmJ Physiol 1991;261:G891-G895.
14. de Vera ME, Geller DA, Billiar TR. Hepatic inducible nitric oxide
synthase: regulation and function. Biochem Soc Trans 1995;23:1008-
15. Kaneda K, Morisawa Y, Kawai Y, Akagawa K, Wake K. Bacterial
aminolipids induce the Kupffer cell activation and give a suppressive
effect on the LPS-induced hepatocyte injury. In: Wise E, Knook DL,
McCuskey RS, eds. CellsoftheHepatic Sinusoid. Volume3. Leiden: The
Kupffer Cell Foundation/TheNetherlands1991:436-438.
16. Fujita S, Arii S, Monden K, Ishiguro S, Nakamura T, Mizumoto M,
ImamuraI. Thepreventiveeffectsof KO432 on endotoxin-inducedliver
injury: liver protection by the modulation of hepatic macrophage
function. SurgToday Jpn J Surg1996;26:29-35.
17. Spitzer JJ. Bacterial endotoxin effects on carbohydrate utilization and
transport. BiochemSoc Trans1995;23:998-1002.
18. Hardardottir I, Grunfeld C, Feingold KR. Effects of endotoxin and
cytokineson lipidmetabolism. Curr Opin Lipidol 1994;5:207-215.
19. Hasselgren PO. Muscle protein metabolism during sepsis. Biochem Soc
20. Weiner FR, Smith PJ, Wertheimer S, Rubin CS. Regulation of gene
expression by insulin andtumor necrosisfactor ? in 3T3-L1 cells. J Biol
21. Yasmineh WG, Parkin JL, Casper JI, Theologides A. Tumor necrosis
factor/cachectin decreases catalase activity of rat liver. Cancer Res
22. Beier K, Volkl A, Fahimi HD. Suppression of peroxisomal lipid ?-oxida-
tion enzymesby TNF-?. FEBSLett 1992;310:273-276.
23. Dhaunsi GS, HanevoldCD, SinghI.Impairmentofperoxisomal ?-oxida-
tion system by endotoxin treatment. Mol Cel Biochem 1994;135:187-
24. Dhaunsi GS, Singh I, Hanevold CD. Peroxisomal participation in the
cellular responseto theoxidativestress of endotoxin. Mol Cel Biochem
25. Feingold KR, Pollock AS, Moser AH, Shigenaga JK, Grunfeld C.
Discordant regulation of proteins of cholesterol metabolism during the
acutephaseresponse. J LipidRes1995;36:1474-1482.
26. Singh I. Biochemistry of peroxisomes in health and disease. Mol Cel
27. SarphieTG, D’SouzaNB, Deaciuc IO. Kupffer cell inactivation prevents
lipopolysaccharide-induced structural changes in therat liver sinusoid:
an electron-microscopic study. HEPATOLOGY 1996;23:788-796.
28. LazoO, ContrerasM, BhushanA, StanleyW, SinghI.Adrenoleukodystro-
phy: impaired oxidation of fatty acids due to peroxisomal lignoceroyl-
CoA ligasedeficiency. Arch BiochemBiophys1989;270:722-728.
29. Bradford M. A rapid and sensitive method for the quantitation of
microgram quantities of protein utilizing the principle of protein-dye
binding. Anal Biochem1976;72:248-254.
30. Contreras M, Sengupta TK, Sheikh F, Aubourg P , Singh I. Topology of
ATP-binding domain of Adrenoleukodystrophy geneproduct in peroxi-
somes. Arch BiochemBiophys1996;334:369-379.
31. Folch J, Lee M, Sloane-Stanley GH. A simple method for the isolation
and purification of total lipids from animal tissues. J Biol Chem
32. Ganser AL, Kerner AL, Brown BJ, Davisson MT, Kirschner DA. A survey
of neurological mutant mice. I. Lipid composition of myelinated tissues
in known myelin mutants. Dev Neurosci 1988;10:99-122.
33. Stewart JCM. Colorimetric determination of phospholipid with ammo-
niumferrothiocyanate. Anal Biochem1980;104:10-14.
34. Fujiki Y, Hubbard S, Fowler S, Lazarow PB. Isolation of intracellular
membranes by means of sodium carbonate treatment: application to
endoplasmic reticulum. J Cell Biol 1982;93:97-102.
35. Hayashi H, NakataK, Hashimoto F. Study of membranefluidity of liver
peroxisomes. In: Fahimi HD, Sies H, eds. Peroxisomes in Biology and
Medicine. Berlin: Springer-Verlag/Heidelberg1987:205-209.
36. Lazo O, Contreras M, Yoshida Y, Singh AK, Weise M, Singh I. Cellular
lignoceroyl-CoA ligases. J LipidRes1990;31:583-595.
Rat liver homogenates(150µg) fromthecontrol, endotoxin, andGdCl3plus
endotoxin were assayed for peroxisomal ?-oxidation as described in the
Materials and Methods. Control, n ? 3; endotoxin, n ? 3; and endotoxin
plus GdCl3, n ? 2. (B) The homogenate (50 µg) were analyzed by Western
blot for multifunctional enzyme (78 kd) in (1) control, (2) endotoxin-
treated, and(3) GdCl3plusendotoxin–treatedanimals.
Endotoxin treatment decreases peroxisomal ?-oxidation. (A)
CONTRERAS ET AL.HEPATOLOGY February 2000
37. Schulze-OsthoffK, BakkerAC, Vanhaesebroeck B, BeyaertR, JacobsWA, Download full-text
FiersW. Cytotoxic activity of tumor necrosisfactor ismediated by early
damage of mitochondrial functions. Evidence for the involvement of
mitochondrial radical generation. J Biol Chem1992;267:5317-5323.
38. Koudstaal J, Dijkhuis FWJ, Hardonk MJ. Selective depletion of Kupffer
cells after intravenous injection of gadolinium chloride. In: Wise E,
Knook DL, McCuskey RS, eds. Cellsof theHepatic Sinusoid. Volume3.
Leiden: TheKupffer Cell Foundation/TheNetherlands1991:87-91.
39. Perlmutter DH, Dinarello CA, Punsal PI, Colten H. Cachectin/tumor
necrosis factor regulates hepatic acute-phase gene expression. J Clin
40. CanonicoPG, WhiteJD, PowandaMC. Peroxisomedepletionofratliver
duringPneumococcal sepsis. LabInvest 1975;33:147-150.
41. Goldfischer S, Essner E. Further observations on the peroxidatic
42. Yokota S. Formation of autophagosomes during degradation of excess
peroxisomes induced by administration of dioctyl phthalate. Eur J Cell
43. IshikawaT, FurunoK, KatoK. Ultrastructural studieson autolysosomes
in rat hepatocyte after leupeptin treatment. Exp Cell Res 1983;144:15-
44. HeikoopJC, vandenBergM, StrijlandA, WeijersPJ, JustWW, MeijerAJ,
Tager JM. Turnover of peroxisomal vesiclesby autophagic proteolysisin
culturedfibroblastsfromZellwegerpatients.EurJ Cell Biol 1992;57:165-
45. Khan M, Contreras M, Singh I. Endotoxin-mediated peroxisomal lipid
composition alterations[Abstract]. FASEBJ 1996;10:522A.
46. Sartori C, Stefanini S, Cimini A, Di GiulioA, CeruMP .Liverperoxisomes
in newborns from clofibrate-treated rats. II. A biochemical study of the
recovery period. Biol Cell 1992;74:315-324.
47. Kyrklund T, Meijer J. Lipid composition of liver peroxisomes isolated
from untreated and clofibrate-treated mice and rats. Comp Biochem
48. Chen JJ, Yu BP . Alterations in mitochondrial membranefluidity by lipid
peroxidation products. FreeRadBiol Med1993;17:411-418.
49. Curtis MT, Gilfor D, Farber JL. Lipid peroxidation increases the
molecular order of microsomal membranes. Arch Biochem Biophys
50. Staubli W, SchweizerW, SuterJ, Weibel ER. Theproliferativeresponseof
hepatic peroxisomes of neonatal rats to treatment with SU-13 437
(Nafenopin). J Cell Biol 1977;74:665-689.
51. Veenhuis M, Douma A, Harder W, Osumi M. Degradation turnover of
peroxisomes in the yeast Hansenula polymorpha induced by selective
inactivation of peroxisomal enzymes. Arch Microbiol 1983;134:193-
52. Wilcke M, Hultenby K, Alexson S. Novel peroxisomal population in
subcellular fractions from Rat Liver. Implications for peroxisomestruc-
tureandbiogenesis. J Biol Chem1995;270:6949-6958.
53. Luers G, Hashimoto T, Fahimi HD, Volkl A. Biogenesis of peroxisomes:
Isolation and characterization of two distinct peroxisomal populations
fromnormal and regenerating rat liver. J Cell Biol 1993;121:1271-1280.
54. Goglia F, Liverini G, Lanni A, Iossa S, Barletta A. Morphological and
functional modificationsof rat liver peroxisomal subpopulationsduring
coldexposure. ExpBiol 1989;48:127-133.
55. Gulati S, Singh AK, Irazu C, Orak J, Rajagopalan PR, Fitts CT, Singh I.
Ischemia-reperfusion injury: biochemical alterations in peroxisomes of
rat kidney. Arch BiochemBiophys1992;295:90-100.
56. Jedlitschky G, Huber M, Volkl A, Muller M, Leier I, Muller J, Lehman
WD, et al. Peroxisomal degradation of leukotrienesby ?-oxidation from
the?-end. J Biol Chem1991;266:24763-24772.
57. Diczfalusy UG, Alexon EH. Role of peroxisomes in the degradation of
prostaglandins. ProgClin Biol Res1992;375:253-261.
58. Schmidt A, Endo N, Rutledge SJ, Vogel R, Shinar D, Rodan GA.
Identification of a new member of the steroid hormone receptor
superfamily thatisactivatedby aperoxisomal proliferatorandfatty acid.
Mol Endocrinol 1992;6:1634-1641.
59. Rose ML, Rusyn I, Bojes HK, Germolec DR, Luster M, Thurman RG.
Role of Kupffer cells in peroxisome proliferator-induced hepatocyte
proliferation. DrugMetabRev 1999;31:87-116.
HEPATOLOGY Vol. 31, No. 2, 2000CONTRERAS ET AL.