Full text of DOI:10.1093/gerona/56.6.B259

Journal of Gerontology: BIOLOGICAL SCIENCES
Copyright 2001 by The Gerontological Society of America
2001, Vol. 56A, No. 6, B259–B267
Aging Lowers Steady-State Antioxidant Enzyme and
Stress Protein Expression in Primary Hepatocytes
David M. Hall,1 Gerald L. Sattler,3 Carol A. Sattler,3 Hannah J. Zhang,1 Larry W. Oberley,2
Henry C. Pitot,3 and Kevin C. Kregel1
1Department of Exercise Science and 2Free Radical and Radiation Biology Program, The University of Iowa, Iowa City.
3McArdle Laboratory for Cancer Research, University of Wisconsin Medical School, Madison.
It has been reported that the isolation and culture of primary hepatocytes can compromise cellu-
lar ability to constituitively express antioxidant enzyme (AE) genes, making it difficult to study
their regulation ex vivo. In the present study, the steady-state expression of manganese-contain-
ing superoxide dismutase, copper- and zinc-containing superoxide dismutase, catalase, and glu-
tathione peroxidase was assessed in primary hepatocytes isolated from young and senescent rats
and cultured in Matrigel. There was no change in steady-state superoxide dismutase protein or
activity levels in cells collected from young animals and cultured for 7 days. Catalase expression
was initially increased, and then it declined 30%. In contrast, superoxide dismutase expression
declined 60% and catalase expression declined 50% in cells from senescent animals. Constitu-
tive and inducible 70-kDa heat shock protein expression increased coincident with declining AE
levels in the young cells but not senescent cells. For both age groups, electron micrographs
showed rounded hepatocytes with abundant rough endoplasmic reticulum, mitochondria, and
peroxisomes. Hepatocytes were organized into clusters of 6–12 cells surrounding a large central
lumen devoid of microvilli. Each cluster also contained smaller microvilli-lined lumens between
adjacent hepatocytes that resembled canniculi. The plasma membranes of these lumens were
sealed from the extracellular space by junctional complexes. Gap junctions in the plasma mem-
brane suggest that hepatocytes were capable of intercellular communication. We conclude that
the Matrigel system can be used to study AE regulation in primary hepatocytes from young and
senescent animals, provided that experiments can be conducted within a time frame of 5–7 days
in culture. These data also support the hypothesis that aging compromises hepatocellular ability
to maintain AE status and upregulate stress protein expression.
GING impairs eukaryotic ability to upregulate stress
protein expression in response to environmental chal-
(14), and ultimately, gene expression (15). We have re-
cently observed that aging lowers the capacity of the liver to
induce AEs in response to environmental heat stress (1) and
significantly increases morbidity (16). Morbidity in the
aged animals was closely linked with an increased genera-
tion of radical biomarkers of cellular oxidative stress. Al-
though these are important in vivo observations that appear
to link hyperthermia and cellular oxidative stress, the mech-
anisms responsible for age-related deficiencies in AE ex-
pression are unknown. Therefore, we have adopted a reduc-
tionist approach to the problem by establishing a primary
hepatocyte culture system that will permit us to address
mechanistic questions related to age-associated changes in
cell function.
Differentiated primary cells hold many advantages over
immortalized cell lines in aging research. However, several
research groups have reported that liver-specific functions
are compromised (17–19) and that nuclear transcription of
the AE genes is significantly lowered (20) following the
isolation and culture of hepatocytes. Recent advances in pri-
mary cell culture environments have shown that liver-spe-
cific functions (e.g., cytochrome-P450 activity, albumin
A
lenges such as heat stress (1,2), exercise (3), and viral infec-
tion (4). In particular, aged organisms have a reduced
capacity to rapidly synthesize protective cellular stress pro-
teins such as heat shock proteins (2,5,6) and antioxidant en-
zymes (1). Mechanisms responsible for these critical age-
related deficiencies in the cellular stress response are un-
clear, but experimental evidence implicates an elevation in
steady state cellular electromotive force (7,8) and poor tran-
scriptional regulation of redox sensitive acute response
(stress) genes (5,9,10).
The primary antioxidant enzymes (AEs) manganese-con-
taining superoxide dismutase (MnSOD), copper- and zinc-
containing SOD (CuZnSOD), catalase, and glutathione per-
oxidase (GPx) are an important subset of the stress proteins
that can protect cells against the damaging effects of hyper-
thermia (11), viral infection (4), and ionizing radiation (12).
Because of its ability to detoxify oxidants directly during
episodic stress, the AE system is an important determinant
of cellular electromotive force (8), which in turn influences
cell signaling events (13), transcription factor activation
B259

B260
HALL ET AL.
production) can be maintained at near-normal levels for up
to 21 days in culture (21,22). However, no group has revis-
ited the issue of steady-state AE expression in cultures of
primary cells subsequent to these developments.
ꢃ 106 cells/well for the six-well dishes and 12–14 ꢃ 106
cells/100-mm dish. The culture medium was changed every
2 days per manufacturer’s instruction, and cells were har-
vested for assay at the indicated times by using Matrisperse
(Becton Dickinson) to depolarize the Matrigel.
Using the hepatocyte growth matrix Matrigel (Becton
Dickinson Labware, Bedford, MA), the present study as-
sessed the steady-state levels of MnSOD, CuZnSOD, cata-
lase, and GPx as well as the expression of the constitutive
(Hsc70) and inducible (Hsp70) isoforms of the 70-kDa heat
shock protein in cultures of hepatocytes isolated from
young and senescent rats. We hypothesized that the stresses
associated with hepatocyte isolation and culture (e.g., dis-
ruption of the liver, exposure to 20% oxygen) would stimu-
late adaptive increases in cellular Hsp70 and AE expression.
We further postulated that cells collected from aged rats
would lag behind their younger counterparts in adapting to
culture conditions.
Protein Immunoblots
Cells for protein immunoblots were homogenized in
0.05M of phosphate buffer (pH 7.8) and then briefly spun at
4ꢂC in a clinical microcentrifuge. The supernatants were
sonicated on ice and protein content was quantified by the
Bradford assay (BioRad Laboratories, Richmond, CA).
Equal protein amounts of cell homogenate were separated
on a one-dimensional 12.5% polyacrylamide gel under stan-
dard denaturing conditions according to the method of
Laemmli (25). After separation by electrophoresis, proteins
were transferred to nitrocellulose membranes (100 V, 4ꢂC, 1
hour). The membranes were blocked overnight at 4ꢂC with
5% dry milk and 2% bovine serum albumin in TBST, which
is 10 mM of tris(hydroxymethyl) aminomethane (Tris), 150
mM of NaCl, pH 8.0, and 0.1% Tween 20. The nitrocellu-
lose membranes were incubated for 1 hour at room tempera-
ture with antibodies specific for MnSOD (1:2000 dilution),
CuZnSOD (1:200 dilution), catalase (1:500 dilution), Hsc70
(1:500 dilution), and Hsp70 (1:1000 dilution). The AE blots
were probed with a goat antirabbit IgG peroxidase anti-
body (Boehringer Mannheim) for 1 hour at room tempera-
ture (1:10,000 dilution for MnSOD and 1:5000 dilution for
CuZnSOD and catalase). The Hsc70 blot was probed with
rabbit antirat IgG peroxidase antibody (1:5000; StressGen
Biotechnologies, Victoria, BC, Canada), and the Hsp70 blot
was probed with goat antimouse IgG peroxidase antibody
(1:1000; Sigma Chemical). Blots were developed by using
the ECL-Western blot detection kit (Amersham Life Sci-
ence, Piscataway, NJ) and BioMaxMR film (Kodak, Roch-
ester, NY).
METHODS
Isolation and Culture of Hepatocytes
All procedures for handling rats were approved by the
University of Wisconsin and The University of Iowa Insti-
tutional Animal Care and Use Committees. Hepatocytes for
primary culture were isolated from 6- and 24-month-old
male Fischer 344 rats that were obtained from Harlan Labs
(Harlan, Madison, WI). Briefly, hepatocytes were isolated
by using the Seglen (23) in situ collagenase perfusion
method with the following modifications.
The liver was perfused via the inferior vena cava with
Ca2ꢀ, Mgꢀ-free Hanks balanced salt solution (phenol red
free, pH 7.4) supplemented with penicillin (100 units/ml),
streptomycin (100 ꢁg/ml), insulin (5 ꢁg/ml), and glucose
(1.5 mg/ml; all from Sigma Chemical Co., St. Louis, MO).
After the initial blanching of the liver, the perfusion was
continued with the same buffer plus 2 mg/ml of Type 1 col-
lagenase (Boehringer Mannheim Biochemicals, Indianapo-
lis, IN) and 0.15 mg/ml of soybean trypsin inhibitor
(Sigma). Dissociated cells were washed twice in cold L15
medium supplemented with penicillin, streptomycin, insu-
lin, and glucose as above, and then they were purified by
Percoll isodensity centrifugation (24). This procedure
yielded a nearly pure population of hepatocytes with 95% or
greater viability. One control aliquot of cells was prepared
for immediate assessment of AE, Hsc70, and Hsp70 protein
levels. A second control aliquot was incubated in a spinner
flask at 4ꢂC for 3 hours followed by determination of AE,
Hsc70, and Hsp70 protein levels. A third control consisted
of freshly isolated liver tissue collected from a littermate.
There were no differences in the respective protein levels
among these three controls; therefore, hepatocytes from two
rats were pooled and incubated in spinner flasks for 3 hours
prior to plating. This allowed us to obtain enough cells for
the multiple time points in these experiments.
Antioxidant Enzyme Activity Gels
SOD, catalase, and GPx activities were analyzed by using
native polyacrylamide gel electrophoresis (PAGE) gels. A
10% gel was used for evaluating SOD activity and an 8%
gel was used for determining catalase and GPx activities.
Protein concentrations were measured by the Bradford
method (BioRad) and proteins (75 ꢁg for SOD and 50 ꢁg
for catalase and GPx) were separated under nondenaturing
conditions. SOD activity was visualized by staining gels in
2.43 mM of Nitro Blue Tetrazolium (NBT) in distilled wa-
ter for 20 minutes, then incubating the gels for 15 minutes in
50 mM of phosphate buffer (pH 7.8) containing 2.8 ꢃ 10ꢄ5
M riboflavin and 28 mM of tetramethylethylenediamine
(TEMED) (26). So that only MnSOD activity would be ob-
served; 0.75 mM of NaCN (which inhibits CuZnSOD activ-
ity) was added to both the NBT and riboflavin–TEMED so-
lutions. However, rat MnSOD is not visualized in these
activity gels; therefore, SOD activity results represent only
CuZnSOD activity. All incubations were performed at room
temperature in the dark. Superoxide dismutase bands were
observed by exposing gels to fluorescent light.
The purified and washed cells were suspended in Hepa-
toSTIM (Becton Dickinson) medium supplemented with
penicillin–streptomycin, glutamine (2 mM), and epidermal
growth factor (10 ng/ml) and then plated in either six-well
or 100-mm tissue culture dishes that had been precoated
with Matrigel (Becton Dickinson). Plating density was 1.25
Catalase activity was measured by incubating gels in
0.003% H2O2 for 10 minutes (27). GPx activity was mea-

AGING, ANTIOXIDANT ENZYMES, AND PRIMARY CELLS
B261
sured by incubating gels in 0.003% cumene hydroperoxide
in 1 mM of glutathione (27). Gels were rinsed with distilled
water and then stained with 2% potassium ferricyanide and
2% ferric chloride until bands were visible. Achromatic
bands indicate removal of peroxides and thus are a direct as-
sessment of catalase or GPx activity. Activity gels were
thoroughly washed following staining procedures, and pho-
tographic images were obtained by using the Digital Imag-
ing and Analysis System (Alpha Innotech Corporation, San
Leandro, CA).
Data Analysis
Densitometries were performed on immunblots by using
the National Institutes of Health Image 1.61 software pack-
age (NIH, Bethesda, MD).
Transmission Electron Microscopy
Primary hepatocytes isolated from senescent rats and cul-
tured in six-well plates were processed for transmission
electron microscopy on days 1, 3, 5, 14, and 21 in culture.
Individual cultures were gently washed with phosphate-
buffered saline (PBS) and fixed at room temperature in 3%
glutaraldehyde in 0.1M of cacodylate buffer, pH 7.4 for 45
minutes. Cultures were then washed in three changes of
0.1M cacodylate and stored overnight at 4ꢂC. Following a
45-minute postfixation on ice with 2% osmium tetroxide in
0.1M cacodylate (pH 7.4), cultures were washed three times
with cold distilled water. A scalpel was used to detach the
Matrigel “wafer” from the edges of each well, and one cut
was made through the center of the wafer. The thick Matri-
gel layer containing clusters of fixed hepatocytes was sepa-
rated from the dish with a plastic Swiss spatula and trans-
ferred to tissue culture dishes containing distilled water,
where they were cut into strips approximately 3–4 mm long
and 1.5 mm wide. Samples were then transferred to tissue
carrier cylinders (Ted Pella, Redding, CA), dehydrated
through a series of graded alcohols, and then placed in small
glass vials with propylene oxide followed by 1 hour infiltra-
tion with 1:1 propylene oxide:Eponate resin (Ted Pella). Af-
ter overnight infiltration with 100% Eponate, samples were
imbedded in freshly made Eponate in flat embedding molds.
Thin sections were cut on a Reichert Ultracut E3 ultrami-
crotome (Reichert, Austria) equipped with a diamond knife,
and they were stained with uranyl acetate and lead citrate
before examination in a Hitachi H-7000 electron micro-
scope (Hitachi Scientific Instruments, San Jose, CA) oper-
ated at 75 kV.
Figure 1. Time course of steady-state manganese-containing su-
peroxide dismutase (MnSOD) protein expression in hepatocyte cul-
tures derived from young and senescent rats. Primary hepatocytes
were isolated from A, young (6-month-old), and B, senescent (24-
month-old), Fischer 344 rats and cultured for 14 days as described in
Methods. Cultures were harvested at the time indicated and MnSOD
protein was quantitated by using Western immunoblots and densito-
metry analyses. Hepatocytes isolated from young rats maintained
steady-state MnSOD expression at or near control levels for 10 days
in culture with only minor decrements in steady-state protein levels at
day 14. In contrast, cultures derived from senescent animals main-
tained MnSOD expression at control levels through 5 days in culture
followed by an approximate 60% decline in steady-state protein ex-
pression at day 14.
tially increased on days 1–3, and then slowly declined to
ꢀ65% of control by day 14. Activity gels specific for
CuZnSOD, catalase, and GPx (Figure 4) demonstrate that
SOD activity was maintained for 14 days in culture,
whereas peroxide-removing capacity was sharply lower on
days 10 and 14.
In contrast to cells derived from young rats, hepatocytes
collected from old animals maintained AE expression at or
near control levels for 5 days in culture followed by pro-
gressively declining AE protein and activity at days 7–14.
Both MnSOD (Figure 1) and catalase (Figure 3) expression
were initially increased above control on days 1–3, followed
by a progressive decline in steady-state protein levels to
ꢀ40% of control by day 14. Interestingly, CuZnSOD ex-
RESULTS
Antioxidant Enzymes and Heat Shock Proteins
We measured a time course of AE protein and enzymatic
activity levels in hepatocyte cultures to gain insight into the
effects of isolation and culture on steady-state cellular anti-
oxidant capacity. Figures 1–3 show that hepatocytes derived
from young rats maintained steady-state AE expression at
or near control levels for 7 days in culture. MnSOD (Figure
1) and CuZnSOD (Figure 2) expression were unaltered
through 7 days in culture and only marginally diminished at
days 10–14. Steady-state catalase levels (Figure 3) were ini-

B262
HALL ET AL.
Figure 3. Time course of steady-state catalase protein expression
in hepatocyte cultures derived from A, young, and B, senescent rats.
Hepatocyte cultures derived from both young and senescent animals
showed an initial increase in catalase expression followed by declin-
ing steady-state catalase expression at days 5–14. Catalase expression
fell 35% in the young cultures and 50% in the old.
Figure 2. Time-course of steady-state copper- and zinc-containing
superoxide dismutase (CuZnSOD) protein expression in hepatocyte
cultures derived from young and senescent rats. Primary hepatocytes
were isolated from A, young (6-month-old), and B, senescent (24-
month-old) Fischer 344 rats and cultured for 14 days as described in
Methods. Cultures were harvested at the time indicated and CuZn-
SOD protein was quantitated by using Western immunoblots and
densitometry analyses. Hepatocytes isolated from young rats main-
tained steady-state CuZnSOD expression at or near control levels for
14 days in culture. In contrast, cells collected from senescent animals
presented lower CuZnSOD protein on day 3 followed by progres-
sively increased protein levels on days 5–10. CuZnSOD expression
declined 75% between days 10 and 14.
Hsc70 expression was acutely depressed in hepatocytes
from both age groups following 1 day in culture. However,
young cells rapidly recovered the ability to synthesize
Hsc70, whereas cells from senescent animals did not.
Steady-state Hsc70 expression was back to control levels by
day 3 in the young cohort, and it was maintained at or near
control levels through 14 days. A slight elevation in Hsc70
expression was observed at day 10 in the young cells (Fig-
ure 5), which was concurrent with the decline in cellular AE
activity (Figure 4). In contrast, steady-state Hsc70 expres-
sion fell to ꢀ50% of control on day 1 in the senescent
group, and to ꢅ25% of control on days 3–10. By day 14,
Hsc70 levels were barely above the limit of detection on im-
munoblots.
A similar trend was observed with Hsp70 (Figure 6).
Steady-state Hsp70 protein expression was maintained at or
near control levels for 7 days in culture in the young group.
On days 10–14 the young cells responded to declining AE
activity with an increase in Hsp70 expression. These results
demonstrate that young cells maintained the capacity to
constitutively express both Hsc70 and Hsp70, and to re-
pression appeared to be elevated in the senescent cells on
days 5–10 (Figure 2), which coincided with declining Mn-
SOD and catalase expression. Increased CuZnSOD expres-
sion on days 7–10 was followed by a sharp 80% decline in
CuZnSOD levels at day 14 (Figure 2).
Catalase activity was well preserved in the senescent
cells through 10 days in culture (Figure 4), and catalase ac-
tivity gels closely matched the level of immunoreactive pro-
tein measured by immunoblot (compare Figures 3 and 4).
However, CuZnSOD activity was declining on days 5, 7,
and 10 while protein levels were increasing, suggesting that
a portion of the new protein was inactive.
Steady-state Hsc70 and Hsp70 protein levels were mea-
sured as indicators of the magnitude of cell stress and the
cellular ability to respond to stress. Figure 5 shows that

AGING, ANTIOXIDANT ENZYMES, AND PRIMARY CELLS
B263
Figure 4. Antioxidant enzyme activities in hepatocyte cultures de-
rived from A, young, and B, senescent rats. Hepatocytes isolated from
young rats maintained steady-state copper- and zinc-containing su-
peroxide dismutase (CuZnSOD) and catalase activity at or near con-
trol levels for 14 days in culture, with glutathione peroxidase (GPx)
activity declining at 7–10 days. In contrast, the cultures of cells from
older rats showed a loss of catalase activity by day 5 and declining
GPx and CuZnSOD activities at day 7. By day 14, CuZnSOD activity
was extremely low and GPx activity was below the detection limits of
the system.
Figure 5. Aging reduces steady-state Hsc70 protein expression in
primary hepatocytes from A, young, and B, senescent rats. Hepato-
cytes isolated from young rats showed an initial decrease in Hsc70 ex-
pression at day 1 in culture, followed by a return to control levels at
days 3–14. In contrast, steady-state Hsc70 expression was reduced by
ꢀ50% at day 1 in culture and ꢀ75% at day 3 in cells collected from
senescent animals. Steady-state levels of Hsc70 protein continued to
decline in the senescent cells. By day 14, Hsc70 protein was less than
10% of that observed in controls.
spond to declining cellular AE levels with an induction of
both proteins. In contrast, Hsp70 expression was not ele-
vated above control levels in the senescent cells at any time
point in the study (Figure 6).
Cell Morphology and Cell-to-Cell Interaction
All micrographs presented (Figures 7–11) are from old
cells. Cells from the young groups had similar morphology
(micrographs not presented). Figure 7 shows light micro-
graphs depicting the typical organization of hepatocytes
within the Matrigel matrix. Hepatocytes maintained a
rounded morphology that is typical of differentiated cells
and formed progressively larger clusters of spherical cells.
Electron micrographs of cells cultured for 1 day (Figure
8) showed many clusters of two or three hepatocytes pos-
sessing cytoplasmic organelles typical of normally func-
tioning cells, including well-defined rough endoplasmic
reticulum (RER), mitochondria, and peroxisomes. By day 3
(Figure 9), hepatocytes were organized into clusters of 6–12
cells surrounding a central lumen devoid of microvilli. Each
cluster also contained small microvilli-lined lumens be-
tween adjacent hepatocytes that resembled canniculi. The
plasma membranes of these lumens were sealed from the
extracellular space by junctional complexes, and gap junc-
tions were present in the lateral plasma membranes between
adjacent cells (Figure 9 insert), suggesting that hepatocytes
were capable of intercellular communication. Individual
cells displayed abundant RER in lamellar arrays, often in
association with mitochondria.
At days 5–14 (Figures 10 and 11), cells were organized
into clusters of 10 or more hepatocytes. Hepatocytes in the
day 5 cultures (Figure 10) contained numerous intact mito-
chondria, lamellar arrays of RER associated with mitochon-
dria, golgi arrays, and peroxisomes. Lateral plasma mem-
branes displayed desmosomes and gap junctions (Figure 10
insert) as well as small microvilli-lined lumens. By day 14,
very large lumens devoid of microvilli were observed (Fig-
ure 11), and cells contained numerous autophagic vacuoles.
However, day 14 cultures still exhibited gap junctions (Fig-
ure 11 insert), RER, intact mitochondria, and peroxisomes.
These morphological data are consistent with the biochemi-
cal assays (Figures 1–6), indicating that basic cell function
was maintained during the culture period.

B264
HALL ET AL.
Figure 6. Aging reduces hepatocellular capacity to express Hsp70
in A, young, and B, senescent rats. Hepatocytes isolated from young
rats maintained steady-state Hsp70 expression at or near control lev-
els through 10 days in culture. Hsp70 expression was elevated at days
10–14. In contrast, steady-state Hsp70 expression was not increased
in cells collected from senescent animals at any time point in the
study.
DISCUSSION
The main findings of the present study indicate that pri-
mary hepatocytes collected from young and senescent rats
and cultured using Matrigel adopt a differentiated pheno-
type, establish complex cell-to-cell interactions, and main-
tain the ability to constituitively express primary AE genes
through 14 days in culture. These results are in contrast to
previously published data reporting that primary hepatocyte
cultures rapidly lose the ability to express AE proteins ex
vivo (20).
Our results also suggest that hepatocytes collected from
senescent rats may be less competent than cells isolated
from younger animals at withstanding the environmental in-
sults associated with disruption of the liver and long-term
culture. Hepatocyte cultures from senescent animals were
morphologically similar to cultures of younger cells, but
they maintained SOD and catalase expression at control lev-
els for 3–5 days as opposed to 7–10 days for the younger
cells. The resulting decrement in catalase expression was not
offset by a compensatory increase in GPx activity. We specu-
Figure 7. Light micrographs of hepatocyte cultures derived from
senescent rats. Hepatocytes maintained a rounded morphology and
formed progressively larger clusters of spherical cells. A, day 1; B,
day 3; C, day 5; and D, day 14 in culture.
late that an early imbalance in the SOD/peroxidase ratio
may have exposed the senescent cultures to higher intracel-
lular fluxes of H2O2 or lipid peroxides. Progressively in-
creasing numbers of cellular autophagic vacuoles in the se-
nescent cultures support this interpretation.
In addition to declining AE expression, Hsc70 levels fell
to ꢀ25% of control in the senescent cells after only 3 days
in culture. Neither Hsc70 nor Hsp70 expression was ele-
vated above control levels in the senescent cultures at any
time point in the study. In contrast, cells from younger rats

AGING, ANTIOXIDANT ENZYMES, AND PRIMARY CELLS
B265
Figure 8. Electron micrographs of hepatocytes from senescent rats
after 1 day in culture. Hepatocytes formed small aggregates of two to
three cells. Peripheral plasma membranes had numerous cytoplasmic
projections; microvilli-lined lumens were located between cells. Cyto-
plasmic lipid (L) deposits occurred near the cell periphery. Rough en-
doplasmic reticulum (RER) was prominent and intact mitochondria
(M) and peroxisomes (P) were numerous; magnification 3600ꢃ.
increased both Hsc70 and Hsp70 expression coincident with
declining cellular AE levels. We interpret these results as
evidence that aging reduces cellular capacity to express heat
shock protein genes ex vivo. These results are consistent
with previous reports suggesting that hepatocytes become
less responsive to epidermal growth factor stimulation with
age (28), exhibit a lowered proliferative capacity following
growth factor stimulation (29), and are less capable of tran-
scriptional activation and expression of the heat shock pro-
tein genes (9,30). On the basis of our previous work (1) and
the present data, we speculate that transcriptional activation
and expression of the AE genes may also be compromised
with age, potentially magnifying the exposure of senescent
cells to oxidants and blunting their ability to mount an ef-
fective stress response.
It should also be noted that detectable levels of the induc-
ible Hsp70 were observed in the control condition for both
young and old cells, which may suggest that hepatocytes
cultured in the Matrigel system are under a low level of
stress. Culturing primary cells is likely a stressful process,
and every attempt was made to minimize stress during each
step of the collection and plating process. Once hepatocytes
were within the Matrigel matrix, individual cells migrated
toward one another and organized themselves into complex
multicellular structures requiring cell-to-cell interaction and
communication. These observations indicate that cells re-
tained the ability to respond to the rigors of collection and
culture. In addition, Hsp70 levels, although detectable, were
still quite low in control conditions.
Figure 9. Electron micrographs of hepatocytes from senescent rats
after 3 days in culture. Day 3 cultures displayed many clusters of six
to eight hepatocytes with small lumens (SL) between adjacent hepa-
tocytes. Junctional complexes (JC) occurred between the luminal and
the lateral plasma membranes of these cells. Individual cells dis-
played abundant rough endoplasmic reticulum (RER) in lamellar ar-
rays and in association with mitochondria (M); magnification 3600ꢃ.
Insert: high-magnification gap junction (GJ) in plasma membranes
between two hepatocytes; magnification 54,000ꢃ.
mal cell function (32). The inducible Hsp70 is expressed at
low levels under putative nonstress conditions and can be
highly induced by environmental stresses such as hyperther-
mia (16,33) and exercise (33,34). Studies using hepatocytes
and splenocytes (35), alveolar macrophages (36), and fibro-
blasts (37) collected from young and old animals have
shown that aging lowers heat-inducible Hsp70 mRNA and
protein expression. The mechanism responsible for declin-
ing Hsp expression appears to involve an age-related loss of
transcriptional activation of the heat shock gene by heat
shock factor (9,35,36). In vivo studies from our laboratory
contrasting heat-inducible Hsp70 accumulation in tissues
collected from young and senescent heat-stressed Fischer
344 rats have shown that aging significantly blunts Hsp70
expression in the liver (16,33). Therefore, the present results
are consistent with data from many laboratories suggesting
that aging compromises cellular ability to upregulate stress
protein gene expression.
It is noteworthy that in addition to decrements in Hsp70
expression, senescent hepatocytes showed a marked decline
in CuZnSOD activity that appeared to occur independent of
decrements in CuZnSOD protein expression. This phenom-
enon was not observed in cultures derived from young ani-
mals, suggesting that aging may not compromise cellular
ability to express CuZnSOD, but that a portion of the pro-
tein may be inactive. A recent study by Yoo and colleagues
Multiple groups have examined the stress response in
aged animals to determine if a deficit in 70-kDa heat shock
protein synthesis could explain the loss of stress tolerance in
older populations. Heat shock proteins act as molecular
chaperones and are intimately involved in cellular protein–
protein interactions (31). Constitutive Hsc70 is ubiqui-
tously expressed throughout the cell and is essential for nor-

B266
HALL ET AL.
Figure 10. Electron micrographs of hepatocytes from senescent
rats following 5 days in culture. Large hepatocyte clusters (ꢆ10 cells)
were evident. Small lumens (SL) between cells contained microvilli;
large lumens (LL) lacked microvilli. Individual cells displayed abun-
dant rough endoplasmic reticulum (RER) often associated with mito-
chondria (M); magnification 2850ꢃ. Insert: high magnification of the
plasma membrane between adjacent hepatocytes, illustrating two
desmosomes flanking a gap junction (GJ); magnification 37,500ꢃ.
(38) reported that cellular expression of CuZnSOD and the
70-kDa heat shock proteins could be induced by heat shock,
hydrogen peroxide (H2O2), and by paraquat, an agent
known to stimulate intracellular reactive oxygen species
production. In addition, these investigators demonstrated
that both oxidative stress and heat shock response elements
on the promoters of the CuZnSOD and Hsp70 genes were
responsive to heat shock, H2O2, and paraquat. These data
suggest that CuZnSOD is a stress protein that can be regu-
lated in a manner similar to the heat shock proteins. From
the present data and our previous in vivo work (1), we pos-
tulate that in addition to lowering cellular capacity to ex-
press heat shock protein genes, aging also reduces cellular
capacity to increase CuZnSOD activity and thus antioxidant
capacity.
In summary, the main findings of the present investiga-
tion establish that primary hepatocytes isolated from young
and senescent animals and cultured with Matrigel adopt a
differentiated phenotype and maintain the ability to express
AEs. Although our results suggest that Matrigel may pro-
vide primary hepatocytes with an environment that pro-
motes cellular organization closer to that observed in vivo,
changes in cellular AE and heat shock protein expression
were evident at time points after 7 days ex vivo. Accord-
ingly, we can conclude that the Matrigel system minimizes,
but does not eliminate, environmental factors that lead to al-
tered gene expression in cultured cells. However, it is im-
portant to note that both AE and heat shock protein expres-
sion were maintained at or near control levels for the first 7
days ex vivo, suggesting that an adequate window of time
exists for interventional studies using this preparation. With
this outcome in mind, we find that the present data show
that cells derived from young animals maintained steady-
state AE expression at or near control levels longer than
cells derived from senescent rats. Moreover, 70-kDa heat
Figure 11. Electron micrographs of hepatocytes from senescent
rats following 14 days in culture. This figure depicts two adjacent
clusters of hepatocytes containing large lumens (LL) devoid of mi-
crovilli and small lumens (SL) with microvilli; junctional complexes
(JC) bordered the small lumens. Autophagic vacuoles (AV) were
large and numerous in many of the cells, but the cells still possessed
lamellar rough endoplasmic reticulum (RER) and numerous mito-
chondria; magnification 2700ꢃ. Insert: high magnification of a gap
junction (GJ) in day 14 culture; magnification 38,000ꢃ.
shock protein expression is relatively well maintained in
hepatocyte cultures from young rats, whereas cultures de-
rived from senescent animals rapidly lose the ability to ex-
press the 70-kDa heat shock proteins. These results are con-
sistent with the hypothesis that age-related deficiencies in
cellular AE and heat shock protein regulation contribute to
declining stress tolerance in the elderly population.
Acknowledgments
This research was supported by National Institutes of Health Grants AG-
12350, AG-14687, CA-66081, CA-81090, and T32 AG-00214 (Interdisci-
plinary Research Training Program on Aging Fellowship to D.M. Hall).
The authors gratefully acknowledge the technical support of Victoria J.
Drake, Linjing Xu, and Alana T. Brennan, and the kind gift of Biocoat re-
agents from Becton Dickinson Labware.
Address correspondence to Kevin C. Kregel, PhD, Department of Exer-
cise Science, 532 FH, The University of Iowa, Iowa City, IA 52242.
E-mail: kevin-kregel@uiowa.edu
References
1. Hall DM, Oberley TD, Moseley PL, et al. Caloric restriction improves
thermotolerance and reduces hyperthermia-induced cellular damage in
old rats. FASEB J. 2000;14:78–86.

AGING, ANTIOXIDANT ENZYMES, AND PRIMARY CELLS
B267
2. Kregel KC, Moseley PL, Skidmore R, Gutierrez JA, Guerriero V.
HSP70 accumulation in tissues of heat-stressed rats is blunted with ad-
vancing age. J Appl Physiol. 1995;79:1673–1678.
3. Seals DR, Taylor JA, Ng AV, Esler MD. Exercise and aging: auto-
nomic control of the circulation. Med Sci Sports Exerc. 1994;26:568–
576.
4. Bender BS, Taylor SF, Zander DS, Cottey R. Pulmonary immune re-
sponse of young and aged mice after influenza challenge. J Lab Clin
Med. 1995;126:169–177.
5. Fawcell TW, Sylvester SL, Sarge KD, Morimoto RI, Holbrook NJ. Ef-
fects of neurohumoral stress and aging on the activation of mammalian
heat shock factor 1. J Biol Chem. 1994;269:32272–32278.
6. Pahlavani MA, Harris MD, Moore SA, Weindruch R, Richardson A.
The expression of heat shock protein 70 decreases with age in lympho-
cytes from rats and rhesus monkeys. Exp Cell Res. 1995;218:1–9.
7. Ames B, Shinenaga MK, Hagen TM. Oxidants, antioxidants, and the
degenerative diseases of aging. Proc Natl Acad Sci USA. 1993;90:
7915–7922.
8. Jones DP, Carlson JL, Mody VC, Cai J, Lynn MJ, Sternberg P. Redox
state of glutathione in human plasma. Free Radic Biol Med. 2000;28:
625–635.
9. Heydari AR, Wu B, Takahashai R, Strong R, Richardson A. Expres-
sion of heat shock protein 70 is altered by age and diet at the level of
transcription. Mol Cell Biol. 1993;13:2909–2918.
22. Rana B, Mischoulon D, Xie Y, Bucher NL, Famer SR. Cell-extracellu-
lar matrix interactions can regulate the switch between growth and
differentiation in rat hepatocytes: reciprocal expression of C/EBP al-
pha and immediate-early growth response transcription factors. Mol
Cell Biol. 1994;14:5858–5869.
23. Seglen PO. Preparation of isolated rat liver cells. Methods Cell Biol.
1976;13:29–83.
24. Kreamer BL, Staecker JL, Sawanda N, Sattler GL, Hsia MTS, Pitot
HC. Use of low-speed, isodensity Percoll centrifugation method to in-
crease the viability of isolated rat hepatocyte preparations. In Vitro
Cell Dev Biol. 1986;22:201–211.
25. Laemmli UK. Cleavage of structural proteins during the assembly of
the head of bacteriophage T4. Nature (London). 1970;227:680–685.
26. Beauchamp C, Fridovich I. Superoxide dismutase: improved assays
and an assay applicable to acrylamide gels. Anal Biochem. 1971;44:
276–287.
27. Sun Y, Elwell JH, Oberley LW. A simultaneous visualization of the
antioxidant enzymes glutathione peroxide and catalase on polyacryla-
mide gels. Free Rad Res Commun. 1988;5:67–75.
28. Liu Y, Guyton KZ, Gorospe M, et al. Age-related decline in mitogen-
activated protein kinase activity in epidermal growth factor-stimulated
rat hepatocytes. J Biol Chem. 1996;271:3604–3607.
29. Mitaka T, Norioka K, Sattler GL, Pitot HC, Mochizuki Y. Effect of
age on the formation of small-cell colonies in cultures of primary rat
hepatocytes. Cancer Res. 1993;53:3145–3148.
10. Papaconstantinou J. Unifying model of the programmed (intrinsic) and
stochastic (extrinsic) theories of aging. Ann NY Acad Sci. 1994;719:
195–211.
30. Fargnoli J, Kunisada T, Fornace AJ, Schneider EL, Holbrook NJ. De-
creased expression of heat shock protein 70 mRNA and protein after
heat treatment in cells of aged rats. Proc Natl Acad Sci USA. 1990;87:
846–850.
31. Welch WJ. Mammalian stress response: cell physiology, structure/
function of stress proteins, and implications for medicine and disease.
Physiol Rev. 1992;72:1063–1081.
32. Parsell DA, Lindquist S. The function of heat shock proteins in stress
tolerance: degradation and reactivation of damaged proteins. Annu Rev
Genet. 1993;27:437–496.
11. Li J-J, Oberley LW. Overexpression of manganese-containing super-
oxide dismutase confers resistance to the cytotoxicity of tumor necrosis
factor ꢇ and/or hyperthermia. Cancer Res. 1997;57:1991–1998.
12. Oberley LW, Buettner GR. Role of antioxidant enzymes in cell immor-
talization and transformation. Mol Cell Biochem. 1988;84:147–153.
13. Guyton KZ, Gorospe M, Wang X, et al. Age-related changes in activa-
tion of mitogen-activated protein kinase cascades by oxidative stress. J
Invest Dermatol. 1998;3:23–27.
14. Jamieson DJ, Storz G. Transcriptional regulators of oxidative stress re-
sponses. In: Scandalios JG, ed. Oxidative Stress and the Molecular
Biology of Antioxidant Defenses. Plainview, NY: Cold Spring Harbor
Laboratory Press;1997:91–115.
15. Dalton TP, Shertzer HG, Puga A. Regulation of gene expression by re-
active oxygen. Annu Rev Toxicol. 1999;39:67–101.
16. Hall DM, Xu L, Drake VJ, et al. Aging reduces adaptive capacity and
stress protein expression in the liver after heat stress. J Appl Physiol.
2000;89:749–759.
33. Kregel KC, Moseley PL. Differential effects of exercise and heat
stress on liver HSP70 accumulation with aging. J Appl Physiol. 1996;
80:262–277.
34. Locke M, Noble EG, Atkinson BG. Exercising mammals synthesize
stress proteins. Am J Physiol. 1990;258:C723–C729.
35. Gutsmann-Conrad A, Pahlavani MA, Heydari AR, Richardson A. Ex-
pression of heat shock protein 70 decreases with age in hepatocytes
and splenocytes from female rats. Mech Ageing Develop. 1999;107:
255–270.
17. Clayton DF, Darnell JE. Changes in liver-specific compared to com-
mon gene transcription during primary culture of mouse hepatocytes.
Mol Cell Biol. 1983;3:1552–1561.
18. Kocarek TA, Schuetz EG, Guzelian PS. Expression of multiple forms
of cytochrome P450 mRNAs in primary cultures of rat hepatocytes
maintained on Matrigel. Mol Pharmacol. 1994;43:328–334.
19. Schuetz EG, Li D, Omiecinski CJ, et al. Regulation of gene expression
in adult rat hepatocytes cultured on a basement membrane matrix. J
Cell Physiol. 1988;134:309–323.
20. Van Remmen H, Williams MD, Heydari AR, et al. Expression of
genes coding for antioxidant enzymes and heat shock proteins is al-
tered in primary cultures of rat hepatocytes. J Cell Physiol. 1996;166:
453–460.
36. Moore SA, Lopez A, Richardson A, Pahlavani MA. Effect of age and
dietary restriction on expression of heat shock protein 70 in rat alveo-
lar macrophages. Mech Ageing Develop. 1998;104:59–73.
37. Liu A, Lin Z, Choi H, Sorhage F, Li B. Attenuated induction of heat
shock gene expression in aging diploid fibroblasts. J Biol Chem. 1989;
264:12037–12045.
38. Yoo HY, Chang MS, Ro HM. The activation of the rat copper/zinc su-
peroxide dismutase gene by hydrogen peroxide through the hydrogen
peroxide response element and by paraquat and heat shock through the
same element. J Biol Chem. 1999;274:23887–23892.
21. Flaherty PM. Expression of cytochrome P450 activity in primary cul-
tures of rat hepatocytes maintained on matrigel. Mol Cell Biol. 1994;
S177s.
Received July 7, 2000
Accepted January 12, 2001
Decision Editor: John A. Faulkner, PhD