Vol. 283, Issue 4, H1322-H1333, October 2002
HSC73-tubulin complex formation during low-flow
ischemia in the canine myocardium
Robert S.
Decker1,2,3,
Marlene L.
Decker1,
Sakie
Nakamura1,
Yu-Sheng
Zhao1,
Sascha
Hedjbeli1,
Kathleen R.
Harris1, and
Francis J.
Klocke1,2
1 Feinberg Cardiovascular Research Institute and
Departments of 2 Medicine and 3 Cell
and Molecular Biology, Northwestern University, The Feinberg School
of Medicine, Chicago, Illinois 60611-3008
 |
ABSTRACT |
Canine myocardium was
exposed to bouts of low-flow ischemia to identify the
interactions that develop between the microtubule-based cytoskeleton
and the heat shock protein 70 (HSP70) family of heat shock proteins in
viable cardiomyocytes. "Moderate" or "severe" low-flow
ischemia was produced in chronically instrumented dogs by
reducing circumflex coronary flow by 50% for 2 h or by 75% for
5 h followed by reperfusion for 2 and 24 h, respectively. Electron and immunofluorescence microscopy demonstrated either partial
or nearly complete depolymerization of the intermyofibrillar microtubules in areas of myofibril disruption and partial dissolution of the perinuclear microtubule girdle. In contrast, centrosomal tubulin
arrays appeared to remain intact following low-flow ischemia. In cardiomyocytes displaying myofibril disruption, constitutively expressed HSP73 (HSC73) colocalized with intact but not disrupted microtubules and with perinuclear and centrosomal tubulin following moderate ischemia. Microtubule depolymerization and high
molecular weight tubulin-HSC73 complexes were present in more severely
ischemic tissue. These results suggest that HSC73 directly
interacts with tubulin and may protect selected elements of the
microtubule network and limit myofibril disruption during reversible
low-flow ischemia.
microtubule network; myofibril disruption; contractile
dysfunction
| |
INTRODUCTION |
THE
MICROTUBULE-BASED cytoskeletal network has long been theorized to
function as a scaffold that stabilizes the contractile apparatus of the
cardiac myocyte (10, 20, 25). Observations from
Kamada's laboratory (13, 21) illustrated that
microtubules depolymerize in ischemically injured and stunned
canine myocardium. Armstrong and Ganote (2) have reported
anomalies in tubulin organization in metabolically damaged adult
myocytes, and microtubule depolymerization has been observed in
"ischemically" injured neonatal myocytes (5).
The overexpression of members of the heat shock protein 70 (HSP70)
family stress proteins and the "small" class of heat shock proteins
has been reported to stabilize the microtubule network and
"protect" cultured neonatal heart cells from ischemic injury (5, 8, 18). Such observations have suggested that microtubules may represent pivotal cytoskeletal elements that stabilize
the myofibril and that specific classes of heat shock proteins may
preserve this relationship.
The present study has focused on identifying protein-to-protein
interactions that develop between tubulin and the HSP70 family of heat
shock proteins (11, 19, 26), because the expression of
these stress proteins was modestly elevated in short-term hibernating and stunned canine myocardium (24). In this experimental
paradigm, myofibrillar thick filaments have been reported to
disassemble from sarcomeres in subendocardial cardiomyocytes of animals
subjected to bouts of low-flow ischemia and reperfusion
(24). This observation raised the question of whether the
enhanced expression of heat shock proteins influences the integrity of
the cardiocyte's contractile apparatus/cytoskeleton during and after
low-flow ischemia. The strategy employed to address this query
utilized a chronically instrumented canine model of low-flow
ischemia in which coronary flow was either reduced moderately
so that perfusion-contraction matching could be maintained or was
restricted more severely so that perfusion-contraction mismatching
would develop (23, 24). Moderate low-flow ischemia
induced a short-term hibernating phenotype during reperfusion, whereas
severely reducing coronary flow provoked myocardial stunning on
reperfusion (23). The objective of these experiments was
to determine whether tubulin complexed with constitutively expressed
HSC73 or inducible HSP70 during and after low-flow ischemia and
whether such changes paralleled myofibril disruption in a model in
which irreversible cardiocyte necrosis was not apparent (24).
| |
MATERIALS AND METHODS |
Experimental Animal Preparation
Experiments were conducted with mongrel dogs of both sexes using
protocols that conform to the principles in the Guide for the
Care and Use of Laboratory Animals published by the National Institutes of Health (NIH Publication No. 85-23, 1996). Experimental procedures also were approved by Northwestern University's
Institutional Animal Care and Use Committee. Animals weighing
23-36 kg were instrumented following an overnight fast and a 3-wk
period of on-site conditioning. Briefly, all animals were surgically
instrumented with a Konigsberg micromanometer, ultrasonic volumetric
transit-time flow probes, a hydraulic occluder, and downstream
catheters to measure pressure. Regional subendocardial segment
shortening and/or wall thickening was monitored by placing ultrasonic
crystal pairs in the central portion of the distributions of the left
circumflex (LCX) and left anterior descending (LAD) vascular beds.
Seven to ten days after surgery, the dogs were lightly sedated with Innovar-Vet (fentanyl, 0.4 mg/ml and droperidol, 20 mg/ml) and placed
upright in a sling to which they had been previously acclimated. Regional coronary flow in the LCX (i.e., test region) was gradually reduced by injecting saline into the hydraulic occluder until the
desired reduction in coronary flow and segmental function was achieved.
The LAD region of the left ventricle, where neither coronary flow nor
function was altered, represented the remote myocardial site and was
examined in parallel with the paired LCX test area. Details of the
instrumentation procedure have been fully reported previously
(23, 24).
In this investigation, 10 animals were subjected to a moderate
reduction in coronary flow of 50% for 2 h that resulted in a 50%
decrease in regional wall function. Five animals were euthanized after
low-flow ischemia, whereas the other five were reperfused for
an additional 2 h before being euthanized. A second group of 12 dogs was exposed to a more severe, sustained 5-h reduction in regional
coronary flow sufficient to reduce regional segment function by
~75%. Five animals were euthanized after low-flow ischemia,
and seven dogs were reperfused for 24 h. Four additional fully
instrumented dogs were conditioned and hemodynamic data collected, but
coronary flow was not reduced; these animals represented sham-instrumented, control preparations. All animals were brought to
the laboratory for final physiological measurements and then euthanized
with an overdose of pentobarbital sodium and potassium chloride. Hearts
from the euthanized animals were biopsied in situ, and test
and remote tissue specimens were fixed immediately in glutaraldehyde
(24). The hearts were then rapidly removed and placed in
ice-cold isotonic saline. Additional test, remote, and sham samples
were either quickly frozen in liquid nitrogen, preserved in a
paraformaldehyde-lysine-periodate fixative, or processed immediately
for the isolation of HSC73/HSP70-tubulin complexes.
Immunofluoresence Microscopy
The distribution of tubulin isoforms was examined in 4-µm
frozen sections of tissue that had been fixed in 4% paraformaldehyde, 100 mmol/l L-lysine, and 10 mmol/l sodium periodate
prepared in PBS (pH 7.4) at the time of euthanasia (21).
The sections were blocked with a 1:10 dilution of normal goat serum in
PBS for 4 h at room temperature to minimize nonspecific adsorption
of primary antibodies. 
-, 
-, and 
-tubulin distribution was
then monitored in frozen sections by using mouse monoclonal antibodies
(Sigma Chemical; St. Louis, MO). Nonimmune mouse IgG1 was employed to control for the nonspecific binding of the primary anti-tubulin antibodies. The antibodies were diluted to a final concentration of 1 µg/ml with PBS supplemented with 1 mg/ml of BSA, and the sections
were incubated with each of the primary antibodies overnight at 4°C
before being washed in PBS (3 × for 30 min). Sections were then
stained for 1 h at 37°C with a secondary, affinity-purified, goat anti-mouse IgG (Cappel, ICN; Aurora, OH) labeled with either FITC
or tetramethyl rhodamine that was diluted 1:300 in PBS plus BSA. After
being washed in PBS-BSA, the sections were mounted with gelvatol and Z
sectioned with a Zeiss 510 laser scanning confocal microscope.
Some sections were doubly stained with affinity-purified, goat
polyclonal antibodies (1 µg/ml) produced against highly conserved peptide sequences derived from human HSC73 and HSP70 (Santa Cruz Biotechnology; Santa Cruz, CA). Affinity-purified rabbit anti-goat IgG
labeled with FITC (Cappel, ICN) was employed to visualize the
distribution of the stress proteins. To monitor the specificity of
these antipeptide antibodies, sections were incubated with either
nonimmune goat IgG1 or supplemented with 100-fold excess of a blocking
peptide that was used to produce the polyclonal anti-HSC73 or HSP70
antibodies (Santa Cruz Biotechnology).
In an effort to quantify the degree of heat shock protein
redistribution to the microtubule network, the fluorescence emission spectra of HSC73 and HSP70 colocalized to microtubules was measured after introduction of an electronic raster across the long axis of a
cardiomyocyte labeled for tubulin and each of the stress proteins. The
fluorescence emission intensity was tabulated from the tubulin window
(i.e., rhodamine channel) and from the stress protein window (i.e.,
fluorescein channel), and a ratio of the two fluorescence emission
intensities (in arbitrary units) was calculated at sites where the
raster intersected with a microtubule. The ratio (rhodamine/fluorescein
intensity) was expressed per microtubule (means ± SE) and was
derived from 20 myocytes from each experiment. Significant differences
in the fluorescence emission ratios were determined using either a one-
or two-way ANOVA followed by post hoc analysis with the
Student-Newman-Keuls test; a P < 0.05 was considered
statistically significant.
Transmission and Immunogold Electron Microscopy
Biopsied samples of test, remote, and sham subendocardium were
preserved in either 2% glutaraldehyde buffered in sodium cacodylate (pH 7.4) or in the formaldehyde-lysine-periodate (FLP) fixative used
for immunofluorescence microscopy. Glutaraldehyde-fixed tissue was
postfixed in 2% osmium tetroxide, dehydrated, and embedded in an epoxy
resin as outlined previously (24) and viewed with a JOEL
100CX electron microscope. The ultrastructural distribution of tubulin
was monitored in FLP-fixed frozen sections using a postembedding
technique. Washed, free-floating 4-µm sections were incubated in a
monoclonal anti--tubulin antibody (Sigma Chemical) diluted 1:100 in
PBS overnight at 4°C. The sections were then washed in PBS and
incubated overnight at 4°C in a 1:20 solution of donkey anti-mouse
IgG1 conjugated to 6-nm colloidal gold particles (Jackson Laboratories;
West Cove, PA). The sections were then processed as described above.
Biochemical Distribution of Tubulin and HSC73/HSP70 Stress
Proteins
Relative changes in the total content of tubulin isoforms and
HSC73/HSP70 that developed during low-flow ischemia and
reperfusion were obtained from frozen sham, test, and remote
subendocardium (200 mg) that was solubilized in five volumes of Laemmli
sample lysis buffer [200 mmol/l DTT, 4% SDS, 160 mmol/l
Tris · HCl (pH 6.8), and 20% glycerol] and heated at 95°C
for 5 min. Other paired samples were solubilized in Laemmli buffer for
15 min at 37°C. The solubilized tissue extracts were clarified by
centrifugation, and the protein concentration of the supernatants was
determined using the Pierce microprotein assay (Pierce; Rockford, IL).
Equal amounts of protein (1 µg/lane) were loaded on 10% SDS-PAGE
minislab gels, and tubulin isoforms and HSC73/HSP70 were separated from one another. Protein was then transferred to a polyvinylidene fluoride
membrane for Western blotting. Sample loading and transfer efficiency
was evaluated by Coomassie blue brilliant staining. Membranes were
blocked in TBST solution [150 mmol/l NaCl, 10 mmol/l Tris · HCl (pH 8.0), 0.05% Tween 20, and 3% nonfat dry milk]
and incubated in a solution of tubulin antibody (1:5,000 dilution of
-, -, and -tubulin monoclonal mouse IgG1;Sigma Chemical) and
in a 1:2,500 dilution of affinity-purified anti-goat polyclonal HSC73/HSP70 antibodies (Santa Cruz Biotechnology). The
membranes were then incubated with a secondary goat anti-mouse/rabbit
anti-goat IgG conjugated to horseradish peroxidase, and the Western
blots were visualized by reacting the blotted membranes with
Supersignal West Pico Luminal Enhancer (Pierce) followed by
autoradiography. Metamorph image software was employed to digitize the
gel bands, and the protein content of each myocardial sample was
evaluated by using purified tubulin isoforms and HSC73 and HSP70
standards (i.e., 200 ng-2 µg concentration range) obtained from
Sigma, Stressgen (Victoria, BC, Canada), or Santa Cruz, respectively.
The amount of each protein was expressed in micrograms of tubulin,
HSC73, or HSP70 per milligram soluble protein (means ± SE).
Protein contents were compared with values derived from
sham-instrumented myocardium using a one- or two-way ANOVA followed by
post hoc testing with the Student-Newman-Keuls test, with a
P < 0.05 being considered statistically significant.
Tubulin-HSC73/HSP70 Interactions
Putative interactions between tubulin and HSC73/HSP70 were
identified in sarcoplasmic and myofibrillar fractions prepared after
homogenizing 100 mg of fresh myocardial tissue in 10 volumes of
low-salt buffer (40 mmol/l NaCl, 1 mmol/l DTT, 0.1 mmol/l EGTA, and
0.1% Triton X-100, pH 7.4) with a Polytron PC-1. The homogenate was
centrifuged at 11,000 g for 30 min at 4°C, and the
supernatant was saved as the sarcoplasmic fraction, whereas the pellet
(myofibril fraction) was washed five times in ice-cold, low-salt
buffer. Sham, test, and remote sarcoplasmic and myofibril fractions
were diluted/homogenized with 0.5 ml of extraction buffer [50 mmol/l Tris · HCl (pH 8.5), 150 mmol/l NaCl, 20 mmol/l EDTA, 1%
Triton X-100, 4 mol/l urea, and 0.5 mmol/l PMSF], and both extracts
were clarified by centrifugation. Aliquots of the myofibril and
sarcoplasmic fractions were layered over a 10-40% linear sucrose
gradient and centrifuged for 16 h at 100,000 g in a
Beckman L8-80M Ultracentrifuge at 4°C. The gradients were
calibrated by using thyroglobulin (669 kDa) and catalase (232 kDa) as
molecular mass standards (Sigma Chemical). Each fraction also was
assayed qualitatively for its protein composition by immunodot blotting
with antibodies directed against actin, desmin, myosin, and tubulin in
addition to HSC73/HSP70, HSP27, and B-crystallin. In those
fractions where stress and cytoskeletal proteins were detected,
aliquots of each fraction were loaded on to a 4% native PAGE gel and
electrophoresed in the cold; thyroglobulin and catalase (Sigma
Chemical) standards were run in parallel. Similarly, samples of each
fraction also were subjected to denaturing gel electrophoresis in 10%
SDS-PAGE as described above.
Characterization of Tubulin-Stress Protein Complexes
In those fractions that disclosed high-molecular-weight
complexes containing tubulin and/or HSC73/HSP70, anti-HSC73/HSP70 antibodies were employed to immunoprecipitate those protein(s) from the
suspected complexes. Aliquots of protein (100 µg) derived from each
of the sucrose gradient fractions were mixed with 10 µg of
HSC73/HSP70 antibody (clone 92-mouse IgG1, StressGen). This antibody
recognizes both the constitutive (HSC73) and inducible (HSP70) forms of
the stress protein with the same degree of affinity. Protein G-agarose
(20 µl, Santa Cruz Biotechnology) was added to the incubation mixture
and incubated for an additional 30 min at 4°C with gentle agitation.
The solution was then centrifuged, and the pellet was solubilized in
Laemmli buffer at 95°C and loaded on a 10% SDS-PAGE gel and
immunoblotted as described above. The supernatant, known amounts (i.e.,
100 ng-1 µg) of purified tubulin and HSP70, and molecular weight
standards were run in adjacent lanes with the immunoprecipitates. The
amount of tubulin and HSC73/HSP70 in the immunoprecipitates was
estimated from a standard curve created with the purified proteins, and
tubulin and stress protein content were expressed as micrograms tubulin
and/or HSC73/HSP70 per 100 µg of total fraction protein (means ± SE). Significant differences were determined using a one-way ANOVA
followed by post hoc testing with the Student-Newman-Keuls test, with a
P < 0.05 being significant.
| |
RESULTS |
Hemodynamic Parameters During Low-Flow Ischemia
When coronary blood flow was reduced moderately for 2 h, both
flow and regional systolic function (i.e., segmental shortening or wall
thickening) decreased in parallel. Segmental function in LCX region was
reduced on average 50 ± 2%, and coronary flow decreased 54 ± 3% compared with baseline values (Table
1). Coronary flow and segmental
function in the remote, left ventricle (LAD) remained at values
observed before occlusion. In animals reperfused for 2 h, coronary
flow returned to 86 ± 4% of baseline flow, and segment function
measured 79 ± 2%, indicating that the LCX (i.e., test)
subendocardium displayed a short-term hibernating phenotype (23,
24). Conversely, severely reducing coronary flow for 5 h
depressed segmental function 76 ± 2% and reduced coronary flow
about 73 ± 5% in the LCX bed (Table 1), whereas flow and function continued to remain unperturbed in the remote LAD myocardium. When the dogs were reperfused for 24 h, flow returned to values seen before occlusion (i.e., 109 ± 6% of baseline), whereas wall function remained markedly depressed (54 ± 8% of the
preocclusion baseline segment shortening/wall thickening; see Table 1).
The flow-function mismatch was indicative of the development of
myocardial stunning (23). The hemodynamic parameters in
the remote left ventricle supplied by the LAD were not altered
following the severe partial coronary occlusion.
Distribution of Tubulin and HSC73/HSP70 in Normal Cardiomyocytes
The microtubule network assumes a rather complex
distribution in the heart cell, forming a densely woven basket of 25-nm
tubules surrounding each nucleus and a somewhat less extensively
arrayed sheath of tubules enveloping each myofibril (Fig.
1a). A better appreciation of
tubulin distribution between neighboring myofibrils was achieved by
employing immunogold electron microscopy. Anti-tubulin-conjugated gold
particles decorated the 25-nm microtubules that weave between adjacent
myofibrils and parallel to the Z line (Fig. 1d). - and/or -Tubulin-positive perinuclear arrays also were readily visible in
most cardiocytes (Fig. 1a); moreover, -tubulin
also could be observed at the center of these sites (Fig.
2a). Many of the canine
cardiocytes appeared to disclose multiple perinuclear centrosomal sites
(Fig. 1a), and fine structural observations confirmed that as many as three centrioles could be visualized in these perinuclear areas (Fig. 2, b and c). In subendocardial
myocytes derived from sham-instrumented animals, HSC73 was distributed
uniformly throughout the sarcoplasm (Fig. 1b). Anti-HSC73
antibody also stained the central areas of these perinuclear
tubulin-positive arrays (Fig. 1, b and c,
arrowheads) and weakly labeled some intermyofibrillar microtubules
(Fig. 1, a and c). In contrast, there was no
evidence that HSP70 stained either intermyofibrillar or perinuclear
microtubules, and little HSP70 colocalization of the centrosomes could
be observed in "normal" cardiocytes (data not shown).
View larger version (183K):
[in this window]
[in a new window]
|
Fig. 1.
Distribution of -tubulin (a) and constitutive heat
shock protein (HSC73) (b) in subendocardial myocytes derived
from a sham-instrumented animal. a: -Tubulin antibody
staining of the perinuclear (N) girdle, arrays (arrowheads), and
intermyofibrillar microtubules (arrows). HSC73 is diffusely distributed
throughout the sarcoplasm and in the perinuclear (N) arrays
(b, arrowheads). HSC73 and -tubulin colocalize to the
perinuclear arrays (c, arrowheads) and weakly stain some
peripheral microtubules. d: Immunogold (arrows) labeling
reveals a complex pattern of -tubulin-positive microtubules between
neighboring myofibrils (M); some Z lines (arrowheads) also display
gold-labeled tubulin. m, Mitochondrion; a-c,
bar = 25 µm; d, bar= 0.1 µm.
|
|
View larger version (210K):
[in this window]
[in a new window]
|
Fig. 2.
a: -Tubulin localization is restricted to
microtubule-organizing centers of normal heart cells. Punctate
-tubulin sites (arrows) are observed at each pole of the myocyte
nucleus (N). b and c: as many as three centrioles
(arrows) have been identified in these perinuclear (N) areas. G, Golgi
apparatus; a, bar = 2 µm; b and
c, bar = 1 µm.
|
|
Integrity of Microtubules During Low-Flow
Ischemia-Reperfusion
Moderate low-flow ischemia.
When coronary flow and function were reduced moderately (in the LCX
test area), the perinuclear microtubule girdle and microtubular arrays
remained intact (Fig. 3a);
however, areas within individual myocytes revealed partially disrupted
intermyofibrillar microtubules (Fig. 3a). In preparations
double labeled for tubulin and HSC73, the structurally intact,
peripheral intermyofibrillar microtubules stained intensely for the
presence of HSC73 (Fig. 3, b and c) but not HSP70
(data not shown). In those regions of the cardiocyte that displayed
poorly ordered (depolymerized/disrupted) microtubules or punctate
deposits of tubulin, few of the recognizable tubulin-positive elements
stained for HSC73 (Fig. 3, b and c) or HSP70
(data not shown). Anti-HSC73 also weakly stained cardiocyte nuclei
following low-flow ischemia (Fig. 3b). Biopsied
samples derived from remote, normally perfused myocardium showed only a
minimal redistribution of HSC73 (Fig.
4b) to cardiocyte
intermyofibrillar microtubules (Fig. 4a). Only the
centrosomes and the proximal portions of microtubules radiating away
from these structures displayed HSC73 immunoreactivity (Fig.
4c). Only small amounts of HSP70 were expressed in the
canine myocardium, and no redistribution of the inducible form of the stress protein was observed in the cardiomyocytes following a moderate
reduction of coronary flow in the LCX or in the normally perfused (LAD
area) myocardium (data not shown).
View larger version (113K):
[in this window]
[in a new window]
|
Fig. 3.
-Tubulin (a) and HSC73 (b)
redistribution in subendocardial myocytes after 2 h of low-flow
ischemia. Both intact intermyofibrillar (arrowhead) and
disrupted (*) microtubules are present in the cardiocyte
(a). HSC73 stains intact microtubules (arrowhead,
b and c) but not disrupted microtubules (*,
a-c). Perinuclear girdle, nucleoplasm, and
tubulin arrays (arrows) also display enhanced HSC73 staining
(b and c). Note intensely stained endothelial
cell (e). N, nucleus; a-c, bar = 2 µm.
|
|
View larger version (149K):
[in this window]
[in a new window]
|
Fig. 4.
-Tubulin (a) and HSC73 (b)
distribution in the remote subendocardial myocytes after 2 h
of low-flow ischemia. HSC73 stains few of intermyofibrillar
microtubules or the perinuclear (N) girdle (a and
b); instead HSC73 colocalizes with centrosomes
(c, arrows) or remains diffusely distributed throughout the
sarcoplasm (*). Note that HSC73-positive microtubules radiate from the
centrosomes (arrows, a-c);
a-c, bar = 2 µm.
|
|
When the hearts were reperfused, there was little change in the
distribution of tubulin in subendocardial cardiocytes. However, intermyofibrillar microtubules stained less intensively for HSC73 than
did their counterparts from the ischemic myocardium (Table 2). Sites of depolymerized
intermyofibrillar microtubules could be observed in cardiocytes from
reperfused hearts although less frequently than in tissue sampled
immediately following low-flow ischemia. In the remote regions
of the left ventricle, no changes in the distribution of tubulin were
apparent, and the pattern of HSC73 staining resembled that observed
before reperfusion (data not shown).
Severe low-flow ischemia.
When coronary flow was reduced more severely for a
prolonged period, intermyofibrillar microtubules then appeared to
completely depolymerize in cardiocytes revealing myofibril
disruption. Intermyofibrillar microtubules appeared to be replaced by a
punctate tubulin-staining pattern (Fig.
5a) that was somewhat
reminiscent of the particulate tubulin pattern reported in
ischemic/ stunned canine cardiac myocytes (13,
21). The punctate tubulin aggregates generally lacked evidence
of stress protein labeling as well (Fig. 5, b and
c). Portions of the perinuclear girdle also appeared
disrupted (Fig. 5, a and d); however, the
remaining intact elements of the girdle stained positively for HSC73
(Fig. 5b). Intranuclear localization of HSC73 also was
observed in cardiocytes disclosing intermyofibrillar microtubule
depolymerization; moreover, HSC73 appeared closely associated with the
inner aspect of the nuclear envelope directly adjacent to the remaining
intact, HSC73-positive portions of the microtubule girdle (Fig. 5,
b and d). Whereas the centrosomes remained
structurally intact following severe low-flow ischemia (Fig. 5,
a, c, and d), no microtubules were
observed radiating away from them (Fig. 5a; compare with
Figs. 3a and 4a); instead, only "blunt-ended"
structures projected from some of the centrosomes (Fig. 5a).
Even though coronary flow was not compromised in the remote
myocardium, some remote cardiac myocytes also displayed microtubule depolymerization, whereas neighboring cardiocytes disclosed
a relatively intact intermyofibrillar microtubule network (Fig.
6a) that stained for both
HSC73 (Table 2) and HSP70 (Fig. 6, b and c).
View larger version (175K):
[in this window]
[in a new window]
|
Fig. 5.
-Tubulin (a) and HSC73 (b) distribution
after 5 h of low-flow ischemia. a:
depolymerized intermyofibrillar microtubules (*, a).
Portions of perinuclear tubulin-positive girdle and centrosomal arrays
remain intact (arrows, a). Nuclei (N), portions of the
perinuclear girdle, and arrays (arrows) disclose HSC73
staining (b and c). Electron microscopy reveals
intact centrioles (arrowhead) in perinuclear regions of myocytes with
disrupted myofibrillar thick filaments (arrows) and condensed nuclear
chromatin (d). a-c, Bar: 2 µm;
d, bar: 1 µm.
|
|
View larger version (153K):
[in this window]
[in a new window]
|
Fig. 6.
Remote subendocardial myocytes derived from 5-h low-flow
ischemic tissue. -Tubulin (a) and heat shock
protein 70 (HSP70, b) colocalize to the microtubules
(arrowheads), the perinuclear (N) girdle, and centrosomes (arrows);
other myocytes display few intact microtubules (*, a) and no
tubulin-HSP70 colocalization (*, c); a-c,
bar = 5 µm.
|
|
Intermyofibrillar microtubules continued to remain depolymerized
following 24 h of reperfusion, especially in severely stunned myocardium (Table 2). Likewise, the perinuclear centrosomal arrays remained intensely labeled with HSC73; however, no microtubules radiated away from these structures, implying little microtubule reassembly transpires in these "stunned" cardiocytes (data not shown). In remote regions of the myocardium, the organization of the
microtubule network was not influenced by reperfusion with the
exception that intermyofibrillar microtubules and cardiocyte nuclei
were labeled with HSP70 (Fig. 6b) as well as HSC73 (Table 2).
Tubulin-HSC73/HSP70 Interactions in Canine Myocardium
Distribution of tubulin and HSC73/HSP70.
Western immunoblotting was used to monitor the fate of tubulin and the
heat shock proteins in test and remote myocardial tissue samples
following low-flow ischemia. Neither tubulin nor HSC73/HSP70 could be detected in the low-salt, partially purified myofibrillar fraction; instead, both protein families were present in the soluble, sarcoplasmic fraction of myocardial homogenates, regardless of the
degree or duration of low-flow ischemia-reperfusion. When the
sarcoplasmic extracts were diluted into Laemmli buffer and solubilized
at 95°C, the relative content of -tubulin was reduced significantly (32.7 ± 6.9%, P < 0.05, n = 5) in the test regions (Fig.
7A) compared with extracts
prepared from either remote (Fig. 7A) or sham myocardium
(Fig. 7A). -Tubulin content declined somewhat more
modestly (17.6 ± 3.5%, P < 0.1, n = 5) in extracts of test tissue (Fig. 7A).
Although no significant change in total HSC73 content could be detected
in gels scanned from test, remote, or sham tissue, the amount of the
constitutively expressed HSC73 (6.7 ± 1.6 µg/mg total protein,
n = 5) was 9.3-fold greater than HSP70 (0.7 ± 0.3 µg/mg total protein, n = 5), the inducible form of
the stress protein (P < 0.05), regardless of the
changes in coronary flow and contractile function (Fig. 7B).
However, in remote regions of the myocardium, an apparent upregulation
of HSP70 expression was observed, especially following severe low-flow ischemia (Fig. 7B). HSP70 content increased from
0.7 ± 0.3 to 2.3 ± 0.4 µg/mg of total protein
(P < 0.05, n = 5) in the remote regions of the left ventricle where flow and function remained normal.
HSC73 content also was elevated slightly (i.e., 
16%, n = 7, P < 0.2) in the remote
myocardium. The amounts of HSC73 extracted from canine myocardium were
comparable to levels expressed in rodent and human hearts, but HSP70
protein content was somewhat lower than that found in other mammals
(15).
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 7.
Temperature-sensitive (37°C/95°C) aggregation
of tubulin isoforms (A) and HSC73/HSP70 (B) is
depicted in these representative Western blots derived from 5-h test,
remote, and sham tissue samples. Note that neither tubulin nor stress
protein isolated from test extracts (A and B)
entered the 10% SDS gel when samples were solubilized at 37°C;
however, when solubilized at 95°C, all isoforms of tubulin and HSC73
and HSP70 can be identified in the test extracts (A and
B). Each lane was loaded with 2 µg of protein from sham,
test, and remote sarcoplasmic extracts.
|
|
Tubulin, HSP70, and HSC73 interactions could be demonstrated in
severely ischemic myocardial extracts solubilized at 37°C but
not in moderately ischemic tissue preparations (Fig. 7,
A and B). Under these conditions, immunoblots
prepared from subendocardial sarcoplasmic fractions obtained from test
areas of the heart revealed that - and/or -tubulin as well as
HSC73 and HSP70 failed to enter 10% SDS-PAGE gels (Fig. 7,
A and B). Conversely, extracts prepared
at 37°C from remote regions of the left ventricle and sham-instrumented animals displayed all forms of both protein families
(Fig. 7, A and B). When severely ischemic
myocardium was reperfused for 24 h, the stress proteins and the
tubulin isoforms remained aggregated with neither class of proteins
being dissociated from one another after incubation at 37°C (data not
shown). No evidence of aggregation was ever detected in the remote
myocardium; furthermore, no tubulin-stress protein aggregation could be
documented in extracts prepared from tissue derived from moderately
ischemic-reperfused hearts (data not shown).
Properties of heterooligomeric tubulin-HSC73/HSP70 complexes.
Putative high molecular complexes between tubulin and HSC73/HSP70
were identified in sarcoplasmic fractions of the ischemic heart
fractionated by linear sucrose density gradient centrifugation. Although presence of both HSC73 and small amounts HSP70 was apparent in
the six heavier fractions of the gradient, the heat shock proteins were
found concentrated in fractions 3 and 4 of the
gradient. When aliquots of each of these fractions were separated on
4% native gels, only HSC73 could be identified in two high molecular aggregates that ranged in size from ~400 to ~800 kDa (Fig.
8). Aliquots of fraction 3 solubilized at 95°C revealed the presence of the HSC73 and - and
-tubulin (Fig. 9, lanes
1-3) and small amounts of -tubulin as well (not
illustrated). Aggregates isolated from fraction 4 displayed
reduced amounts of -tubulin and little evidence of either - or
-tubulin (Fig. 9, lanes 2 and 3). In addition,
a 30-kDa peptide also could be immunoblotted from fractions 3 and 4 with anti-HSC73 antibody (Fig. 9, lane
1). The other HSC73/HSP70-positive fractions subjected to
denaturing gel electrophoresis lacked tubulin. "Test"
fraction 3 consistently disclosed large amounts of the 800-kDa tubulin-HSC73/HSP70 aggregate (Fig. 8, 60% vs. 40% in fraction 3). To confirm that the aggregates reflected specific protein-protein interactions between tubulin and the HSP70 family of
stress proteins, anti-HSC73/HSP70 antibodies were employed to
immunoprecipitate proteins that interacted with the stress proteins in
fraction 3. The immunoprecipitates yielded only - (Fig.
10, lane 2), - and
-tubulin (data not shown) and HSC73 (Fig. 10, lane 5).
The anti-HSP70/HSC73 antibody routinely immunoprecipitated a constant
amount of HSC73 from fraction 3; however, the antibody only
brought down significant amounts of tubulin following severe low-flow
ischemia and reperfusion (Table
3). Whereas anti-HSP70/HSC73 antibody
immunoprecipitated tubulin, anti-tubulin antibodies failed to
immunoprecipitate either HSC73 or HSP70, suggesting that the tubulin
was sequestered within these heterooligomeric complexes. Furthermore,
the 30-kDa HSC73-positive peptide that was observed in immunoblots from
fractions 3 and 4 (Fig. 9) was not present in the
immunoprecipitates. The origin of this HSC73 cross-reactive peptide
remains unresolved at this juncture, although a HSC73-positive peptide
of similar size has been identified in "heat shocked" rat
myocardium and is believed to represent a degradation fragment of HSC73
(9). The presence of these high molecular aggregates also
could be identified in extracts prepared from reperfused hearts (not
illustrated). The content and size of the aggregates was similar to
those complexes isolated immediately following low-flow
ischemia (Table 3).
View larger version (7K):
[in this window]
[in a new window]
|
Fig. 8.
Western blots of HSC73-positive high-molecular
weight-aggregates prepared from sucrose gradient test
fractions 3 and 4 derived from 5-h
low-flow ischemic myocardium. Twenty micrograms of protein from
each fraction were loaded on to native 4% PAGE gels, and the gels
illustrate the presence of two large HSC73-positive/HSP70-negative
complexes that range in molecular mass from 800 kDa down to 400
kDa. Thyroglobulin (669 kDa) and catalase (232 kDa) were run in
parallel as molecular mass standards.
|
|
View larger version (8K):
[in this window]
[in a new window]
|
Fig. 9.
Western blots of - and -tubulin and HSC73 in
denatured test fractions 3 and 4 prepared from
extracts of 5-h low-flow ischemic myocardium. One microgram of
fraction protein was loaded in each lane of a 10% SDS-PAGE slab gel.
Lane 1 shows the presence of two HSC73-positive polypeptides
in both fractions with a molecular mass of 73 and 30 kDa, respectively.
Lanes 2 and 3 reveal that most of the - (i.e.,
75%) and -tubulin is present only in fraction 3.
|
|
View larger version (15K):
[in this window]
[in a new window]
|
Fig. 10.
-Tubulin (A) and HSC73 (B)
immunoprecipitated from fraction 3 of a 5-h low-flow
ischemic extract. Lanes 1 and 4 are
purified -tubulin standard (1 µg) or purified HSP70 (0.5 µg),
respectively. Lanes 2 and 5 are
immunoprecipitated -tubulin or HSC73, respectively. Lane
3 shows residual -tubulin in fraction 3 after immunoprecipitation, whereas lane 6 demonstrates that
the anti-HSP70/HSC73 antibody depletes the fraction of HSC73.
|
|
View this table:
[in this window]
[in a new window]
|
Table 3.
-Tubulin and HSP70/HSC73 content in immunoprecipitates isolated from
test fraction 3 of 5 h ischemic myocardial extracts
|
|
| |
DISCUSSION |
A graded reduction in coronary flow to the subendocardium induced
a complex redistribution of the constitutively expressed heat shock
protein, HSC73, to elements of the microtubule network in both test
and, to a lesser degree, remote canine cardiocytes. Although HSC73
stained intermyofibrillar microtubules, no HSC73-tubulin aggregates
could be isolated following moderate low-flow ischemia, implying the development of relatively "weak" interactions between tubulin and the stress protein. However, when coronary flow was reduced
more severely, microtubules depolymerized and high molecular weight
tubulin-HSC73 aggregates could be isolated from extracts of
ischemic tissue. Furthermore, these aggregates persisted for as
long as 24 h following reperfusion. The results of these
experiments have demonstrated that HSC73 and tubulin directly interact
with one another following the reduction of coronary flow and
"weak" as well as "strong" interactions appear to exist.
Selective preservation of centrosomal tubulin following severe low-flow
ischemia further suggests the existence of at least two pools
of tubulin, one of which appears to form complexes with HSC73. The
presence of some microtubule depolymerization in the remote myocardium
also argues that low-flow ischemia, in-and-of-itself, may not
be the sole stimulus responsible for the disruption of the microtubules
(24). Such observations strengthen the argument that
tubulin-HSC73 interactions may serve to stabilize portions of the
microtubule network under conditions recognized to disrupt microtubules
(13, 21).
Although the subcellular mechanism(s) that afford myocardial
"protection" remain largely unknown, heat shock/stress proteins are
believed to mediate cardioprotection, in part, as molecular chaperones
that target, fold, and stabilize cardiocyte cytoskeletal proteins
(3, 17, 30). Nevertheless, only a single report has
identified the cardiocyte's microtubule network as a potential target
for protection by members of the HSP70 family of stress proteins
(5, 8). Furthermore, overexpression of the constitutive (HSC72/73), not the inducible (HSP70) form, of the stress protein appeared to be responsible for stabilizing microtubule integrity in
"ischemic" neonatal rat heart cells (5). Brown
and colleagues (6) reported that the constitutive form of
HSP70 family of stress proteins (i.e., HSC73) acts as a chaperone,
which assists the folding of nascent tubulin at perinuclear centrosomes
in HeLa cells. The microinjection HSC73 into cultured HeLa cells also was demonstrated to "protect" centrosomes by illustrating that the
cells could nucleate new microtubule assembly immediately following
heat shock rather than requiring de novo synthesis of the stress
protein (7). From such experiments, the hypothesis has
evolved that HSC73 must directly interact with tubulin to facilitate
assembly and stabilization of the nascent microtubule network following
heat shock and that the centrosome is the likely site of such interactions.
The results from the present investigation provide additional evidence
supporting the contention that constitutively expressed HSC73 directly
interacts with tubulin in the ischemic myocardium. Two forms of
HSC73-tubulin interaction appear to develop following low-flow
ischemia. The colocalization of HSC73 and tubulin infers that
some form of "weak" interaction was established (Table 2) in
cardiocytes following moderate low-flow ischemia (Table 1), because high-molecular-weight HSC73-tubulin complexes could not be
isolated from either the ischemic or reperfused myocardium. The
nature of these weak interactions has not been defined, but such
colocalization may serve to stabilize the structure of
intermyofibrillar microtubules, which like cytoplasmic microtubules
(6, 7), appear most sensitive to heat shock and
ischemic stress. In contrast, tubulin-HSC73 complexes that have
been isolated from severely ischemic tissue (Figs. 8-10)
appear similar in size and composition to those recently identified in
the ciliated protozoan, Tetrahymena thermophila, following
heat shock (28). The stress protein appears to
preferentially target centrioles, which are central components of
centrosomes and basal bodies (7, 28). Perhaps the presence of HSC73 at these sites facilitates the oligmerization of cytosolic HSC73 during ischemic stress, creating the
high-molecular-weight complexes that sequester and protect centriolar
tubulin. Nevertheless, following reperfusion, the centrosome appears
incapable of promoting microtubule assembly from existing depolymerized
tubulin for at least 24 h, so it seems plausible that the
formation of such HSC73-tubulin complexes may limit the ability of
anti-tubulin antibodies to immunoprecipitate HSC73. The formation of
such oligomeric complexes may ensure that intermyofibrillar
microtubules ultimately can be reassembled from the "protected"
centrosome during recovery from reversible cell injury. The presence of
hetero-oligomeric complexes, depolymerized tubulin, and
disrupted myofibrils in stunned myocardium further
suggests that the three phenomena may be causally linked. Future
experiments must be extended to include prolonged reperfusions to test
this hypothesis.
Another intriguing scenario regarding the functional significance of
the HSC73-positive centrosomal complexes has emerged from several cell
culture experiments. Ciechanover's laboratory (4) has
provided evidence demonstrating that HSC73 also can function as a
chaperone for the ubiquitination of abnormal/damaged proteins,
including cytoskeletal proteins, and others have reported that this
activity occurs at the centrosome (1). The colocalization of HSC73 and tubulin also could be indicative of the ubiquitination of
"damaged" tubulin and, perhaps, other proteins (4, 14) at the centrosome. The presence of the multicatalytic proteasome in
enriched centrosomal fractions is consistent with these observations (1, 14, 27). Because a modest but significant decrease in
tubulin develops following low-flow ischemia and reperfusion, future experiments will have to address the relationship between HSC73,
tubulin turnover, and proteasome activity during low-flow ischemia to test the validity of this hypothesis. Conversely, others (12, 16, 30) have proposed that the formation of high-molecular-weight heterooligomeric complexes of heat shock proteins
actually may limit complex breakdown, thereby conferring a degree of
protection to subcellular organelles during stressful circumstances.
Perhaps the high levels of HSC73 expression in the canine myocardium
reflect the multiple functions that this stress protein must subserve
in maintaining cardiomyocyte homeostasis.
In summary, the present experiments have demonstrated that
depolymerization of intermyofibrillar microtubules parallels myofibril disruption during low-flow ischemia. Gradually reducing
coronary flow has provided evidence for two forms of HSC73-tubulin
interactions in the canine cardiocyte. It is conceivable that the weak
and strong interactions that develop may reflect the severity of the ischemic insult. Because the intermyofibrillar site has been
recognized as a site of thick filament disruption (24),
the redistribution of heat shock proteins to microtubules at these
sites appears compatible with the hypothesis that the HSC73 stabilizes
portions of the microtubule network and theoretically "protects"
the myofibril. Conversely, severe low-flow ischemia
depolymerizes intermyofibrillar microtubules, leaving only the
centrosomal microtubule sites intact. The development of
high-molecular-weight complexes that are composed of HSC73 and all
three tubulin isoforms demonstrates the highly specific nature of these
protein-to-protein interactions. These results support the contention
that this constitutively expressed stress protein may represent a first
line of defense in preserving selected elements of the cardiocyte's
microtubule network. Because myofibrillar disruption parallels
microtubule depolymerization, these results also provide additional
evidence supporting the contention (2, 20) that an intact
microtubule-based cytoskeleton may contribute to the stabilization of
myofibril structure in the cardiac myocyte.
| |
ACKNOWLEDGEMENTS |
The authors express thanks to the American Heart Association for
the Grant-in-Aid that partially funded this research study and to the
Feinberg Cardiovascular Research Institute for its continued support.
The authors express gratitude to Dr. Lester Binder for providing a
sample of a monoclonal antibody directed against -tubulin for use in
this investigation.
| |
FOOTNOTES |
Address for reprint requests and other correspondence:
R. S. Decker, Dept. of Medicine, Searle 2-575, S207,
Northwestern Univ. Feinberg School of Medicine, 303 E. Chicago Ave.,
Chicago, IL 60611 (E-mail:
r-decker{at}northwestern.edu).
The costs of publication of this
article were defrayed in part by the
payment of page charges. The article
must therefore be hereby marked
"advertisement"
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
June 6, 2002;10.1152/ajpheart.00062.2002
Received 25 February 2002; accepted in final form 17 May 2002.
| |
REFERENCES |
1.
Anton, LC,
Schubert U,
Bacik I,
Princiotta MF,
Wearsch PA,
Gibbs J,
Day PM,
Realini C,
Rechsteiner MC,
and
Bennink JR.
Intracellular localization of proteosomal degradation of a viral antigen.
J Cell Biol
146:
113-124,
1999[Abstract/Free Full Text].
2.
Armstrong, SC,
and
Ganote CE.
Flow cytometric analysis of isolated adult cardiomyocytes: vinculin and tubulin fluorescence during metabolic inhibition and ischemia.
J Mol Cell Cardiol
24:
149-162,
1992[Web of Science][Medline].
3.
Benjamin, IJ,
and
McMillan DR.
Stress (heat shock) proteins, molecular chaperones in cardiovascular disease.
Circ Res
83:
117-132,
1998[Abstract/Free Full Text].
4.
Bercovich, B,
Stancoviski I,
Mayer A,
Blumenfeld N,
Laszlo A,
Schwartz AL,
and
Ciechanover A.
Ubiquitin-dependent degradation of certain protein substrates in vitro requires the molecular chaperone HSC70.
J Biol Chem
272:
9002-9010,
1997[Abstract/Free Full Text].
5.
Bluhm, WF,
Martin JL,
Mestril R,
and
Dillmann WH.
Specific heat shock proteins protect microtubules during simulated ischemia in cardiac myocytes.
Am J Physiol Heart Circ Physiol
275:
H2243-H2249,
1998[Abstract/Free Full Text].
6.
Brown, CR,
Hong-Brown LQ,
Doxsey SJ,
and
Welch WJ.
Molecular chaperones and the centrosome. A role for TCP-1 in microtubule nucleation.
J Biol Chem
271:
833-840,
1996[Abstract/Free Full Text].
7.
Brown, CR,
Hong-Brown LQ,
Doxsey SJ,
and
Welch WJ.
Molecular chaperones and the centrosome. A role for HSP73 in centrosomal repair following heat shock treatment.
J Biol Chem
271:
833-840,
1996[Abstract/Free Full Text].
8.
Dillmann, WH.
Heat shock proteins and protection against ischemic injury.
Infect Dis Obstet Gynecol
7:
55-57,
1999[Medline].
9.
Donnelly, TJ,
Sievers RE,
Frank GJ,
Vissern MD,
Welch WJ,
and
Wolfe CL.
Heat shock protein induction in rat hearts. A role for improved myocardial salvage after ischemia reperfusion?
Circulation
85:
769-778,
1992[Abstract/Free Full Text].
10.
Goldstein, MA,
and
Entman ML.
Microtubules in mammalian heart muscle.
J Cell Biol
80:
183-195,
1979[Abstract/Free Full Text].
11.
Hartl, F.
Molecular chaperones in cellular protein folding.
Nature
381:
571-579,
1996[Medline].
12.
Ito, H,
Kamei K,
Iwamoto I,
Inaguma Y,
Nohara D,
and
Kato K.
Phosphorylation-induced change of the oligomerization state of B-crystallin.
J Biol Chem
276:
5346-5352,
2001[Abstract/Free Full Text].
13.
Iwai, K,
Hori M,
Kitabatake H,
Kurihara H,
Uchida K,
Inoue M,
and
Kamada T.
Disruption of microtubules as an early sign of irreversible ischemic injury: immunohistochemical study of in situ canine hearts.
Circ Res
67:
694-706,
1990[Abstract/Free Full Text].
14.
Johnston, JA,
Ward CL,
and
Kopito RR.
Aggresomes: a cellular response to misfolded proteins.
J Cell Biol
143:
1883-1898,
1998[Abstract/Free Full Text].
15.
Knowlton, AA,
Kapadia Torre-Amione S, G,
Durand JB,
Bies R,
Young J,
and
Mann DL.
Differential expression of heat shock proteins in normal and failing human hearts.
J Mol Cell Cardiol
30:
811-818,
1998[Web of Science][Medline].
16.
Lambert, H,
Charette SJ,
Bernier AF,
Guimond A,
and
Landry J.
HSP 27 multimerization mediated by phosphorylation-sensitive intermolecular interactions at the amino terminus.
J Biol Chem
274:
9378-9385,
1999[Abstract/Free Full Text].
17.
Liang, P,
and
MacRea TH.
Molecular chaperones and the cytoskeleton.
J Cell Sci
110:
1431-1440,
1997[Abstract].
18.
Mestril, R,
Chi SH,
Sayen K,
O'Reilly K,
and
Dillmann WH.
Expression of inducible stress protein 70 in rat heart myogenic cells confers protection against simulated ischemia-induced injury.
J Clin Invest
93:
759-767,
1994[Web of Science][Medline].
19.
Morimoto, RI.
Cells in stress: transcriptional activation of heat shock genes.
Science
259:
1409-1410,
1993[Free Full Text].
20.
Rappaport, L,
and
Samuel JL.
Microtubules in cardiac myocytes.
Int Rev Cytol
113:
101-143,
1988[Web of Science][Medline].
21.
Sato, H,
Hori M,
Kiakaze M,
Iwai K,
Takashima S,
Kurahara H,
Inoue M,
and
Kamada T.
Reperfusion after brief ischemia disrupts the microtubule network in canine hearts.
Circ Res
72:
361-375,
1993[Abstract/Free Full Text].
22.
Schwartz, AL.
The ubiquitin-proteasome pathway and the pathogenesis of human diseases.
Annu Rev Med
50:
57-74,
1999[Web of Science][Medline].
23.
Sherman, AJ,
Harris KR,
Hedjebeli S,
Yaroshenko Y,
Schafer D,
Schroff S,
Sung J,
and
Klocke FJ.
Proportionate reversible decreases in systolic function and myocardial oxygen consumption after modest reductions in coronary flow: hibernation versus stunning.
J Am Coll Cardiol
29:
1623-1631,
1997[Abstract].
24.
Sherman, AJ,
Klocke FJ,
Decker RS,
Decker ML,
Kozlowski KA,
Harris KR,
Hedjbeli S,
Yaroshenko Y,
Nakamura S,
Parker MA,
Checchia PA,
and
Evans DB.
Myofibrillar disruption in hypocontractile myocardium showing perfusion-contraction matches and mismatches.
Am J Physiol Heart Circ Physiol
278:
H1320-H1334,
2000[Abstract/Free Full Text].
25.
Watkins, SC,
Samuel JL,
Marotte F,
Bertier-Savalle B,
and
Rappaport L.
Microtubules and desmin filaments during onset of heart hypertrophy in rat: a double immunoelectron microscope study.
Circ Res
60:
327-336,
1987[Abstract/Free Full Text].
26.
Welch, WJ,
Kang HS,
Beckmann RP,
and
Mizzen LA.
Response of mammalian cells to metabolic stress: changes in cell physiology and structure/function of stress proteins.
Curr Top Microbiol Immunol
167:
31-55,
1991[Web of Science][Medline].
27.
Wigley, WC,
Fabunmi RP,
Lee MG,
Marino CR,
Muallem S,
DeMartino GN,
and
Thomas PJ.
Dynamic association of proteosomal machinery with the centrosome.
J Cell Biol
145:
481-490,
1999[Abstract/Free Full Text].
28.
Williams, NE,
and
Nelson EM.
HSP70 and HSP90 homologs are associated with tubulin in hetero-oligomeric complexes, cilia and the cortex of Tetrahymena.
J Cell Sci
110:
1665-1672,
1997[Abstract].
29.
Williams, RS,
and
Benjamin IJ.
Protective responses in the ischemic myocardium.
J Clin Invest
106:
813-818,
2000[Web of Science][Medline].
30.
Xiao, XZ,
and
Benjamin IJ.
Human genetics '99: The cardiovascular system. Stress-response proteins in cardiovascular disease.
Am J Hum Genet
64:
685-690,
1999[Web of Science][Medline].
Am J Physiol Heart Circ Physiol 283(4):H1322-H1333
0363-6135/02 $5.00
Copyright © 2002 the American Physiological Society
TOP
ABSTRACT
INTRODUCTION
