| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Institute of Experimental and Clinical Pharmacology and Toxicology and
* Institute of Anatomy, University-Hospital Eppendorf, Hamburg, Germany; and
Institute of Experimental and Clinical Pharmacology and Toxicology, Erlangen, Germany
1Correspondence: Institute of Experimental and Clinical Pharmacology and Toxicology, Friedrich-Alexander-University Erlangen-Nuremberg, Fahrstrasse 17, D-91054 Erlangen, Germany. E-mail: thomas.eschenhagen{at}pharmakologie.uni-erlangen.de
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-sarcomeric actin increased by 98% and 40%, respectively.
Morphologically, stretched EHTs exhibited improved organization of
cardiac myocytes into parallel arrays of rod-shaped cells, increased
cell length and width, longer myofilaments, and increased mitochondrial
density. Thus, stretch induced phenotypic changes, generally referred
to as hypertrophy. Concomitantly, force of contraction was two- to
fourfold higher both under basal conditions and after stimulation with
calcium or the ß-adrenergic agonist isoprenaline. Contraction
kinetics were accelerated with a 14–44% decrease in twitch duration
under all those conditions. In summary, we have developed a new
in vitro model that allows morphological, molecular, and
functional consequences of stretch to be studied under defined
conditions. The main finding was that stretch of EHTs induced cardiac
myocyte hypertrophy, which was accompanied by marked improvement of
contractile function.—Fink, C., Ergün, S., Kralisch, D.,
Remmers, U., Weil, J., Eschenhagen, T. Chronic stretch of engineered
heart tissue induces hypertrophy and functional improvement.
Key Words: tissue engineering • cell culture • EHT • atrial natriuretic factor
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Substantial progress has
been made over the last years to elucidate mechanisms leading to
hypertrophy. Early studies in animal models using the technique of
aortic banding have shown that mechanical stress and neurohumoral
activation are the two major stimuli for cardiac hypertrophy
(2
3
4
5)
. These principle factors are difficult to isolate
in vivo. Important progress has been made by using isolated
tissue-cultured myocytes to study either growth response to humoral
factors (6)
or mechanical stress under defined conditions.
The latter became possible with the use of neonatal rat cardiac
myocytes cultured on deformable silicone dishes subjected to stretch, a
technique introduced in the field by Komuro and Yazaki
(7
8
9
10)
. These studies have now identified a whole concert
of signaling mechanisms that link short-term stretch (0.5–60 min) to
the induction of genes such as atrial natriuretic factor (ANF),
ß-myosin heavy chain (ß-MHC), or skeletal -actin (sACT), which
in earlier studies have been identified to accompany cardiac
hypertrophy both in vivo (11)
and in
vitro (12)
. These changes appear to be part of the
reprogramming of cardiac gene expression toward a fetal phenotype and
are marker genes of hypertrophy. Thus, mechanical stress of isolated
cardiac myocytes appears to be sufficient to turn on hypertrophic gene
programs. This process appears to involve stretch-induced release of
angiotensin II and endothelin-1, activation of protein kinase C, the
mitogen-activated protein (MAP) kinases ERK, p38, and c-Jun
NH2-terminal kinases (JNK), as well as
stress-activated protein kinase (SAPK), and S6 kinase (8
, 9
,13
14
15
16
17
). Another possible link between stretch,
angiotensin II/endothelin-1 release, and hypertrophy could be
activation of the sarcolemmal
Na+/H+ exchanger, which
would induce an increase in intracellular Na+
and, consecutively, in intracellular Ca2+
(18
, 19)
. The increase in Ca2+ would
explain the positive inotropic response to stretch that develops in
minutes, i.e., the second phase of the response to acute increases in
loading (Anrep effect; 18), as well as the hypertrophic response,
possibly via activation of the calcineurin/NFAT3/GATA4 pathway
(20)
.
Although these in vitro studies provided important
mechanistic information, they cannot elucidate whether growth in
response to mechanical stress impairs or improves cardiac function.
This is, however, a crucial question with regard to potential
therapeutic intervention. Hypertrophy, i.e., the enlargement of myocyte
size, encompasses more than one phenotype. Myocardial function appears
to be normal or increased in hypertrophy that accompanies exercise
training or development even if the absolute magnitude of hypertrophy
is equal to that produced by hypertension, valve disease, or myocardial
infarction (21)
. Conversely, in the latter case the
myocardium develops less force, calcium transients are prolonged, and
the inotropic response to ß-adrenergic stimulation is blunted,
changes that may at least partly result from the altered gene
programming (22)
. Thus, hypertrophy can present as a
‘normal’, physiological response to altered demands, and it is the
only intrinsic compensation of the heart to maintain pump function
after loss of myocardial mass. On the other hand, it may represent a
maladaptive response in cardiac pathology. The mechanisms that cause
transition from beneficial growth to malfunction are still poorly
defined.
In the present study we made use of a technique that allows
reconstitution of cardiac myocytes to a 3-dimensional, spontaneously
and coherently beating heart tissue-like structure (23)
,
which we will refer to as engineered heart tissue (EHT). EHTs allow the
impact of various interventions on force, kinetics, and frequency of
the contraction to be studied together with morphological and molecular
parameters in the same sample under defined conditions in
vitro.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Briefly, minced ventricles from 9- to
11-day-hatched chick embryos (White Leghorn, VALO eggs; Lohmann,
Cuxhaven, Germany) were exposed to a fractionated trypsin/collagenase
digestion. For reduction of noncardiac myocytes, the cell suspension
was preplated for about 1 h. Unattached cells were pelleted,
washed, and resuspended in culture medium (DMEM, 10% heat-inactivated
horse serum, 2% chick embryo extract (Gibco BRL; Germany), 100 µg/ml
streptomycin, and 100 U/ml penicillin G (1x P/S; Gibco BRL) at a
density of 3.1 x 106 cells/ml to cast EHTs.
Neonatal rat cardiac myocytes
Rat cardiac myocytes were isolated from 1- to 3-day-old neonatal
Wistar rats (University of Hamburg breed, from Charles River, Bad
Kisslegg, Germany) by a modification of the technique described by
Webster et al. (24)
. Briefly, hearts from 50–70 pups were
minced and subjected to a serial trypsin digestion to release single
cells. After the final digestion, cells were washed and preplated for
1–2 h in complete culture medium. Unattached cells were pelleted and
suspended in culture medium at a density of 8.9 x
106 cells/ml to cast EHTs.
Engineered heart tissue
The principal technique has been described previously
(23)
and is shown in Fig. 1a
. In general, a cell/collagen master mix was prepared for 6
EHTs and kept on ice until casting. EHTs with embryonic chick cardiac
myocytes were composed of 1.3 ml collagen type I (rat tail; 3.7 mg/ml,
Upstate Biotechnology, Lake Placid, N.Y,), 1.3 ml twofold concentrated
DMEM supplemented with 20% heat-inactivated horse serum, 4% chick
embryo extract and 2x P/S, 172 µl 0.1 M NaOH for neutralization of
the collagen solution, and 1.4 ml cell suspension
(4.2x106 cells). EHTs with neonatal rat cardiac
myocytes were shown previously to depend on the presence of Matrigel
(25)
. They were composed of 1.1 ml collagen type I, 1.1 ml
twofold concentrated DMEM, 370 µl Matrigel (Harbor Bio-Products,
Tebu, Germany), 370 µl 0.1 M NaOH, and 1.2 ml cell suspension
(10.4x106 cells). For each EHT, 0.7 ml of the
ice-cold cell/collagen mix was poured into a well (11x17x4 mm) of the
culture dish that contained a layer of silicone rubber with six
rectangular wells, each holding one set of Velcro-coated silicone tubes
(7 mm length, 3 mm outer, 2 mm inner diameter) kept at a fixed distance
with a metal wire spacer (Fig. 1a
). The mixture was allowed
to gel at 37°C for 60 min before culture medium was added to the
dish. Medium changes were performed after overnight incubation and then
every other day.
|
Chronic mechanical stretching
EHTs were cultured for 4 days as described above and then
exposed to an unidirectional and phasic stretch in a special motorized
stretching device for 6 days. To determine the optimal stretch extent,
the first experiments were done with EHTs (chick) by stretching them by
+1–20% of the original (spacer) length. A cartoon of this stretching
device in shown in Fig. 6a
. Based on these results, the
following experiments were done in the stretching device shown in Fig. 1b
. EHTs were connected at both ends to stretching bars,
which pulled the matrix apart with a constant frequency of 1.5 Hz and,
as a result from the first experiment, to an extent of +20% of
original length. An original recording of the phasic stretch is
demonstrated in Fig. 1c
. The entire arrangement was held in
a CO2 incubator at 37°C. The culture medium was
changed every other day. Unstretched control EHTs were held under
identical conditions without stretch.
|
Histology
EHTs were carefully removed from the holding rods after the
force measurement and fixed, with the metal wire spacer attached, at
4°C in 4% formaldehyde in phosphate-buffered saline (PBS), pH 7.4.
After dehydration in graded concentrations of ethanol and paraffin
infiltration according to standard procedures, the specimens were
removed from the wire spacer and embedded in paraffin blocks. Sections
(10 µm) were cut parallel to the plane of the tissue, stained with
hematoxylin-eosin (H&E), and photographed with a Zeiss Axioplan
microscope.
For toluidine blue staining and electron microscopy (EM), two stretched and two unstretched EHTs from two independent series of experiments were fixed overnight in phosphate-buffered glutaraldehyde (5.5%), cut in half, and prepared for embedding in Epon (Serva, Heidelberg, Germany), parallel and perpendicular to the plane of the tissue, as follows. EHTs were fixed for 2 h in 2% OsO4/saccharose-phosphate [1:2] at 4°C, dehydrated in graded series of ethanol, incubated with propylene oxide for 2 h 15 min and with propylene oxide-Epon [1:4] for 12 h, and subjected to polymerization of the Epon for 5-10 h at 37°C, 12 h at 45°C, 12-24 h at 60°C. For toluidine blue staining, semithin sections (1 µm) of both orientations were cut, stained, and photographed at 400-fold magnification with a B/W negative film. 13 x 18 cm prints were used to quantify cell diameter. For EM, ultrathin sections (90 nm) of both orientations were cut with an MT 6000 microtome (Sorvall, Newtown, Conn.). The sections were contrasted with uranyl-acetate for 20 min and lead-acetate for 5 min and examined with an EM 300 electron microscope (Phillips, Eindhoven, The Netherlands). Altogether, 5 toluidine blue-stained semithin cross sections and 5 EM ultrathin sections were cut from each block (= 40 sections each).
Force measurement
After 10 days in culture, EHTs were removed either from the
culture dish or the stretching device and subjected to isometric force
measurement in thermostatted organ bathes as described previously
(23)
. After 15 min equilibration without pacing, EHTs were
electrically stimulated with rectangular pulses (10 ms, 20–40 V) at a
standard frequency of 1.5 Hz (chick) or 2 Hz (rat). Preload was
stepwise adjusted to Lmax, i.e., the length at
which the preparation developed maximal force. The inotropic response
to calcium was investigated by a cumulative increase in the
extracellular calcium concentration from 1.8 to 12.6 mmol/l (chick) or
0.4 to 2.0 mmol/l (rat). The effect of the ß-adrenergic agonist
isoprenaline (1 µmol/l) was tested at a calcium concentration of 1.8
mmol/l (chick) and 0.4 mmol/l (rat). Force was evaluated after
equilibration (5 min). Time to 80% relaxation
(T2) was determined before and after
isoprenaline. The length of contraction experiments was 3 to 5 h.
During that time basal force of contraction decreased by less than
30%.
Determination of protein, RNA and DNA content, and cell number
EHTs were either removed directly from the culture dish or the
stretching device or carefully removed from the holding rod after
contraction experiments and rinsed three times with PBS. RNA and DNA
were extracted with a commercially available kit (TRIzol; Gibco BRL)
according to the manufacturer‘s instructions. Briefly, frozen EHTs
were homogenized in 600 µl of TRIzol with a Polytron (Kinematica AG
Littau, Switzerland) and subjected to phenol-chloroform extraction. RNA
was precipitated from the supernatant. DNA was extracted and
precipitated from the phenol and interphase and finally solubilized in
8 mmol/l NaOH. Concentration was photometrically quantified at 260 nm.
For determination of protein and cell count, EHTs were digested with
500 µl 0.1% collagenase in PBS, pH 7.4 for 1 h at 37°C under
constant shaking. Cells were counted microscopically in a 10 µl
aliquot. The rest of cell suspension was centrifuged 10 min at 500 rpm.
The sediment was washed twice with PBS, pH 7.4 and finally solubilized
with 100 µl sodium dodecyl sulfate (SDS) -Laemmli buffer
(26)
. Total protein was determined in duplicate in the
Laemmli buffer according to Bradford (27)
using
-globuline as a standard.
3H thymidine incorporation
DNA synthesis was determined by measuring incorporation of
3H thymidine as described previously
(23)
. One microliter 3H thymidine
(20 Ci/mmol, 1 µCi/µl; NEN, Bad Homburg, Germany) was added to the
EHTs in 1 ml medium. After 6 h of incubation at 37°C in the
presence or absence of stretch, EHTs were washed 3 x 10 min with
2 ml PBS, dissected from the Velcro, and dissolved in 1 ml 0.1% acetic
acid (18 h, 4°C). The samples were rapidly filtered over fiberglass
filters, washed twice with 6% trichloric acid, and once with 95%
ethanol. Radioactivity was counted after overnight incubation.
Northern blot analysis
RNA blotting, cDNA labeling, hybridization, and quantification
were performed essentially as described previously (28)
.
To correct measurements for minor loading differences, the membranes
were rehybridized with a 32P-labeled cDNA coding
for the housekeeping enzyme glyceraldehyde-3-phosphate dehydrogenase
(GAPDH).
Western blot analysis
50 µg of total protein from EHTs, solubilized in SDS-Laemmli
buffer, were separated on 10% SDS-polyacrylamide gel electrophoresis
and blotted essentially as described previously (29)
.
Immunochemistry was performed with the mouse monoclonal antibody
against -sarcomeric actin (Sigma, St. Louis, Mo.; dilution 1:5000),
an alkaline phosphatase-coupled anti-mouse antiserum (1:5000, Dianova,
Germany) and a color reaction with NBT/BCIP.
Statistics
All values presented are arithmetic means ±
SE. Curves were fitted with the PC-based curve fitting
program (GraphPads). Student‘s t test for unpaired
observations was used in all experiments. A P value of less
than 0.05 was considered significant.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
), hooked
onto the stretching bars, and exposed to phasic stretch in the
motorized stretching device (Fig. 1b
). Motorized revolving
of an elliptical rod (minimal diameter 10, maximal diameter 12 mm)
produced a phasic sinusoidal increase by 20% of original (spacer)
length with a frequency of 1.5 Hz (Fig. 1c
). Chick and rat
EHTs behaved similarly; differences will be described. Microscopic data
were derived primarily from chick EHTs, molecular data mainly from rat
EHTs, and functional data from both.
Effect of stretch on morphology
EHTs typically show a denser tissue-like structure made up from
longitudinally oriented well-developed cardiac myocytes at the free
lateral edges and a sparse population of small cells without consistent
orientation in the center (Fig. 2a, c
). The cell gradient from the edges to the center is most
likely due to load differences between the edges and the center of the
matrix (23)
. Six days of stretch resulted in a
macroscopically detectable increase of the cell-dense portion at the
free edges of EHTs (Fig. 2a, b
). Microscopically, stretched
EHTs exhibited a higher density of longitudinally orientated rod-shaped
cells, with a larger cell-to-nucleus ratio that formed a loose heart
tissue-like structure (Fig. 2c, d
). In addition, cells from
the sparsely populated center of the matrix also increased in size and
H&E staining intensity. To quantify the influence of stretch on cell
size, toluidine blue-stained semithin sections from the lateral free
edges were evaluated for the frequency distribution of individual cell
diameters of different size, regardless of the position at which the
cells were cut (Fig. 3a, b
). Care was taken not to count cell groups as one, but
this was not a serious problem. The statistical analysis showed an
increase in mean cross-sectional cell diameter in stretched EHTs by
41% (mean 7.7±0,1 vs. 5.4±0.1 µm; Fig. 3c
). Counting of
all cells larger than 1 µm was chosen for statistical simplicity and
because this evaluation also takes into account changes in mean
width/length. Yet it was obvious that the standard morphometric
parameter, cell diameter at the nucleus, showed similar differences
between stretched and unstretched EHTs (see Fig. 3a, b
and
Fig. 4
).
|
|
|
To get some information about subcellular alterations induced by
stretch, >800 cells from two stretched and two unstretched EHTs (40
sections from the lateral free edges) were evaluated by electron
microscopy. EM confirmed the significant stretch-induced increase in
cell size (Fig. 4a, b
). It was obvious from inspection that
stretch increased both the density (Fig. 4a, b
) and the
length of individual myofilaments (Fig. 4c, d
), as well as
the overall density of mitochondria (approximately by 20%) and the
association of mitochondria with the myofilaments (Fig. 5a, b
). Even though a formal morphometric analysis was not
performed, we feel safe in estimating the density of myofilaments to be
at least twofold higher in stretched EHTs. Despite the problems of
orientation of individual myofilaments in the plane of the section
(better longitudinal orientation could theoretically lead to
overestimation of length), a marked increase in myofilament length was
also apparent. Thus, the mean number of sarcomeres in series was at
least twofold higher in stretched EHTs, and the maximal number of
sarcomeres in series ever observed in ~400 unstretched and 400
stretched cells was 3–4 in control and 6–8 in stretched EHTs. In
addition, nuclear membranes were generally smooth in stretched and
wrinkled in unstretched EHTs (Figs. 4
, 5)
. Smoothening of the nuclear
membrane is another hallmark of cardiac hypertrophy (30)
.
Cardiac myocytes in unstretched EHTs contained more rough endoplasmic
reticulum (Figs. 4
, 5)
than in stretched EHTs, which is a classical
sign of high protein synthesis rates. Its loss, in contrast, is
regarded as indicative of differentiation.
|
Effect of stretch on contractile function
Contractile function was determined 10 days after casting
(=6 days after stretch) in standard organ baths under isometric
conditions. Stretch did not significantly influence spontaneous
frequency early after suspension in the organ bath (91±4 beats/min in
chick (n=31) and 142 ± 11 in rat (n=14).
However, stretched EHTs developed markedly increased contractile force
both under basal and stimulated conditions. The dependency of this
effect on the degree of the 6 day stretch was tested in a device in
which EHTs were exposed not to a uniform, but a gradual increase in
length by 1–20% of the original length (Fig. 6a
). Stretch up to +3% did not affect force of contraction,
indicating a threshold of the stretch effect. Stretch between +3 and
+20% induced a 2.5 - to 3.8-fold increase in force development, with
no clear ‘dose dependency’ (Fig. 6b
). Higher degree of
stretch was not tested.
In EHTs stretched for 6 days under our standard condition (20%
stretch), force of contraction was higher than in unstretched controls
at all preload levels (Fig. 7a, b
). Force of contraction at Lmax in
stretched chick and rat EHTs was increased by 238 and 188% over
unstretched controls, respectively. The difference remained
approximately constant at all levels of extracellular calcium (Fig. 7c, d
). As described previously (25)
, the
calcium sensitivity of rat EHTs was markedly higher than that of chick
EHTs, with EC50 values of 3.7 and 0.4 mM,
respectively. Stretch did not influence calcium sensitivity. A
maximally effective concentration of the ß-adrenergic agonist
isoprenaline increased twitch amplitude by 19.5 ± 5.5% and
19.4 ± 3.7% in unstretched and stretched chick, and by 35.2 ± 5.4% and 48.8 ± 9.8% in unstretched and stretched rat EHTs,
respectively. Twitch duration was significantly shorter in stretched
EHTs, an effect mainly due to a significant decrease in time-to-80%
relaxation (T2; Fig. 8c, d
). The difference between stretched and unstretched EHTs
was further enlarged under maximal ß-adrenergic stimulation. The
isoprenaline-induced decrease in T2, i.e., the
positive lusitropic effect, amounted to 16 and 12% in unstretched and
stretched chick and 35 and 55% in unstretched and stretched rat EHTs,
respectively (Fig. 8c, d
).
|
|
Effect of stretch on RNA/DNA content, cell number, 3H
thymidine incorporation, and gene expression
Stretched chick EHTs exhibited an increased RNA/DNA ratio (+100%;
Fig. 9a
) and an increased protein content per cell (+50%; Fig. 9b
); DNA content did not differ. 3H
thymidine incorporation on day 2 of stretch did not differ
significantly between stretched and unstretched chick EHTs (1339±410
vs. 1075±394 dpm/EHT, n=6), but when expressed as percent
of the respective controls of each series, amounted to 130 ± 8%
of control (P<0.05). 3H thymidine
incorporation also tended to be higher directly after initiation of
stretch (110% of control, n=2) and at day 3 (104% of
control, n=2). In contrast, total cell count per EHT,
determined by collagenase-mediated cell isolation and microscopic cell
counting, was significantly smaller in stretched EHTs than unstretched
controls (149,580±11,650 [n=12] vs. 207,140±25,870
cells/EHT [n=7], P=0.04). Similar results were
obtained in rat EHTs. ANF mRNA levels were twofold higher in stretched
rat EHTs (Fig. 9c, e
). Similarly, stretch induced a 40%
increase in -sarcomeric actin levels (Fig. 9d, f
).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-sarcomeric actin levels increased.
4) Peak contraction amplitude was markedly enhanced and
relaxation was accelerated. Thus, under these tissue culture
conditions, stretch induces quantitative and qualitative phenotype
changes that are generally assumed to indicate cardiac hypertrophy. The
stretch-induced changes are accompanied by improvement and not by
impairment of contractile function. If this conclusion can be
generalized to other tissue culture work on cardiac myocytes (which has
yet to be shown), the data suggest that the mechanisms of hypertrophy
identified so far under cell culture conditions apply to normal growth
processes and not necessarily to those that lead, under in
vivo conditions, to deterioration of cardiac function.
Under in vivo conditions, the distinction between
‘physiological’ (compensated) and ‘pathological’ (decompensated or
maladaptive) hypertrophy remains rather descriptive, and only a few
factors have been associated with the transition between these two
principal forms of cardiac hypertrophy. One candidate may be
endothelin-1 (ET-1), which was up-regulated in the (clinically)
decompensated but not the compensated state of LV hypertrophy in
salt-sensitive rat (31)
. Since treatment with an ET-1
receptor antagonist delayed clinical decompensation and death in these
animals, the idea was raised that other (unknown) factors cause
hypertrophy during compensation and that ET-1 initiates a pathological
process, leading finally to decompensation. This relates to the
therapeutically important question of whether decompensated hypertrophy
represents just the consequence of too much of the same factors that
trigger physiological hypertrophy or whether decompensated hypertrophy
is the consequence of a qualitatively distinct program.
Given the complexity of the in vivo situation where
mechanical load, perfusion pressure, autocrine/paracrine, and systemic
hormonal stimulation are necessarily interwoven, attempts have been
made to establish in vitro cell culture models that allow
the effect of hormonal and mechanical factors on cardiac myocytes to be
studied under defined conditions. Simpson and co-workers
(12)
used neonatal rat cardiac myocytes to show more than
15 years ago that 1-adrenergic stimulation induces cardiac
hypertrophy and that the Gq-phospholipase C-protein kinase C signal
transduction cascade plays a central role in this process. The main
conclusions of these early studies have only recently been verified by
elegant work with transgenic animals (32
33
34
35
36)
. The groups
of Yazaki and Izumo (7
8
9
10)
developed and characterized a
model in which neonatal rat cardiac myocytes are cultured on deformable
silicone dishes and exposed to stretch. These experiments showed that
mechanical stress of cardiac myoctes initiates similar hypertrophic
gene programs as stimulation with Gq-coupled receptor agonists, thereby
supporting the concept that load itself represents a major trigger for
cardiac hypertrophy (4)
. These and other studies have
focused on the short-term affection (minutes to hours) of cardiac gene
expression, namely, hypertrophic markers such as ANF, skeletal
-actin, and ß-MHC, and have delineated a variety of signaling
cascades involved in this response (9
, 10
, 15
, 16)
.
The experimental model described in the present study offers a number
of unique features that may help to complement these data.
1) Most important, it allows exact and reproducible
measurements of the functional consequence of stretch and other
hypertrophic interventions. This should help to answer important
questions. Does a different quality or quantity of stretch alone or
stretch in the presence of other hypertrophic stimuli (that are
necessarily present in vivo) ultimately lead to
deterioration of contractile function? Does chronic stimulation with
cardiotrophin/LIF-1 [that appear to stimulate assembly of sarcomeric
units in series (37)
] affect contractile function other
than stimulation with phenylephrine (that stimulate assembly of
sarcomeric units in parallel? 2) EHTs are remarkably stable
(tested for up to 5 wk) when compared with the papillary muscles (a
dying preparation with an ischemic core) and with cardiac myocytes
cultured on deformable silicone dishes. The stability of EHTs allowed
investigation of long-term consequences of stretch on cardiac
morphology and function, which showed (to our knowledge, for the first
time) that stretch actually induces enlargement of individual cardiac
myocytes, increases in mitochondrial and myofilament density, and
functional improvement. 3) The present experiments were
performed in the continuous presence of 10% serum plus 2% chick
embryo extract, culture conditions generally considered as being
maximally growth promoting. This is in contrast to earlier experiments
that used serum-starved cells exclusively to induce hypertrophy. EHTs,
in contrast to 2-dimensional cultures, allow for the addition of serum
because they are not overgrown by nonmyocytes (23)
. Thus,
the 3-dimensional structure dramatically reduces proliferative
activity. In direct comparisons, mean 3H
thymidine incorporation amounted to 120,000
dpm/106 cells in 2-dimensional cultures
(relatively stable between days 0 and 6 of culture) and 1000–6000 dpm
in EHTs (declining between day 1 and 6; unpublished experiments and ref
23
). Similar data have been reported for 3-dimensional
fibroblast cultures (38)
. The effects of serum are an old
debate in cell culture, and it is clear that cardiac myocytes in
vivo are normally in contact with the filtrate of plasma only,
devoid of components of coagulation cascade. Nevertheless, serum
contains growth factors, hormones, trace elements, and vitamins the
cell in vivo is provided with and that are absent in
standard synthetic media. Thus, it is likely that interventions in
serum-starved cells in vitro induce normalization of an
atrophic state rather than hypertrophy, i.e., the "morbid enlargement
or overgrowth of an organ or part due to an increase in size of its
constituent cells" (39)
. 4) EHTs are made up
by a mixed population of cells, including cardiac myocytes
[~80–90% (23)
], fibroblasts, and other unidentified
cells. As an advantage, this condition closely reflects the
physiological situation of an intact heart in which the biological
response is necessarily an integrated response of different cells. On
the other hand, the increased complexity may obscure clear-cut
cause-and-effect relationships. Thus, the effect of stretch on the
percentage of myocytes vs. nonmyocytes is difficult to determine
exactly. However, stretch neither markedly affected
3H thymidine incorporation nor increased cell
number. In contrast, the number of cells that could be isolated by
collagenase digestion of EHTs was significantly lower in stretched
EHTs. Given that 3H thymidine incorporation (if
any) modestly increased, we believe that the lower cell count reflects
the better tissue development in stretched EHTs and consequently the
greater difficulty in cell isolation. Hence, it is unlikely that
changes in cellular composition account for the observed changes in
contractile function or gene expression.
Stretch of EHTs induced a 2- to 4-fold increase in contractile force,
but only a 1.4-fold increase in mean cell diameter and a 1.5- to 2-fold
increase in RNA/DNA and protein/cell ratio and actin expression,
indicating that force per tissue mass increased. In addition, twitch
kinetics were accelerated and not prolonged, and the response to the
ß-adrenergic agonist isoprenaline was not blunted, but, if anything,
improved in rat EHTs (difference not significant). This suggests a
stretch-induced qualitative change in cardiac phenotype toward a more
adult form of contraction. Increased density of mitochondria (as shown
previously in exercise hypertrophy) and better development of
myofilaments could contribute to this improvement. Also, a higher
percentage of cells that actively contribute to the coherent
contraction of EHTs, i.e., better formation of a network and a higher
degree of cardiac cell differentiation may be important. Stretch
increased ANF mRNA levels, but did not up-regulate ß-myosin heavy
chain (MHC; the fetal isoform with a low ATPase activity; data not
shown), which may correspond to the increased contraction velocity of
stretched EHTs. A similar disparity was recently described in rats in
which volume overload was associated with an increase in ANF and BNF,
but not with a fetal isoform switch (40)
that is generally
associated with decreased contractile function. In contrast, pressure
overload led to both molecular changes (39)
. These data
are compatible with the idea that stretch under our conditions imitate
volume rather than pressure overload, but further work is needed to
clarify this question.
One may argue that motorized stretch not only imposes increased load on the cells, but also improves metabolic supply (by ‘stirring’ the culture medium), and that this causes hypertrophy and functional improvement. We cannot completely exclude this possibility, but consider this hypothesis to be unlikely for the following reasons. EHTs are very thin (~0.4 mm) and well accessible for medium, arguing against relevant diffusion barriers. A small degree of stretch (< 4%) had no effect, even though the medium was also stirred.
In summary, the present experiments show that stretch of reconstituted heart tissue induces an increase in mean cardiac cell size, which, under cell culture conditions, is generally referred to as hypertrophy. This response was seen in the presence of 10% serum and additional growth factors and was accompanied by marked improvement of contractile function. These data suggest that many processes that have been identified as being involved in the development of ‘cardiac myocyte hypertrophy in vitro’ are indistinguishable from those necessary for physiological growth of cardiac myocytes. Future experiments with the new model should help to differentiate factors or factor combinations that induce hypertrophy with improved function from those inducing hypertrophy with decreased function.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-skeletal muscle actin mRNAs accumulate in hypertrophied adult rat heart. Circ. Res. 59,551-555[Abstract]
proteins increases arrythmogenic effects of ß-adrenergic stimulation in the heart. J. Mol. Cell. Cardiol. 30,1917-1928[Medline]
q overexpression induces cardiac contractile failure in mice. Proc. Natl. Acad. Sci. USA 94,8121-8126This article has been cited by other articles:
|
|
 |
|
  T. Bupha-Intr, J. W. Holmes, and P. M. L. Janssen Induction of hypertrophy in vitro by mechanical loading in adult rabbit myocardium Am J Physiol Heart Circ Physiol, December 1, 2007; 293(6): H3759 - H3767. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. N. Morritt, S. K. Bortolotto, R. J. Dilley, X. Han, A. R. Kompa, D. McCombe, C. E. Wright, S. Itescu, J. A. Angus, and W. A. Morrison Cardiac Tissue Engineering in an In Vivo Vascularized Chamber Circulation, January 23, 2007; 115(3): 353 - 360. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Tobita, L. J. Liu, A. M. Janczewski, J. P. Tinney, J. M. Nonemaker, S. Augustine, D. B. Stolz, S. G. Shroff, and B. B. Keller Engineered early embryonic cardiac tissue retains proliferative and contractile properties of developing embryonic myocardium Am J Physiol Heart Circ Physiol, October 1, 2006; 291(4): H1829 - H1837. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W.-H. Zimmermann, M. Didie, S. Doker, I. Melnychenko, H. Naito, C. Rogge, M. Tiburcy, and T. Eschenhagen Heart muscle engineering: An update on cardiac muscle replacement therapy Cardiovasc Res, August 1, 2006; 71(3): 419 - 429. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Sekine, T. Shimizu, J. Yang, E. Kobayashi, and T. Okano Pulsatile Myocardial Tubes Fabricated With Cell Sheet Engineering Circulation, July 4, 2006; 114(1_suppl): I-87 - I-93. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Eschenhagen and W. H. Zimmermann Engineering Myocardial Tissue Circ. Res., December 9, 2005; 97(12): 1220 - 1231. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Baar New dimensions in tissue engineering: possible models for human physiology Exp Physiol, November 1, 2005; 90(6): 799 - 806. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. LAMMERDING, R. D. KAMM, and R. T. LEE Mechanotransduction in Cardiac Myocytes Ann. N.Y. Acad. Sci., May 1, 2004; 1015(1): 53 - 70. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. D. I. Anderson, F. Wang, and D. G. Gardner Role of the Epidermal Growth Factor Receptor in Signaling Strain-dependent Activation of the Brain Natriuretic Peptide Gene J. Biol. Chem., March 5, 2004; 279(10): 9287 - 9297. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Radisic, L. Yang, J. Boublik, R. J. Cohen, R. Langer, L. E. Freed, and G. Vunjak-Novakovic Medium perfusion enables engineering of compact and contractile cardiac tissue Am J Physiol Heart Circ Physiol, February 1, 2004; 286(2): H507 - H516. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Akhyari, P. W. M. Fedak, R. D. Weisel, T.-Y. J. Lee, S. Verma, D. A. G. Mickle, and R.-K. Li Mechanical Stretch Regimen Enhances the Formation of Bioengineered Autologous Cardiac Muscle Grafts Circulation, September 24, 2002; 106(12_suppl_1): I-137 - I-142. [Abstract] [Full Text] [PDF]
|
|
|
|
