Download Anterior and posterior left ventricular sarcomere lengths behave

Anterior
lengths
and posterior left ventricular
sarcomere
behave similarly
during ejection
J. M. GUCCIONE,l
‘Department
Baltimore,
San Diego,
W. G. O’DELL,’
A. D. McCULLOCH,2
Guccione,
J. M., W. G. O’Dell, A. D. McCulloch,
and
W. C. Hunter.
Anterior
and posterior left ventricular
sarcomere lengths behave similarly during ejection. Am. J. Physiol.
272 (Heart Circ. Physiol. 41): H469-H477,
1997.-Previous
studies of regional
differences
in myocardial
deformation
between the anterior
and posterior walls of the canine left
ventricle
were based on strain, which is not an absolute
measure of deformation.
We thus compared sarcomere lengths
at anterior and posterior sites during ejection in isolated dog
hearts. Cineradiographic
imaging
of regional
deformation
with radiopaque
markers implanted
near the midwall in five
hearts and just below the epicardium
in six hearts, combined
with postmortem
histology, allowed sarcomere length reconstruction throughout
the cardiac cycle. The amount of sarcomere shortening
accompanying
left ventricular
ejection was
similar
in both walls of the left ventricle
for sarcomeres
located at epicardial
and midwall sites. The mean sarcomere
length (taken at the middle of the ejecting range) was also
similar between the anterior and posterior sites when averaged over all hearts. The similarity
of sarcomere
function
held not only at end systole but throughout
ejection and over
wide ranges of ventricular
pre- and afterloads.
Hence functional measurements
of relative myocardial
shortening
may
not be indicative of regional sarcomere length heterogeneity.
strain;
myocardium;
AND
W. C. HUNTER1
of Biomedical
Engineering,
School of Medicine,
The Johns Hopkins
Maryland
21205-2195; and “Department
of Bioengineering,
University
La Jolla,
California
92093-0412
ventricular
function
DIFFERENCES in the patterns of myocardial
deformation between the anterior, lateral, and posterior walls of the canine left ventricle have been observed (3, 17). This is not surprising because the
beating heart is a complex three-dimensional (3-D) and
fiber-wound structure with mechanical properties that
are nonlinear, anisotropic, time varying, and probably
spatially inhomogeneous. These previous studies were
based on a description of deformation in a limited
region of the wall by means of strain tensors. A tensor
description of strain provides information on relative
stretches in any direction. However, strain is not an
absolute measure of deformation because it is measured relative to an arbitrary reference configuration
such as end diastole.
Recently, our laboratory developed a method that
allows the “reconstruction”
of absolute sarcomere
lengths (SLs) and orientations at any transmural layer
in the left ventricular
wall during any phase of the
cardiac cycle (11). It is not possible with current techniques to observe directly the in vivo behavior of
sarcomeres in the beating heart. Until this method was
developed, the only techniques used for measuring SLs
transmurally involved histological procedures on hearts
fixed at specific left ventricular
pressures so as to
simulate different nonsystolic states in the cardiac
REGIONAL
0363-6135/97
$5.00
Copyright
I:C 1997
University,
of California,
cycle. Relative strain analyses in the anterior and
posterior walls of the left ventricle were correlated with
histological measurements by Villarreal and Lew (17)
and Fann et al. (31, who measured the fiber orientations
and related these fiber directions to the direction of
maximum strain. These studies did not, however, attempt to measure absolute SLs or correct precisely for
any artifactual deformation arising from the histological procedures.
We thus examined the canine left ventricle for potential regional differences in sarcomere function during
ejection. We compared SLs reconstructed at posterior
and anterior sites at different transmural depths in
beating canine hearts operating under a wide range of
hemodynamic loads. SL was reconstructed by combining cineradiography of local deformations with postmortem histology of the same region.
METHODS
Reconstruction
of sarcomere deformation
(as the component in the myofiber direction
of 3-D deformation>
in the
isolated beating canine heart was accomplished
by cineradiography of local wall deformation
calibrated
to SL by postmortem histology. The main stages of the procedures were sterile
implantation
of radiopaque
marker beads, preparation
of the
isolated heart, recording
of the bead motion for a series of
pressure and volume excursions in the isolated beating heart,
reconstruction
of 3-D strains in the region of the implanted
markers, histological
analysis of the geometric relationship
between the local sarcomere vector and the bead positions,
and reconstruction
of strain in the direction
of the sarcomeres. These methods
follow closely those described
by
Rodriguez et al, (11); only a brief summary is given here.
Analysis
of Regional
Deformation
Bead implantation.
The initial stage involved implantation
of arrays of stainless steel beads to mark local points within
the ventricular
wall for the subsequent
cineradiographic
imaging.
Mongrel
dogs (B-20
kg) were anesthetized
with
pentobarbital
sodium (30-40 mg/kg), and a left lateral thoracotomy was performed under sterile conditions. Anterior
and
posterior sites in the left ventricular
free wall were selected
that were free of large vessels and near the equator (one-third
of the distance from base to apex). The anterior
site was
located just lateral to the left anterior
descending
coronary
artery. The posterior
site was located just lateral
to the
posterior
descending
coronary artery. At each site, sets of
three beads each were placed into roughly equilateral
triangles (1 cm on a side) at -6 and 0.5 mm from the epicardial
surface. We will refer to data collected at these two depths as
“midwall”
and “epicardial,”
respectively.
The mean depths
below the epicardium
actually obtained at the anterior
and
posterior midwall sites were 4.8 t 0.1 (SE) and 6.0 t 0.9 mm,
respectively.
This difference
in depth was not significant
the American
Physiological
Society
H469
H470
ANTERIOR-POSTERIOR
LV SARCOMERE
according to a one-way repeated-measures
analysis of variance (P > 0.23) (5).
Surgical
attachment
of mitral
ring. After a 7- to IO-day
recovery period, the implanted
dog was reanesthetized
as
before, and its heart was isolated and metabolically
supported by cross circulation
from a second anesthetized
dog
(14). Oxygenated
blood from the support dog was supplied
directly
to the coronary
bed of the isolat!ed heart, and
coronary venous blood was collected through
a vent in the
right ventricle for return to the support dog. After cutting all
mitral chordae tendineae, a Teflon support ring was sewn into
the fibrous ring surrounding
the mitral valve. The rigid
support ring allowed
the insertion
of a water-filled
latex
balloon into the left ventricle and secured the heart to a servo
pump apparatus (Vivitro Systems, Superpump
System, Victoria, BC).
Hear-t suspension. The heart was positioned with its apexbase axis horizontal,
supported
by a pericardial
“cradle”
consisting of a plastic sheet. The heart was oriented such that
the anterior bead site was toward the most elevated aspect of
the heart. This placed the posterior site generally below the
central left ventricular
axis in a lateral view. There was a
concern that the gravitational
weight
of the heart could
deform the left ventricle
and thereby alter the anteriorposterior mechanical
relationships.
To test this, in two hearts,
we rotated the heart-pump
system around its long axis and
compared the local deformations
of the anterior and posterior
sites at two angular orientations
-45’ apart. During systolic
ejection, strain data at each site showed negligible
differences
as a function of volume. During passive filling, however, an
effect of heart orientation
was noted. Thus, for the subsequent analysis, we focused on the systolic strain data.
Loadirzg
conditions.
Biplane
tine-X-ray
imaging
(at 90
frames/s > was performed
to track anterior-posterior
bead
positions over a wide range of cavity volumes and pressures
in the beating isolated heart. Even though the heart was
isolated, the cardiac cycle followed realistic pressure-volume
loops. Computer
feedback with simulated
arterial
loading
and venous filling conditions
governed the servo pump to
provide physiological
patterns of hemodynamic
loading ( 15 1.
End-diastolic
volumes
ranged
from 20 to 50 ml, stroke
volumes ranged from 7 to 30 ml, and ejection fractions were
20-607~.
Rwonstmcting
3-D deformations
fi-om cirwfilms. An automated image-analysis
system was used to transform the bead
positions recorded from the two orthogonal
sets of tine-X-ray
films into 3-D position coordinates
for each time frame. From
the 3-D trajectories
of each bead, a local measure of the 3-D
deformation
was computed individually
for each layer in both
the anterior and posterior sites based on the relative motions
of neighboring
beads. With the use of the method of Rodriguez
et al. ( II), a hybrid deformation
gradient tensor was used to
compute the deformation
on the plane defined by the three
beads implanted
in any one layer. One or more beads from an
adjacent layer were used to compute the out-of’-plane deformation components.
Histological
Analysis
L
and Sam~mez-c Rccorzstmctim
The next step involved
sectioning
the fixed heart for
histological
analysis of the local sarcomere pool. Thin (3-pm)
sections were cut at the location of each transmural
bead
triangle. Within each section, average SLs and orientations
were observed from multiple sites within the bead plane with
light microscopy
( X 100 objective).
Analysis
of orthogonal
sections provided the average orientation
of myocardial fibers
out of the plane of the beads. While the section at low power
was viewed, the coordinates
of the centers of the three bead
LENGTHS
implant locations were determined
with respect to the coordinate system used to measure sarcomere orientation
within
the bead plane.
A transformation
matrix was computed to relate the sarcomere orientation
and length, the “sarcomere
vector,” at the
locations of the bead planes at the epicardial
surface and
midwall
to those in the intact fixed heart before the hearts
were sectioned. This transformation
involved mapping
the
histological
bead locations to those in the intact heart before
fixation. Combining
this transformation
with the out-of-plane
angle from the orthogonal
sections allowed reconstruction
of
the average sarcomere vector in the intact fixed heart. The
description
of the evolution of the sarcomere vector over time
throughout
the cardiac cycle was then accomplished
by
applying to this sarcomere vector the deformation
between
the intact fixed heart and that observed from the 3-D trajectories of neighboring
beads during the various loading cycles.
Statistical
Methods
for Sarcomere
Deformation
Analysis
L
Experimental
results in this study were from eight mongrel dogs. Five other dogs were excluded from the analysis:
three hearts had weak contractile
strength and two showed
contracture
banding when viewed histologically.
Of the eight
valid experiments,
six hearts yielded epicardial
deformation
data that could be compared
between
the anterior
and
posterior sites, and five provided midwall deformation
data at
both sites. Unfortunately,
in only three hearts did bead
implantation
yield deformation
data simultaneously
at both
depths and both sites. Because comparison
between depths
was possible in only three hearts, we focused our analysis on a
comparison
of sarcomere dynamics between the anterior and
posterior walls.
We also focused our analysis on the behavior of SLs during
the systolic ejection phase. To obtain a strictly objective
criterion
for determining
the beginning
and the end of
ejection phase, we used only data from the period when the
volume was decreasing
over the central 90% of the volume
range covered in that cardiac cycle. Data were pooled together
from all ejections of the same heart, covering a wide range of
preload and afterload.
The chamber volume strongly affects SL, but the relationship between SL and volume (VI may vary at different sites
(anterior vs. posterior) or between different hearts tested. We
therefore performed
a two-way repeated-measures
analysis
of covariance (5 ). We used the following regression
equation:
SL r Constant + Volume + Wall + Heart + Volume. Wall +
Wall. Heart + Volume Heart + Volume. Wall Heart, where
Volume is a continuous
factor (i.e., a regressor) and Wall is a
discrete factor (anterior vs. posterior).
The regression coefficient for the term Volume *Heart indicates the average effect
of the anterior vs. posterior site on the SL-V relationship.
By
combining
the coefficients
for the terms
Volume
and
Volume. Heart, we obtained the average amount of sarcomere
shortening
accompanying
ejection (-1%/N).
The mean SL
during ejection (SL,,, > was obtained at each site by combining
the regression coefficients for the terms Constant and Wall.
So that this calculation
represented
an SL in the middle of the
ejecting range, the mean volume averaged over all ejections
for a single heart was subtracted
from the raw volume data
for that heart before applying the regression. Data for ML/N
and SL,,, are presented as means -+ SE. P < 0.05 was accepted
as statistically
significant.
A preliminary
analysis of a subset
of the data was performed previously (7 ).
l
l
ANTERIOR-POSTERIOR
LV SARCOMERE
RESULTS
Changes in ventricular
volume during the cardiac
cycle are clearly expected to be strongly
coupled with
changes in the local SL. Thus we plotted SL as a
function of left ventricular
volume and compared the
behavior of the sarcomeres
at anterior
and posterior
sites. Figure I shows epicardial
(A) and midwall
(B)
data throughout
the cardiac cycle in which the stroke
volume was maximal.
For example, midwall
sarcomeres at the anterior site (Fig. 1B) began their cardiac
cycle a t the maximal SL (upper right corn er of loop)
before shortening
somewhat
during
the isovol umic
cant raction phase. During the ejection phase (thick line
and arrow), the midwall anterior sarcomeres
shortened
steadily as volume fell so that the trajectory
was nearly
linear throughout
ejection. During
isovolumic
relaxation, these sarcomeres
were elongated slightly, and,
A
E
2.42.32.22.1
-
2.01.9
1.8
1
f
25
I
30
I
35
I
40
LV
E
2.4-
r^
5
s
i!i
2.3-
Volume,
I
45
I
I
1
50
55
60
ml
xi
!!
E
2.22.1-
0
2
ss
3
3
a.I
1
2.01.91.8
IO
I
15
I
20
I
I
25
30
LV Volume,
ml
I
35
I
40
I
45
Fig. 1. Typical
epicardial
(A) and midwall
(B) sarcomere
length vs.
left ventricular
(LV) volume
loops. Each loop traces simultaneous
progression
of sarcomere
length
and volume
throughout
a cardiac
cycle. Solid lines, trajectory
of changes in anterior
sarcomere
lengths;
dashed lines, posterior
sarcomere
lengths;
thick lines, ejection
phase.
Arrows,
sequence
in time of progression
around
loop. These data
correspond
to hemodynamic
loading
state in which stroke
volume
was maximal.
A and B were taken from different
hearts.
LENGTHS
H471
subsequently, they were restretched to their enddiastolic length during filling.
The data shown in Fig. 1 are typical in that sarcomeres at all sites and depths shortened steadily during
ejection (see arrows and the thick sections of each
trace). At either depth, the &SW& (i.e., the slope of the
approximately
linear trajectory during ejection) was
similar between the anterior and posterior sites. Moreover, the mean SL+ (taken at the middle of the ejecting
range) was also similar between the anterior and
posterior sites.
On the other hand, we observed no consistent effect of
site on the relationships between SL and cavity volume
during the other phases of the cardiac cycle, i.e., during
filling and the isovolumic phases. Note that the loops in
Fig. 1 all have different shapes, with different relationships between trajectories during ejection and filling.
Some of the diastolic differences may have been related
to differential effects of gravitational
forces between
the anterior and posterior sites. In our experiments,
the heart was situated with its long axis approximately
horizontal and oriented such that the anterior site
faced upward and the posterior site was located at
approximately four o’clock. If we were to rotate the left
ventricle around its long axis, we would then have
altered the differential offset of the gravitational field
at the two sites. As seen in Fig. 2, such rotation did
indeed affect that portion of the trajectories relating SL
to volume during diastolic filling, but there was no
significant effect during systolic ejection. Thus, for the
remainder of the data analysis, we focused exclusively
on the relationship between SL and V during ejection.
The results presented so far were for cardiac cycles in
which stroke volume was maximal, but we also compared regional SLs in hearts operating under a wide
range of hemodynamic loads. At any myocardial site,
the relationship between SL and V during ejection was
consistent irrespective of the hemodynamic loading
conditions that brought the left ventricle to that volume. This result is reminiscent of the strain-volume
relationship that Douglas et al. (2) also found to be
approximately independent of hemodynamic loads. Both
at the epicardium (Fig. 3) and at midwall (Fig. 4), the
relationships between SL and V during ejection for
different preloads and afterloads were approximately
linear, had similar slopes, and had comparable SLs at
matching volumes. All ejections of the same heart were
pooled together to obtain a slope (ZSL/UQ and intercept (SL,j >for each site.
XSL/SV values at the epicardial and midwall sites
were not significantly different between the anterior
and posterior walls (P > 0.28 and 0.44, respectively;
Table 1, Fig. 5). When averaged over all hearts, XSL/AV
was 7.3 rim/ml at the anterior epicardium and 8.3
rim/ml at the posterior epicardium. Individually, ilSL/lV
was 2.2-2.8 rim/ml greater at the posterior epicardium
than at the anterior epicardium in three hearts, differed less than 0.6 rim/ml between the two walls in two
hearts, and was 2.5 m-n/ml greater at the anterior
epicardium than at the posterior epicardium in one
heart. On average, XSLIAV was 8.9 rim/ml at the
H472
ANTERIOR-POSTERIOR
LV SARCOMERE
LENGTHS
A
Fig. 2. Effect of gravity on anterior
epicardial (A), posterior epicardial (B),
anterior midwall (C), and posterior midwall (II) sarcomere length vs. IX volume throughout cardiac cycle in which
stroke volume was maximal. At all 4
sites, relationship between sarcomere
length and volume before (solid lines)
and after (dashed lines) rotating heart
by 45’ about LV long axis is not significantly affected during ejection phase
(thick lines). Filling phase of cardiac
cycle was affected to the greatest extent, especially at anterior epicardium.
25
B
30
35
40
45
LV Volume,
50
55
60
25
30
ml
C
35
40
45
50
LV Volume,
ml
35
50
55
60
55
60
D
E 2.37
2
AC
%2.2f
ml
E 2.33,
2
g)2.2s
84
2 2.1-
25
30
35
40
45
LV Volume,
anterior midwall and 9.7 rim/ml at the posterior midwall. ASL/AV was 2.9-3.4 rim/ml greater at the posterior midwall than at the anterior midwall in two hearts,
differed less than 0.8 rim/ml between the two walls in
two hearts, and was 1.4 rim/ml greater at the anterior
midwall than at the posterior midwall in one heart.
Similarly,
there was no consistent anterior-posterior
difference in the mean SL, (P > 0.33 at the epicardium
and P > 0.96 at the midwall; Table 1, Fig. 6). When the
SL, was averaged over all hearts, it was 2.00 urn at the
anterior epicardium
and 2.06 urn at the posterior
epicardium.
Individually,
SL, was 0.11-O. 17 urn longer
at the posterior epicardium than at the anterior epicardium in three hearts, was 0.07-0.10 urn shorter at the
50
ml
55
60
25
30
40
45
LV Volume,
ml
posterior epicardium
than at the anterior epicardium
in two hearts, and differed less than 0.03 urn between
the two walls in one heart. On average, SL, was 2.01
urn at the anterior midwall and 2.01 urn at the posterior
midwall. SL, was 0.03-0.05 urn shorter at the posterior midwall than at the anterior midwall in two hearts,
differed less than 0.01 urn between the two walls in two
hearts, and was 0.07 urn longer at the posterior midwall than at the anterior midwall in one heart.
Hence we observed similar SL, values between the
anterior and posterior walls despite potential regional
differences in the myofiber strain. To maintain
such
comparable systolic lengths, sarcomeres in regions of
high systolic fiber stretch (relative to an unloaded
ANTERIOR-POSTERIOR
LV SARCOMERE
‘- 2.4
5
F 2.3
3
Q) 2.2
z
g 2.1
w”
_ Posterior
/
Anterior
10
20
30
40
LV Volume,
50
ml
60
70
8 O-
H473
LENGTHS
anterior epicardium in one heart. On average, the SLUnl
was 1.96 urn at the anterior midwall and 1.94 urn at the
posterior midwall. The SL,nl was 0.04-0.12 urn shorter
at the posterior midwall than at the anterior midwall in
two hearts, differed less than 0.02 urn between the two
walls in two hearts, and was 0.04 urn longer at the
posterior midwall than at the anterior midwall in one
heart.
A more revealing analysis would be to compare the
changes in SL between S1,,,, and average SL,j. At the
epicardium, the difference (SL,j - SL,nI > was significantly greater in the posterior wall than in the anterior
wall (P < 0.04; Table 1, Fig. 6). In fact, this was the case
in all six hearts studied (Fig. 6A). Moreover, SL, SL Lln1was greater at the posterior epicardium than at
the anterior epicardium both before and after rotation
of the left ventricle around its long axis. At the midwall
(Fig. 6B), however, there was no consistent anteriorposterior difference in SL,j - SL,nj (P > 0.55).
In contrast to the “unlo’aded” diastolic state, we also
compared anterior vs. posterior end-diastolic SLs at the
maximal volume (EDSL,,,,) we achieved in each heart.
The EDSL,,,,, values at both the epicardial and midwall
6 O-
A
4 oE
2 onu
2.4
z!L
n
n
I
I
IO
20
30
I
LV
40
Volume,
I
50
ml
I
60
5
2.3
z
1.8
1
1
70
Fig. 3. Epicardial
sarcomere
lengths on anterior
and posterior
walls
during
ejection
(A) and corresponding
LV pressures
(B) plotted
as
functions
of LV volume
for a wide range of end-diastolic
and stroke
volumes
in a typical heart (same heart as Fig. 1A ). A: dashed lines,
linear regressions
for all ejections;
solid symbols,
mean sarcomere
length during
ejection;
open symbols,
sarcomere
length at middle of
filling
at lowest
preload;
circles,
anterior
wall; squares,
posterior
wall. B: 0, pressure-volume
data point at middle of filling at lowest
preload.
ventricular configuration) require shorter sarcomeres
when the left ventricle is unloaded than do sarcomeres
in regions of low-systolic fiber stretch. For an approximately unloaded ventricular state, we chose the configuration at the middle of filling at the lowest preload of
each heart. The SL in this state is called SL,,,]. Note
that left ventricular
pressure in this state was -4
mmHg, so that it does not represent a completely
unloaded reference state. However, this state does
represent the lowest load in terms of pressure available
in our data.
SL unl values at the epicardial and midwall sites were
not significantly different between the anterior and
posterior walls (P > 0.11 and 0.39, respectively; Table 1,
Fig. 6). When the SL,nI was averaged over all hearts, it
was 2.01 urn at the anterior epicardium and 1.94 urn at
the posterior epicardium. Individually,
the SL,nl was
0.04-0.14 urn shorter at the posterior epicardium than
at the anterior epicardium in five hearts and was 0.08
urn longer at the posterior epicardium than at the
IO
I
I
15
20
I
I
I
25
30
LV Volume,
ml
25
30
Volume,
ml
I
1
35
40
4 5
35
40
45
160
E 100
!f
;
80
ii-i
k
60
10
15
20
LV
Fig. 4. Midwall
sarcomere
lengths
on anterior
and posterior
walls
during
ejection
(A) and corresponding
LV pressures
(B) plotted
as
functions
of LV volume
for a wide range of end-diastolic
and stroke
volumes
in a typical heart (same heart as Fig. 1B). Symbols
same as
in Fig. 3.
H474
ANTERIOR-POSTERIOR
LV SARCOMERE
Table 1. Comparison
of sarcomere lengths and
shortening
between anterior and posterior sites
Significance
2&L I N
Epicardial
(n = 6)
Anterior,
run/ml
Posterior,
r-m-i/ml
Midwall
(12 = 51
Anterior,
r-m/ml
Posterior,
rim/ml
Posterior
SL,,
Epicardial,
urn
Midwall,
pm
SLmI
Epicardial,
urn
Midwall,
urn
SL, - %nl
E‘picardial,
urn
Midwall,
urn
EDSL,,,,
Epicardial,
urn
Midwall,
urn
7.3 i 1.1
8.3 t 0.5
NS
8.9 + 1.5
9.7? 1.0
NS
-anterior
0.05 -+ 0.05
0.00 -e 0.02
-0.06
-0.03
t 0.03
t 0.03
NS
NS
NS
NS
0.12 5 0.04
0.03 + 0.04
P < 0.05
NS
0.02 -+ 0.04
0.02 + 0.06
NS
NS
Values
are means
-+ SE; n2, no. of hearts.
XSL/.W,
sarcomere
shortening
per ejected volume;
SL,,, mean sarcomere
length during
ejection;
SL,,,ll, mean
sarcomere
‘length
during
filling
at lowest
preload;
SL,, - SL,,,,,, difference
between
SL,,, and SL,,,,,; EDSL,,,,,, ,
end-diastolic
sarcomere
length
at greatest
preload;
NS, not significant. Significance
was for test of anterior
vs. posterior
data. All tests
used analysis
of covariance
with volume
as a regressor,
except when
comparing
difference
in SL,,,,, , which used analysis
of variance.
sites were not significantly
different between the anterior and posterior
walls (P > 0.88 and 0.79, respectively; Table 1). When the EDSL,,,
was averaged over
all hearts, it was 2.19 urn at the anterior epicardium vs.
2.17 urn at the posterior
epicardium.
At the midwall,
the EDSL,,,,
averaged 2.18 urn anteriorly
compared
with 2.20 urn posteriorly.
LENGTHS
table result. In fact, two previous
studies (3, 17) reported that myocardial
strain during a cardiac cycle
evolved differently
at anterior vs. posterior sites. Even
in the hearts studied here, there were often conspicuous anterior-posterior
differences
in epicardial
sarcomere length changes during the transition
from rest to
activity in the isovolumic
contraction
phase. We suspected that anterior-posterior
differences in strain might
be related to anterior-posterior
differences in the structure of the left ventricular
wall, in particular,
differences in wall curvatures.
When Guccione et al. (7) built
such differences
of curvature
into a mathematical
model, it indeed predicted that epicardial SLs during
ejection would differ between the anterior and posterior
sites if SLs were uniformly
distributed
when the left
ventricle was in its unloaded passive configuration.
Thus the anterior-posterior
similarity
of ejecting
sarcomere
function
that we observed
must develop
despite inherent
mechanical
differences
between
the
anterior and posterior regions of the left ventricle. This
suggests that the ability of cardiac cells to alter their
structure
in response to mechanical loads allows sarcomere function during ejection to be nearly equilibrated
A
0
‘E
W
DISCUSSION
0
1
Anterior
This study demonstrated
that sarcomeres
at anterior
vs. posterior
sites functioned
similarly
during ejection
in isolated canine hearts. The amount of sarcomere
shortening
accompanying
left ventricular
ejection was
similar in both walls of the left ventricle for sarcomeres
located just below the epicardium
and also for those
near the midwall. The mean SL (taken at the middle of
the ejecting range) was also similar between the anterior and posterior sites when averaged over all hearts.
The similarity
of sarcomere
function held not only at
end systole but throughout
ejection and over wide
ranges of ventricular
pre- and afterloads.
This report presents the first observations
of sarcomere behavior
during
active ejections
of an intact
heart. Moreover,
this report also presents
the first
comparison
of lengths between sarcomeres
from different regions of the same heart. Surprisingly,
although
there have been previous observations
of SLs in fixed
hearts, no regional and only two transmural
comparisons (6,12) have been reported.
The regional similarity
of sarcomere function during
ejection that we observed was not necessarily
an inevi-
I
1
Posterior
I
LV Wall
B
2 o-*
E
$
1 5-
s
?
J 1 o2
=
g
5a.I
>
0-c I
*
a
t
I
Anterior
I
I
Posterior
I
’
LV Wall
Fig. 5. Amounts
of sarcomere
shortening
per ejected volume
USL/
JW in anterior
and posterior
LV free walls at epicardium
(A) of 6
hearts
and at midwall
(B) of 5 hearts.
Each symbol
represents
a
different
heart.
ANTERIOR-POSTERIOR
LV SARCOMERE
0
Ial
I
-----
--9
a
I
----Q
-0.2
Fig. 6.
mean
length
between
and at
heart.
)
I
sL~nI
I
sLej
SLej
I
- sLuni
1
Variations
between
anterior
(A) and posterior
(PI walls in
sarcomere
length
during
ejection
(SL,., 1, mean
sarcomere
during
filling
at smallest
preload
(Sill,,+,
and difference
SLlj and SL,,,,, (SL,,, - SL,,,,, 1 at epicardium
(A ) of 6 hearts
midwall
(B) of 5 hearts.
Each symbol
represents
a different
between re gions via myocyte adaptation.
This is espefunction (for
cially so if ej ecti w aspects of sarcomere
example, sarcomere shortening
and the mean SL,, > are
the important
controlling
variables
feeding back to
modify myocyte structure.
In fact, cell stretching
has
been shown to induce adaptations
in isolated myocytes
(9, 13, 18). More specifically, in a theoretical
construct,
Arts et al. (1) showed that including
both SL at the
beginning of ejection and IISL/AV as controls for myocyte growth produced an adapting structure
that reacted to altered hemodynamic
loads much as observed
in vivo.
If local myocyte adaptation occurs such that anteriorposterior
sarcomere
function
is equilibrated
during
ejection, then the mechanical differences between anterior and posterior sites [as incorporated
into the mathematical model (7>] must induce sarcomere nonuniformity when the left ventricle is passive and externally
unloaded. This prediction was confirmed by our experimental data. Note that the model answers
the inverse
question, What sarcomere behavior would occur if SLs
were uniform at all sites in the passive unloaded state?
We chose this question because the actual distribution
of SLs throughout
an unloaded ventricle is not known
presently
and because analytic techniques
to incorpo-
LENGTHS
H475
rate such differences
in finite-element
models are just
in development,
although
they have been used for
simple cylindrical
models (8). Thus, based solely on the
differences
in geometry between the two walls of the
ventricle, the mathematical
model predicted that sarcomeres would work during ejection at consistently
longer
lengths in the epicardium
of the posterior
site compared with the anterior site.
However, we hypothesize
that if cells in the posterior
epicardium
were chronically
forced to work at extended
lengths, they would be stimulated
by their mechanical
environment
to lay down more sarcomeres
in series
than their anterior counterparts,
so that ejecting SLs
could be equilibrated
between
the two walls. Consequently, when the passive ventricle
is then unloaded,
sarcomeres
in the posterior
epicardium
would become
shorter
than anterior
ones as the greater number in
series are compressed.
This was, in fact, observed in
our data. In every heart, wall mechanics caused sarcomeres to shorten
more posteriorly
in the transition
from a midejecting to an unloaded configuration
(SL,.i SL unl in Fig. 6), and, as a result, the epicardial sarcomeres in the unloaded state were shorter posteriorly
in
five of six hearts.
At midwall
depth, the mathematical
model (7) predicted that ejecting SLs would be similar at both the
anterior
and posterior
sites (within
0.05 urn) if SLs
were equal in the unloaded state. When the train of
thought above is followed, it predicts that the mechanical environment
surrounding
midwall myocytes would
supply little adaptive stimulus
to alter SLs differentially between the anterior and posterior
sites. Consequently, we would predict that anterior
and posterior
midwall
SLs would remain similar
in an unloaded
heart even after myocyte adaptation was allowed. This
was, in fact, the general trend observed (Fig. 6).
If differential
growth
by myocytes
adapts them so
that sarcomere
function
tends toward uniformity
for
the ejecting conditions,
then differential
compressions
and tensions within the ventricular
wall are likely to be
present when such an adapted left ventricle
is unloaded passively, even though there is no net transmural pressure gradient distending
the ventricle.
Such a
ventricle is then said to contain residual stresses (4).
These can be revealed by judiciously
cutting apart an
unloaded ventricle, which then deforms as the residual
stresses
are relieved.
Such deformations
are called
residual strains,
which were studied in rat hearts by
Omens and Fung ( 10). One consistent
result from that
study (Fig. 7 in Ref. 10) was the difference in residual
strains between the anterior and posterior
sites on an
equatorial ring from the left ventricle. For example, the
circumferential
component of residual strain was significantly greater anteriorly,
which implies that a circumferentially
oriented line segment was elongated significantly more on the anterior
side of the unloaded
ventricle
compared with its length in the stress-free
sectioned
state attained
after
releasing
residual
stresses.
The anterior-posterior
differences
in residual strain
that Omens and Fung (10) noted are quantitatively
H476
ANTERIOR-POSTERIOR
compatible with the anterior-posterior
differences in SL
that we observed under our unloaded conditions.
Before we could make such a comparison,
we needed to
estimate
SLs in the stress-free
configuration.
After
residual stresses have been removed by sectioning an
unloaded ventricle,
it then seems likely that all SLs
would be uniform. Rodriguez et al. (12) have confirmed
this by showing that after residual stress is removed
from rat ventricles,
the transmural
distribution
of SLs
becomes uniform at 1.84 urn. Combining
the data of
Rodriguez et al. with that of Omens and Fung (lo), one
can thus predict SLs in unloaded passive rat hearts at
both the anterior
and posterior
sites. In making this
prediction,
we assumed that epicardial
and midwall
fibers were oriented at the average fiber angles we
observed and that the longitudinal
principal residual
strain could be derived by assuming tissue incompressibility. At the epicardial layer, anterior sarcomeres
were
predicted to be 1.88 urn compared with 1.80 urn posteriorly. This 0.08~urn elongation
of anterior
epicardial
sarcomeres
predicted from residual strains measured
in rat hearts compares favorably with the elongation of
0.06 urn that we observed in our unloaded state. At the
midwall,
the difference in predicted SLs for unloaded
rat hearts was much smaller
(anterior
= 1.83 urn;
posterior
= 1.80 urn), identical
to the insignificant
0.03~urn difference that we found. However,
note that
all absolute SLs were somewhat
longer in our unloaded
state (e.g., anterior epicardium
= 2.01 urn; posterior
epicardium
= 1.94 urn). This is presumably
because we
never fully achieved an unloaded state (chamber pressures in our unloaded state were still positive),
and
thus the sarcomeres
were somewhat
distended beyond
their truly unloaded lengths.
All of our data on SLs during ejection were obtained
after chordal transection;
on the other hand, myocytes
would have adapted with the chordae intact. A recent
preliminary
study (16) on two open-chest anesthetized
dogs suggests that, at end systole, chordal transection
increases the longitudinal-radial
shear in the anterior
papillary
muscle and inner wall. Although
chordal
transection
certainly
caused some rearrangement
in
cardiac geometry, there are several reasons to expect
that it may not have meaningfully
altered our anteriorposterior comparison.
1) The regions we studied did not
directly
overlie papillary
insertion
points;
we implanted beads near the ventricular
equator (more basal
than papillary roots), and the circumferential
location
was slightly more septal (i.e., just lateral to the anterior
and posterior
descending
coronary
arteries).
2) The
major difference we observed occurred at the epicardium, and we did not examine strains from the inner
half of the wall. 3) Only a differential
effect (e.g., if one
papillary muscle were stronger than the other) is likely
to have influenced
our anterior-posterior
comparison.
Note that the geometry of our finite-element
model for
ventricular
mechanics
(7) was also obtained
after
chordal transection.
Further
limitations
of our SL data should be noted.
Rodriguez et al. (11) estimated that the typical peak-topeak noise that is associated
with reconstructing
the
LV SARCOMERE
LENGTHS
SL was on the order of 0.04 urn. This is ~13% of the
typical amount of sarcomere shortening
over the ejecting range (i.e., 0.3 urn; see Figs. 3A and 4A). Hence our
reconstruction
approach might not have been able to
detect small differences in sarcomere shortening
of this
order. Moreover, the bead array used here consists of
approximately
two layers at least 5 mm apart, which
exceeds the typical separation in the previous report by
Rodriguez et al. Due to potential transmural
inhomogeneities in myocardial strain, the greater the separation,
the more error that could be introduced
during sarcomere reconstruction.
Rodriguez et al. found an uncertainty of co.02 urn when fiber orientation
deviated
from the plane of markers
by <lo”. In this study, all
midwall layers satisfied this angular criterion,
but the
larger
separation
may have introduced
somewhat
greater uncertainty
due to transmural
inhomogeneity
of strain.
Although
we primarily
examined SLs during ejection, uncertainties
in the diastolic configuration
of the
isolated hearts and potential but unknowable
differences in diastolic
suspension
of the heart between
experiment
and in vivo may have affected our ability to
interpret
data (such as SL,,,,I and EDSL,,;,, in Table 1)
that were derived during diastole. Because passive
myocardium
becomes
dramatically
stiffer
as it is
stretched, we expect less influence from diastolic suspension on EDSL,,,;,, than on any other diastolic
SL.
Moreover, although the magnitude
of SL,,,,, was clearly
affected when we altered diastolic suspension
by rotating the isolated heart, SL,,,,, at the anterior epicardial
site remained
significantly
longer than that at the
posterior site.
In summary,
sarcomeres
at anterior
vs. posterior
sites in isolated canine hearts functioned
similarly
throughout
ejection over wide ranges of ventricular
pre- and afterloads. A mathematical
model for ventricular mechanics (7) predicted that epicardial SL,i values
would differ between the anterior and posterior
sites if
SLs were uniformly
distributed
when the left ventricle
was in its unloaded passive configuration.
Thus the
anterior-posterior
similarity
of ejecting sarcomere function that we observed must develop despite inherent
mechanical differences between the anterior and posterior regions of the left ventricle. This suggests that the
ability of cardiac cells to alter their structure
in response to mechanical
loads allows sarcomere function
during ejection to be nearly equilibrated
between
regions via myocyte adaptation.
Such systolically
governed adaptation, however, will lead to nonuniform
SLs
(residual strains) in the passive unloaded ventricle.
The authors
thank
Dr. James H. Anderson,
Michael
Samphilipo,
and Carolyn
Magee for assistance
with the biplane
cineradiographic
procedures.
We also appreciate
the help of Dr. Willard
Graves
and
David Mearns
for assistance
with the marker-tracking
software.
We
also thank
Dr. Brian K. Slinker
for expert
advice on the statistical
analysis
of our data. Sincere
thanks
also go to Ken Rent,
who
provided
invaluable
surgical
assistance,
to Eric Lee and Beth Jones,
who spent many hours tracking
markers
on film, and to Michele
K.
Leppo for performing
the histological
procedures
and sectioning.
This work
was supported
by National
Heart,
Lung,
and Blood
Institute
Grant
HL-30552
(to W. C. Hunter).
ANTERIOR-POSTERIOR
J. M. Guccione
was the recipient
of Individual
National
Research
Service Award
lF32-HL-08492
from the National
Heart,
Lung, and
Blood Institute.
This support
is gratefully
acknowledged.
Address
for reprint
requests:
J. M. Guccione,
Washington
Univ.
Campus
Box 1185, One Brookings
Dr., St. Louis, MO 63130-4899.
Received
14 February
1996;
accepted
in final
form
2 July
1996.
REFERENCES
1. Arts, T., F. W. Prinzen,
L. H. E. H. Snoeckx,
J. M. Rijcken,
and R. S. Reneman.
Adaptation
of cardiac structure
by mechanical feedback
in the environment
of the cell: a model study.
Biophys.
J. 66: 953-961,
1994.
Douglas,
A. S., E. K. Rodriguez,
W. G. O’Dell,
and W. C.
Hunter.
Unique
strain
history
during
ejection
in canine
left
ventricle.
Am. J. Physiol.
260 (Heart
Circ. Physiol.
29): Hl596H1611,1991.
Fann,
J. I., G. E. Sarris,
N. B. Ingels,
Jr., M. A. Niczyporuk,
K, L. Yun, G. T. Daughters
II, G. C. Derby,
and D. C. Miller.
Regional
epicardial
and endocardial
two-dimensional
finite deformations
in canine left ventricle.
Am. J. PI2ysioZ.
261 (Heart
Circ.
Physiol.
30): H1402-H1410,
1991.
Fung,
Y.-C. Biod.ynamics:
Circulation.
New York:
SpringerVerlag, 1984.
Glantz,
S. A., and B. K. Slinker.
Primer
of Applied Regression
and Analysis
of Variance. New York: McGraw-Hill,
1990.
Grimm,
A. F., H. L. Lin, and B. R. Grimm.
Left ventricular
free wall and intraventricular
pressure-sarcomere
length distributions.
Am. J. PhysioL.
239 (Heart
Circ. Physiol.
8): HIOIH107,1980.
Guccione,
J. M., A. D. McCulloch,
and W. C. Hunter.
Three-dimensional
finite element
analysis
of anterior-posterior
variations
in local sarcomere
length
and active
fiber stress
during
left ventricular
ejection.
In: Advances
in Rioengineering,
edited by J. M. Tarbell.
New York: American
Society of Mechanical Engineers,
1993, vol. 26, p. 571-574.
8. Guccione,
J. M., A. D. McCulloch,
and L. K. Waldman.
of intact
ventricular
mvocardium
Passive
material
DroPerties
LV SARCOMERE
9.
10.
11.
12.
13.
14
15.
16.
17.
18.
LENGTHS
H477
determined
from a cylindrical
model.
J. Biomech.
E12g.
113:
42-55,199l.
Komuro,
I., T. Kaida,
Y. Shibazaki,
M. Kurabayashi,
Y.
Katoh,
E. Hoh, F. Takaku,
and Y. Yazaki.
Stretching
cardiac
myocytes
stimulates
protooncogene
expression.
J. Biol. Chem.
265: 3595-3598,199O.
Omens,
J. I-I., and Y. C. Fung.
Residual
strain
in rat left
ventricle.
Circ. Res. 66: 37-45,
1990.
Rodriguez,
E. K., W. C. Hunter,
M. J. Royce,
M. K. Leppo,
A. S. Douglas,
and H. K. Weisman.
A method
to reconstruct
myocardial
sarcomere
lengths
and orientations
at transmural
sites in beating
canine hearts. Am. J. Physiol.
263 (Heart
Circ.
Ph:ysioZ. 32 ): H293-H306,
1992.
Rodriguez,
E. K., J. H. Omens,
L. K. Waldman,
and A. D.
McCulloch.
Effect of residual
stress on transmural
sarcomere
length
distributions
in rat left ventricle.
Am. J. Physiol.
264
(Heart Circ. Physiol.
33 ): H 1048-H
1056, 1993.
Sadoshima,
J., L. Jahn,
T. Takahashi,
T. J. Kulik,
and S.
Izumo.
Molecular
characterization
of the stretch-induced
adaptation of cultured
cardiac cells: an in vitro model of load-induced
cardiac hypertrophy.
J. Biol. Chem. 267: 10551-10560,
1992.
Suga, H., and K. Sagawa.
End-diastolic
and end-systolic
ventricular
volume clamper
for isolated
canine heart. Am. J. Physiol.
233 (Heart
Circ. Ph.ysiol. 2): H718-H722,
1977.
Sunagawa,
K., D. Burkhoff,
K. 0. Lim,
and K. Sagawa.
Impedance
loading
servo pump system
for excised canine ventricle. Am. J. Physiol.
243 (Heart Circ. Physiol.
12): H346-H350,
1982.
Takayama,
Y., I. J. LeGrice,
J. W. Holmes,
and J. W. Covell.
Effects
of chordal
uncoupling
on deformation
in the papillary
muscle (Abstract).
FASEB J. 8: A591, 1994.
Villarreal,
F. J., and W. Y. W. Lew. Finite
strains
in anterior
and posterior
wall of canine left ventricle.
Am. J. PhysioZ.
259
(Heart Circ. Physiol.
28): H1409-H1418,
1990.
Yazaki,Y.,
I. Komuro,
T.Yamazaki,
K. Tobe, K. Maemura,
T.
Kadowaki,
and R. Nagai.
Role of protein
kinase system in the
signal transduction
of stretch-mediated
protooncogene
expression and hypertrophy
of cardiac
myocytes.
Mol. Cell. Biochem.
119: 11-16,
1993.