Download Fiber Types and Myosin Types in Human Atrial and Ventricular

794
Fiber Types and Myosin Types in Human Atrial and
Ventricular Myocardium
An Anatomical Description
Patrice Bouvagnet, Jocelyne Leger, Francoise Pons, Claude Dechesne, and Jean J. Leger
From the Institut National de la Sante et de la Recherche Medicale, Laboratoire de Pathologie Moleculaire du Muscle, Institut de
Biologie, Montpellier, France
SUMMARY. Hybridomas were prepared from mice immunized with myosin from the enlarged
left ventricle of a 53-year-old female with an obstructive cardiomyopathy. The specificity of 15
monoclonal antibodies to myosin heavy chains was assessed by the reactivity of muscle extracts
and of chymotryptic myosin fragments of different sizes with these antibodies, as determined by
the immune replicate technique; some of the monoclonal antibodies cross-reacted only with the
ventricular V3-type myosin from hypothyroid rats, whereas the other antibodies cross-reacted
both with the latter and with the ventricular VI-type myosins from normal young rats. Immunological heterogeneity of the fibers from human atrial muscles and from human ventricular
muscles was detected by some of the antimyosin antibodies by means of indirect immunofluorescence. Histochemical fiber heterogeneity was also detected by adenosine triphosphatase staining
of the same tissues. Because of the close correspondence observed between the immunological
and histochemical responses of atrial fibers, it has been postulated that at least two distinct types
of myosin exist in the human atrium, each myosin form being histochemically related to either
a- or /3-like ventricular myosin heavy chains. In contrast, there was no direct correspondence
between the two experimental approaches in human ventricles, and it is postulated that at least
three distinct types of myosin exist within the human ventricles, one VI-type myosin, presumably
corresponding to the very rare fibers with an alkaline-stable adenosine triphosphatase activity,
and two other V3-type myosins corresponding to immunologically different fibers, each having
an alkaline-labile adenosine triphosphatase activity. Monoclonal antibodies that can distinguish
among the different myosin variants were further used to provide the basis for an anatomical
description of fiber types and myosin types within the human atrial and ventricular myocardium
in the whole hearts of two young boys who died sudden violent deaths. Small zones of myosin
variation were seen to be scattered, but probably not randomly distributed, within large areas of
myocardium in which the cellular distribution of myosin was constant; the large areas had one
myosin distribution specific for each cardiac cavity. No clear-cut conclusions can yet be made
concerning the physiological role of the regional variations observed in the distribution of the
different molecular forms of myosin (Circ Res 55: 794-804, 1984)
CARDIAC myosins exhibit molecular heterogeneity
within the ventricles and atria of some mammals,
according to evidence derived from several different
experimental approaches (Hoh et al., 1978; Flink
and Morkin, 1979; Klotz et al., 1981; Lompre et al.,
1981; Sartore et al., 1981; Chizzonite et al., 1982;
Gorza et al., 1982; Litten et al., 1982; Banerjee,
1983). These molecular variants of myosin heavy
chains play a role in controlling myocardial function,
both under normal conditions of development and
during heart disease (Schwartz et al., 1981; Rupp,
1982; Swynghedauw and Delcayre, 1982). Regional
variations in muscle fiber distribution, corresponding to different distributions of molecular variants
of myosin heavy chains, have been detected in beef
atrial and ventricular myocardium by immunofluorescence: it has been suggested that certain muscle
fibers having a specific myosin content may be
specialized for faster conduction (Sartore et al., 1981;
Gorza et al., 1982).
Several recent publications mention the possible
existence of molecular heterogeneity in human cardiac myosins, but relatively less work has been done
on myosins in human hearts. Human ventricular
and atrial myosins have different enzymatic activities and different light chain contents (Klotz et al.,
1983). Sequence microheterogeneities have been detected in the light chain 2 of human ventricular
myosin (Klotz et al., 1982), and variations in light
chain contents under pathophysiological conditions
have been seen in human atria (Cummins, 1982)
and in human ventricles (Tuschmid et al., 1983).
Human ventricular and atrial myosins have different
heavy chains, as shown by peptide mapping (Klotz
et al., 1983), but migrate as a unique and identical
band under nondissociating conditions (Lompre et
Downloaded from http://circres.ahajournals.org/ by guest on May 12, 2016
Bouvagnet et a/./Isomyosins in Human Atria and Ventricles
al., 1981; Clark et al., 1982; Klotz et al., 1983).
Human ventricular myosin reacts four to six times
more with monoclonal antibodies specific for the V3
myosin heavy chains of embryonic chicken ventricle
than with monoclonal antibodies specific for the VI
myosin heavy chains of adult rabbit ventricle (Clark
et al., 1982). Schiaffino et al. (1983) and Mercadier
et al. (1983) have shown that normal human heart
contains 0-15% of total ventricular myosin which
reacts with antibodies to the rat and beef VI myosin
isoform; no similar information is available concerning human atrial myosin. A recent histochemical
study of normal human heart has conclusively
shown that heterogeneity with respect to fiber
ATPase activity exists not only between atrial and
ventricular myocardia but also within normal atrial
and ventricular myocardium (Thornell and Forsgren,
1982).
A rather new experimental method for studying
molecular variations in any molecule is the use of
monoclonal antibodies; these permit a detection of
even slight antigenic differences between closely
related molecules, and the hypothetically heterogeneous molecule under study can be used as an
immunogen. This method has recently been used to
detect previously known and currently unknown
molecular heterogeneities of myosins in certain animal muscles (Bader et al., 1982; Chizzonite et al.,
1982; Winckelmann et al., 1983; Miller et al., 1983;
Wachsberger et al., 1983). Thus, we prepared hybridomas from mice immunized with myosin from
the left ventricle of a hypertrophic human heart and
observed that muscle fiber heterogeneity can be
detected by indirect immunofluorescence within
both human atrial and ventricular myocardium. This
muscle fiber heterogeneity detected by antimyosin
heavy chain antibodies is variably distributed in
each cavity of the human heart. We compare it to
that detected by histochemistry in beef, human, and
rat hearts, by immunofluorescence in rat ventricles,
and by other methods in human hearts. We discuss
the functional significance of this variable cellular
distribution of the different molecular forms of atrial
or ventricular myosins in terms of their location in
human heart.
795
ventricle and the left atrium were frozen directly in liquid
nitrogen and stored at —80°C, if not immediately used.
Serial tissue sections (6-8 /xm) were cut with a cryostat at
—20°C and either processed for immediate immunofluorescence and histoenzymological studies or stored at
—80°C until use. Identical results were observed under
both conditions.
The anatomical description of fiber types and myosin
types within the human heart was made on the whole
heart of two 16-year-old boys who died sudden violent
deaths. Seventeen tissue samples, each about 1 cm long,
were excised from different regions of the hearts immediately after death (Fig. 1). Four samples came from the
right atrium and three from the left atrium; six samples
came from the right ventricle and four from the left
ventricle. Two atrial samples (3 and 12) came from regions
of postulated preferential conduction pathways (James
and Sherf, 1971); samples 2 and 13 came from the right
and left auricle, respectively, whereas the three other
samples (1, 4, and 11) came from common atrial myocardium. Two ventricular samples (5 and 6) came from regions of postulated preferential conduction pathways,
RA
Methods
Tissue Sources
Adult human hearts were obtained either within 12
hours after death or from renal transplant donors (Nephrology and Urology Units, Hopital Saint-Charles, Montpellier). Prior to surgery, informed consent from the donor's family was obtained.
Myosin for mouse immunization was prepared (Klotz
et al., 1981) from the hypertrophic left ventricle of a 53year-old female who had an obstructive cardiomyopathy
(heart weight: 690 g; degree of hypertrophy: about 200%).
Immunofluorescence studies were routinely made on
fragments from autopsied adult hearts weighing 350 g or
less and apparently free of any cardiac disease. Tissue
samples excised from the posterolateral wall of the left
L
V
LA
FIGURE 1. Location of excised tissue samples (Netter, 1974). Upper
panel: right atrium (RA) and ventricle (RV) opened and viewed from
the front. Lower panel: left ventricle (LV) flap opened in the posterolateral wall and sectioned left atrium (LA) from a posterolateral view.
Specimens were excised from: RA, 1, anterior trabeculated wall; 2,
auricle; 3, crista terminalis; 4, interatrial septum above the limbus of
the fossa ovalis. RV, 5, interventricular septum below the supraventricular crest; 6, septa! band; 7, anterior papillary muscle; 8, anterior
free wall; 9, conus arteriosus; 10, interventricular septum near the
apex. LA, 11, posterior free wall; 12, superior free wall; 13, auricule.
LV, 14, mid-portion of the muscular interventricular septum; 15,
mid-portion of the posterolateral free wall; 16, anterior papillary
muscle; 17, apex.
Downloaded from http://circres.ahajournals.org/ by guest on May 12, 2016
Circulation Research/Vol. 55, No. 6, December 1984
796
and 15 came from free walls, and samples 10 and 17 came
from the apex, whereas sample 9 was excised from the
conus arteriosus located near the pulmonary artery (Fig.
1). All tissue samples were prepared for immunofluorescence and histoenzymological studies as already described.
Normal Wistar rats (6 weeks old) and hypothyroid rats
of the same strain (5 mg of propylthiouracil forcibly fed
daily for 4 weeks) were killed by a sharp blow on the
neck. The hearts were rapidly removed and processed like
human hearts.
Monoclonal Antibodies
Mice (strain CB 20 or Biozzi HR) were immunized with
myosin from a stock solution prepared from a hypertrophic left ventricle (see tissue sources). This myosin,
used as an immunogen, was previously denatured with
2% sodium dodecyl sulfate (Schwartz et al., 1980) and
then dialyzed in 50 mM sodium pyrophosphate, pH 7.5.
The mice were immunized twice subcutaneously with 100
/xg of myosin emulsified in Freund's adjuvant at intervals
of a few weeks. The mice were boosted intraperitoneally
with 100 ng myosin 3 days before the fusion. Isolated
splenocytes were fused with a P3-X63-Ag8-653 mouse
myeloma in the presence of polyethylene glycol, as described by Di Pauli and Raschke (1978). The supernatants
of 1024 hybridoma cell cultures from two fusions were
screened for human ventricular myosin with a direct solid
phase radioimmunoassay (RIA). Most of the positive hybridomas were not immediately cloned, but were kept in
culture in RPMI 1640 with 15% fetal calf serum. A few
cell lines which appeared to be of special interest for the
detection of myosin heterogeneity (see Results) were additionally cloned using a fluorimetric cell sorter; the corresponding immunoglobulins were purified from the ascite fluid of mice injected by each hybridoma line by
passage through protein A Sepharose columns. For use in
immunofluorescence studies, culture supernatants or purified ascite fluids were pooled and their IgG content was
precipitated by the addition of ammonium sulfate at a
final concentration of 50%; the collected pellets were
solubilized in 150 mM NaCl, 10 mM sodium phosphate,
pH 7.4 (PBS), and stored at -20°C until use.
The direct solid phase RIA used for screening cell line
supernatants for immunoglobulin production and for subsequent antibody titrations was processed as follows: an
excess of human ventricular myosin (50 ttl at 20 ttg/ml in
0.5 M NaCl, 50 mM sodium pyrophosphate, pH 7.4) was
coated in each well of a 96-well polyvinyl microtiter plate.
Fifty microliters of each hybridoma medium were then
incubated for 3 hours at 25°C with the coated myosin.
After washing, I25I-Iabeled goat antimouse IgG was added
and incubation continued for 3 hours at 25°C. After
washing, the radioactivity contained in each well was
counted. A cell culture supernatant containing no antimyosin antibody was used as a control. The competitive
RIA for testing the specificity of any antibody to different
human myosins was essentially the same as described by
Winckelman et al. (1983).
To test the specificity of the antibodies to the myosin
heavy chains, total muscle extracts in SDS buffer (Chizzonite et al., 1982), myosin samples, and chymotryptic
treated myofibril samples (Klotz et al., 1981) were first
separated by electrophoresis in SDS PAGE using the
buffer system of Laemmli (1970). The protein bands were
then electrophoretically transferred to nitrocellulose paper
according to the method of Towbin et al. (1979) and, after
incubation with each cell line supernatant, the immune
bands were revealed, using rabbit antimouse IgG coupled
with peroxidase (Matus et al., 1980).
Immunofluorescence Study of Heart Cells
Indirect immunofluorescence was performed on transverse sections of frozen cryostat tissue samples from different regions of the human heart with appropriate dilutions of the antibody solutions, according to the method
of Gorza et al. (1982). Atrial and ventricular sections were
processed with fluorescein-labeled rabbit antimouse IgG
specially prepared to avoid nonspecific cross-reaction
(Nordic Immunologic Laboratories). The slides were
mounted in a mixture consisting of an equal volume of
PBS and glycerol. Sections were examined with a Leitz
Orthoplan microscope equipped with epifluorescence optics. Controls for immunofluorescence included sections
incubated with a cell culture supernatant either free of
monoclonal antibody or containing a monoclonal antibody
to renin (a generous gift from B. Pau).
Enzyme Histochemistry of Heart Cells
Serial sections adjacent to those used for immunofluorescence studies were stained for myofibrillar ATPase
(myosin adenosine triphosphatase, calcium activated, EC
3.6.1.3) according to the method of Brooke and Kaiser
(1970). The initial preincubation was made at pH 10.2,
10.4, 10.6, and 10.8 according to the method of Thornell
and Forsgren (1982). The sections were examined under a
Leitz Orthoplan photomicroscope.
Results
Antibody Characterization and Selection
Fifteen stable hybridomas specific for human ventricular myosin were selected by direct RIA from
two separate cell fusions; two hybridomas were
generated from a CB 20 mouse, and the other 13
hybridomas were from a Biozzi HR mouse. Both
immunizations used
SDS-denatured
myosin
(Schwartz et al., 1980) prepared from the hypertrophic left ventricle of a 5 3-year-old female who
had an obstructive cardiomyopathy. The 15 cell lines
were cloned and their supernatants reacted only
with the myosin heavy chains and not with the
myosin light chains or other trace contaminants
present in the whole muscle extract or in the myosin
preparation as determined by the immune replicate
technique (Fig. 2, a,a' and b,b') (Towbin et al.,
1979). The specificity of the antibodies to myosin
heavy chains was further tested by the reactivity of
the two main chymotryptic fragments of the myosin
heavy chains (e.g., rod and subfragment 1) (Fig. 2,
c,c',c") and in some cases to smaller chymotryptic
fragments of purified myosin rod (Fig. 2, d,d',d")
or subfragment 1 (Klotz et al., 1981) to these antibodies. These different blotting experiments enabled
us to verify that the antibodies were directed to the
myosin heavy chain molecule, even though the myosin used as an immunogen contained trace contaminants. Most of the antibodies reacted differently
with four different human myosins from the ventri-
Downloaded from http://circres.ahajournals.org/ by guest on May 12, 2016
Bouvagnet et a/./Isomyosins in Human Atria and Ventricles
MyHC-
797
MyHC - ROD —,
SF1 -«»
«
Ach rv<
Achn- —
VLC1.
VLC2a
a'
b
b
c d
cle, atrium, skeletal muscle, and uterus as determined by competitive radioimmunoassay (data not
shown). These competition assays and the assays of
specificity to different chymotryptic myosin fragments (Fig. 2, d,d',d") suggest that most of the
antibodies were presumably specific for different
epitopes along the myosin heavy chain molecule.
In a preliminary study, the 15 monoclonal antibodies were applied at different dilutions on serial
cryostat sections of human atrial and ventricular
tissue sections, to test their histoimmunological
cross-reactivity with the cardiac fibers detected by
fluorescein-labeled rabbit antimouse IgG. All 15 antibodies had a high response in terms of fluorescence
level, and the response varied with the antibody
dilution. Fiber typing was carried out on more than
c' d'
cd"
FIGURE 2. a, b, c, d: SDS polyacrylamide gel electrophoreses (SDS PAGE) of different preparations of myosin or myosin fragments detected by Coomassie blue,
and 2 a', b', c', c", d', d": corresponding immunoreplicates reacting with different hybridoma supematants and revealed using rabbit antimouse igC coupled
with peroxidase. a: total muscle extract on 11% SDS
PACE: MyHC, VLC1, VLC2 are the myosin heavy and
light chains, respectively; a': immunoreplicate of a
with a monoclonal antibody called 4F4. b: myosin on
11% SDS PACE; b': immunoreplicate of b with the
4F4 monoclonal antibody; c: myosin rod and subfragment 1 obtained after limited chymotryptic treatment
of myofibrils on 14% SDS PAGE; c', c": immunoreplicate of c with 4F4 and another monoclonal antibody
called 2C9, respectively; d: chymotryptic fragments of
purified myosin rods on 14% SDS PAGE; d', d":
immunoreplicates of d with 4F4 and 2C9 monoclonal
antibodies, respectively.
50 tissue blocks coming from 10 different human
atria and ventricles. The controls carried out with
culture medium and labeled with antimouse IgG
displayed a total absence of staining in all atrial and
ventricular fibers (data not shown).
Fiber Typing in Human Atria
To determine whether any of the 15 available
monoclonal antibodies would react differently with
the hypothetically multiple molecular forms of myosin in human heart, tissue samples were initially
excised from the free anterior wall of the right
ventricle and of the anterior wall of the right atrium
of different adult human hearts. A few of these
muscle sections adjacent to those used for im-
A
A
«*
^
»*®«M&
rf
«2>
FIGURE 3. Three serial tissue sections from the anterior trabeculated wall of the right atrial human myocardium (sample 1 in Fig. 1) processed
for indirect immunofluorescence with antiventricular human myosin antibodies (a) and (b) and for the histochemical staining of myosin ATPase
activity after preincubation at pH 10.6 (c). Hybridomas which uniformly stained all atrial fibers (a) constituted the 1A group. The others, which
preferentially stained certain fibers (b), constituted the 1A group. Arrowhead: an atrial fiber highly reactive with the 2A hybridoma group (b) and
exhibiting weak ATPase activity after alkaline preincubation id). Arrow: an atrial fiber weakly reactive with the 24 hybridoma group (b) and
exhibiting high ATPase activity after alkaline preincubation (d). Initial magnification, lOOx. Bar: 20 fim.
Downloaded from http://circres.ahajournals.org/ by guest on May 12, 2016
Circulation Research/Vol. 55, No. 6, December 1984
798
TABLE 1
Correspondence of Fiber Types as Deduced from Indirect
Immunofluorescence and Histochemical ATPase Staining of
Serial Sections of Human Atrial Myocardium
Atrial fiber type
Immunofluorescence
Hybridoma group: 1A
2A
ATPase staining
pH incubation: 10.6
A+
At
Homogeneous
Intense
Weak
Pale
Dark
munofluorescence studies were stained for histochemical determination of myofibrillar ATPase.
Nine of the 15 hybridoma supernatants examined
on atrial tissue sections reacted with the same affinity for all atrial fibers, and the immunofluorescence
level observed varied according to the antibody
dilution. This group of hybridomas, whose supernatants reacted homogeneously with all atrial fibers,
was designated as 1A (Fig. 3a). The supernatants of
the six other hybridomas,-at specific antibody dilutions, displayed a heterogeneous pattern of fluorescence reactivity (Fig. 3b): some fibers designated as
A+ fibers were bright; others fibers designated as
A— fibers were relatively dark, and one group of
fibers showed various intermediate stainings. This
group of hybridomas, whose supernatants reacted
heterogeneously with the atrial fibers, was designated as 2A. By using serial sections, we were able
to verify that all the 2A hybridoma supernatants
consistently reacted with the same atrial fibers (A+
fibers), and conversely, not with the A— fibers.
In order to compare the heterogeneity of human
atrial fibers detected here by antimyosin antibodies
to that previously reported in histochemical deter-
minations of ATPase activities (Thornell and Forsgren, 1982), serial sections of the same atrial samples
were incubated at pH 10.6 and then stained for
ATPase activity. The atrial fibers which were poorly
labeled immunologically with the 2A-hybridoma
group (A— fibers), were always much more heavily
stained for histochemical ATPase activity than the
fibers which were highly labeled with the 2A hybridoma group (A+ fibers) (see arrow in Fig. 3, b
and c, and arrowhead in Fig. 3, b and c, respectively).
Fibers with intermediate stainings both in fluorescence and ATPase activity could be observed. Therefore, there is apparently a close correspondence
between the immunological and histological methods for detecting heterogeneity in human atrium
(see Table 1).
Fiber Typing in Human Ventricles
Nine of the 15 hybridoma supernatants studied
stained all ventricular fibers homogeneously, but at
variable immunofluorescence levels according to the
antibody dilution used (Fig. 4a); this group of cell
lines was designated as IV. The supernatants of the
six other available hybridomas stained certain sparse
fibers intensely (designated as V+ fibers) and reacted weakly and uniformly with the other ventricular fibers (designated as V— fibers) (Fig. 4b); some
very sparse fibers were intermediately stained. This
group of hybridomas was designated as 2V. The
experimental controls in the absence of any hybridoma supernatants displayed a total absence of
staining in all ventricular fibers (data not shown).
Serial cryostat sections of ventricular samples adjacent to those used for immunofluorescence studies
were incubated at different alkaline pH and stained
for ATPase activity (Thornell and Forsgren, 1982).
The few ventricular fibers which reacted intensely
with the supernatants of the 2V-hybridoma group
FIGURE 4. Two serial tissue sections from the free anterior wall of the right ventricle (sample 8 in Fig. 1) processed for indirect immunofluorescence
with antiventricular human myosin antibodies: (a) and (b). Hybridomas which uniformly stained all ventricular fibers (a) constituted the IV
group. The others, which preferentially stained certain sparse fibers (b), constituted the 2V group. Initial magnification, 40x. Bar: 50 pm.
Downloaded from http://circres.ahajournals.org/ by guest on May 12, 2016
Bouvagnet et a/./Isomyosins in Human Atria and Ventricles
799
FIGURE 5. Three serial tissue sections from the anterior papillary muscle of the left ventricle (sample 16 in Fig. 1) processed for indirect
immunofluorescence with a 2V hybridoma (a) and for ATPase activity after preincubation at pH 10.4 (b) and at pH 10.6 (c). Arrow: a group of
three ventricular fibers whose ATPase activity was inhibited at both pH and which corresponds for immunofluorescence to V+ fiber type.
Arrowhead: a fiber which exhibited a slightly higher ATPase activity at pH 10.6 but which reacted like other V- fibers for ATPase activity at
pH 10.4 or for immunofluorescence. Initial magnification, 1OOX.
(V+ fibers) showed no ATPase activity after alkaline
preincubation of the tissue in the pH range between
10.2 and 10.8 (see arrow in Fig. 5, a-c). The other
ventricular fibers which immunologically reacted
weakly with the supernatants of the 2V hybridoma
group (the V— fibers), exhibited homogeneous
ATPase activity after preincubation of the tissue at
pH 10.4 or below (Fig. 5b), but curiously exhibited
heterogeneous ATPase activity after preincubation
of the tissue at higher pH (Fig. 5c). Under the latter
conditions, a few V— fibers were uniformly and
highly stained for ATPase activity (arrowhead in
Fig. 5c), whereas most of the V— fibers exhibited an
intermediate stain for ATPase activity. These data
are summarized in Table 2. Therefore, both the
immunological and the histochemical method detect
heterogeneity in human ventricle, but the close correspondence between the two experimental approaches does not exist as in the studies on human
atrium.
Comparative Fiber Typing in Human Atrium
and Ventricle
We then made a comparative study of the immunological reactivities of the supernatants of the
15 hybridomas on serial sections of ventricular fibers
and atrial fibers, with both muscle sections applied
to the same slide and with the supernatant at the
same dilution. Two of the six hybridoma supernatants of the 2V group, which heavily stained a few
ventricular fibers (V+ fibers in Fig. 4b), displayed a
heterogeneous pattern of reactivity wih atrial fibers
(Fig. 3b); this indicates that the labeled atrial fibers
previously designated as A+ fibers differ not only
from the unlabeled atrial A— fibers, but also from
the unlabeled ventricular V— fibers. This also indi-
cates that unlabeled atrial A - fibers differ from the
labeled ventricular V+ fibers (Table 3). However,
this hybridoma group, 2V(2A), does not support the
evidence for a difference between atrial and ventricular labeled fibers (A+ vs. V+ fibers) and between
atrial and ventricular unlabeled fibers (A— vs. V—
fibers). The four other supernatants of the 2V hybridoma group uniformly and weakly stained all
atrial fibers: this indicates that the atrial fibers previously reactive with the 2V(2A) hybridoma group
(A+ fibers) were different from the more reactive
group of ventricular fibers (V+ fibers). This hybridoma group was called 2V(1A). In contrast, four of
the nine cell line supernatants which uniformly and
intensely stained all ventricular fibers (Fig. 4a, IV
hybridoma group) displayed a heterogeneous pattern of fluorescence reactivity in atrium (Fig. 3b).
This indicates that the unreactive atrial A— fibers
were different from all ventricular fibers which were
reactive with this 1V(2A) hybridoma group (V— or
V+ fibers). The supernatants of the last five hybridomas of the set uniformly and intensely stained
TABLE 2
Correspondence of Fiber Types as Deduced from Indirect
Immunofluorescence and Histochemical ATPase Staining of
Serial Sections of Human Ventricular Myocardium
Ventricular fiber type
Immunofluorescence
Hybridoma group: IV
2V
ATPase staining
pH incubation: 10.6
Ve
V-
Homogeneous
Intense
Weak
Pale
Interm/Dark
Some fibers displayed an intermediate (interm.) reactivity.
Downloaded from http://circres.ahajournals.org/ by guest on May 12, 2016
Circulation Research/Vo/. 55, No. 6, December 1984
800
TABLE 3
Classification of the Set of the Hybridomas Relative to Their
Reactivity for Ventricular and Atrial Fibers
Hybridoma
groups
Immunofluorescence fiber response
ntric
Ventricles
Name
No.
V+
V-
A+
A-
2V (2A)
2V (1A)
IV (2A)
IV (1A)
2
4
4
5
Intense
Intense
Intense
Intense
Weak
Weak
Intense
Intense
Intense
Weak
Intense
Intense
Weak
Weak
Weak
Intense
the atrial and ventricular fibers 1V(1A) and possiblycorresponded to epitopes common to all human
cardiac myoxins. To summarize, human atrium, as
well as human ventricle, is composed of at least two
different antigenic types of muscle fiber, both specific for each cardiac cavity.
Anatomical Distribution of Human Heart Fibers
Immunofluorescence microscopy was further performed on 17 different regions of the hearts of two
16-year-old boys, to determine the fiber distribution
and, consequently, the cellular distribution of the
multiple molecular forms of atrial and ventricular
myosin recognized by the antibodies of the 2A and
2V hybridoma groups, respectively. About 100 serial
7-/im-thick frozen sections per tissue block were
examined.
The most frequent pattern of reactivity observed
in all atrial samples treated with the supematants of
the 2A hybridoma group is shown on Fig. 6a: Many
atrial fibers were either intensely (i.e., the A+ fibers)
or weakly (i.e., the A— fibers) stained, whereas many
others were intermediately stained. All the tissue
sections of the atrial samples which were excised
from one region of postulated conduction pathways
(12) or from common atrial myocardium (4 and 11)
contained only a half-and-half mixture of the A+
and A— fibers. In a few longitudinally oriented
bundles, A+ and A— fibers (reactive or unreactive,
respectively, with antibodies from the 2A hybridoma group) were observed near each other (Fig.
6b): this indicates the apparently discontinuous arrangement of the different types of atrial fibers. The
four other atrial samples examined contained half
FIGURE 6. Tissue sections from four different regions of human atrial myocardium processed for indirect immunofluorescence with the supematants
of the 1A hybridoma group, a: transverse section of the anterior wall of the RA (no. 1 in Fig. 1), b: longitudinal section of the interatrial septum
(no. 4 in Fig. 1), c: transverse section of the subendocardial area of the left auricle (no. 13 in Fig. 1), d: transverse section of the subendocardial
area of the crista terminalis (no. 3 in Fig. 1). Note that the fiber patterns in (a) and (b) are observed in all atrial tissue samples, whereas the fiber
patterns in (c) and (d) are observed in only a few sections of specifically located atrial samples. Initial magnification: a, c, d, 40x. Bar: 50urn;b,
100X. Bar: 20 pm.
Downloaded from http://circres.ahajournals.org/ by guest on May 12, 2016
Bouvagnet et a/./Isomyosins in Human Atria and Ventricles
of the tissue sections with the equilibrated mixture
of heterogeneously reactive fibers described above;
the other half of the tissue sections corresponded to
two opposite situations in terms of the regional
distribution of fiber types. One situation, which
corresponds to the presence of a majority of weakly
reactive A— fibers with a few highly reactive A+
fibers (Fig. 6c), was observed in both auricles (samples 2 and 13). The other situation, which corresponds to the presence of a majority of highly reactive A+ fibers, was observed only in the subendocardial area of the right atrium (samples 1 and 3,
Fig. 6d). However, samples 3 and 12, which were
both excised from regions of a postulated conduction
pathway displayed a different fiber pattern. The
same regional distribution of the different atrial
fibers was observed in both hearts. From this study
of about two thousand sections from seven different
atrial regions in two normal human hearts, it can be
concluded that human atrium contains two different
fiber types which are invariably located and exist in
roughly the same proportion.
The most frequent pattern of reactivity observed
in all ventricular samples treated with the supernatants of the 2V hybridoma group is shown in Figure
7a. Most of the ventricular fibers reacted weakly but
uniformly with the antimyosin antibodies, and only
a few fibers were highly reactive. The most frequent
pattern of ventricular fibers therefore consists of a
majority of V— fibers with a small amount of the
V+ fibers, ranging from approximately 2 to 5% of
all fibers. This regional distribution of fiber types
was observed in all ventricular sections from samples 6, 8, 14, 15, and 16, and on some sections of
samples 5 and 7. The number of rare, highly reactive
V+ fibers increased significantly in some sections
from sample 5, which contained a maximum proportion of about 15% V+ fibers. However, this
801
increase in V+ fibers was not observed in sample 6
which, like sample 5, was excised from a region of
postulated conduction pathways. On the few sections from sample 5, where the V+ were numerous,
they were arranged in bundles (Fig. 7b). Conversely,
these highly reactive V+ fibers were totally absent
in most sections from samples excised from the apex
(10 and 17) and from the conus arteriosus (9). In
terms of the myosin distribution, no differences were
observed between endocardial and epicardial areas.
The same regional variation in the distribution of
the V+ and V— fibers was observed in both hearts.
On the basis of this study of about 3500 sections
from 10 different ventricular regions in two human
hearts, it can be concluded that human ventricle
contains two different fiber types but that the
amount of the V+ type fibers varies roughly between zero to 15% of the total fibers, depending on
the location of the ventricular tissue.
Discussion
The present study shows that human atrium and
human ventricle each contain muscle fibers that
differ in their antigenic reactivity with a set of myosin heavy chain specific antibodies prepared from
an enlarged human ventricle. On the basis of these
observations, we can assume that at least two antigenic types of myosin are present in the ventricle,
as well as in the atrium. Since different hybridoma
supernatants are able to detect these antigenic differences, and since the epitopes of the immunoglobulins used are located at different positions along
the myosin chains, it is likely that the two (or more)
atrial myosins and the two (or more) ventricular
myosins detected each have different primary structures. In addition, since the set of antimyosin hybridomas can be grouped into four subclasses according to the unique or dual specific reactivity of their
FIGURE 7. Tissue sections from two different regions of the human ventricular myocardium processed for indirect immunofluorescence with the
supernatants of the 2V hybridoma group, a: transverse section of the subendocardial region of the mid-portion of the posterior left ventricular
wall (no. 25 in Fig. 1), b: oblique section of the region below the supraventricular crest of the RV (no. 5 in Fig. 1). Note that the fiber pattern in
(a) is observed in almost all tissue samples, whereas the fiber pattern in (b) is observed in only a few sections of specifically located ventricular
samples. Initial magnification, 40x. Bar: 50 pm.
Downloaded from http://circres.ahajournals.org/ by guest on May 12, 2016
Circulation Research/Vo/. 55, No. 6, December 1984
802
FIGURE 8. Tissue sections from the left ventricle of a normal young rat (a and b) and of
a hypothyroid rat (c and d) processed for
indirect immunofluorescence with 2V antimyosin antibodies. Some 2V antibodies reacted only with the V3-type of rat ventricular myosin (a) and (c); other 2V antibodies
reacted in a single way with the VI- and
V3-types of rat ventricular myosin (b).
supernatants with the atrial or ventricular fibers
(Table 3), it can be assumed that the two human
ventricular myosins (corresponding to the V+ and
V— fibers) are molecularly different from the two
human atrial myosins (corresponding to the A+ and
A— fibers) (Table 3). In general, our immunological
results support the view that human atrium and
human ventricle each contain at least two molecular
variants of myosin heavy chains, as do rat, rabbit,
and beef atria and the ventricles of many animals
(Hoh et al., 1978; Flink and Morkin, 1979; Klotz et
al., 1981; Lompre et al. 1981; Chizzonite et al., 1981;
Sartore et al., 1982; Gorza et al., 1982; Litten et al.,
1982; Banerjee, 1983). Heterogeneity of cardiac fibers with respect to myosin composition consequently appears to be a general property of mammalian hearts, regardless of the size of the animals.
The heterogeneous reaction of the fibers to
ATPase after alkaline preincubation within human
atrium was also demonstrated in beef atrium by
Gorza et al. (1982), but, curiously, these results differ
from those of Thornell et al. (1982), who detected a
heterogeneous ATPase reaction within human
atrium only after acid preincubation. Reservations
are therefore required in any interpretation of the
ATPase data. By comparing the histoenzymatic response to the immunological typing of the same
bovine atrial fibers, Gorza et al. (1982) have observed that fibers which were ATPase resistant at
alkaline pH, cross-reacted with anti-beef atrium (or
anti-Vl or anti-myosin heavy chain a) serum, and
that the other minor atrial fibers which lost their
ATPase activity at alkaline pH, cross-reacted with
anti-beef ventricle (or anti-V3 or anti-myosin heavy
chain /3) serum. Because of their similar histochemical behavior, it is tempting to suggest that human
atrial A— fibers with an alkaline stable ATPase
activity are structurally related to a-like myosin
heavy chains and that human A+ fibers with an
alkaline labile ATPase activity are related to /3-like
myosin heavy chain; the set of antibodies used in
the present work would then react either with V3type myosin heavy chains (2A-hybridoma group) or
with VI- and V3-type myosin heavy chains (1Ahybridoma group), and none of the antibodies
would react with VI-like myosin heavy chains
alone. The presence of two different molecular variants of human atrial myosin, each having a structural relationship with one of the two main types of
ventricular myosin, but all of which are different
(see Results and Table 3), is in accordance with the
results of Clark et al. (1982) which showed the
simultaneous presence of VI-type and V3-type epitopes in atrial myosins from chicken, pig, and beef.
The patterns of alkaline pH sensitivity of the
histochemical ATPases of all human ventricular V+
fibers and of most of the V— fibers are rather similar
to those observed within the ventricles of beef and
hypothyroid rats, which lose their ATPase activity
above pH 10.4 and contain mainly the V3-type of
ventricular myosin (Sartore et al., 1981; Weisberg et
al., 1983). The other very rare human V— fibers
which still have an ATPase activity at pH 10.6 (Fig.
5c), behave histochemically like the ventricular fibers of euthyroid rats which contain mainly the VItype of ventricular myosin. The finding that none
of the 15 monoclonal antibodies used here would
react with any human fiber containing only VI-type
myosins, was further tested on ventricular fibers
prepared from a 1-month-old euthyroid rat and
from a hypothyroid rat. We thus observed that some
of our 2V hybridoma stained intensely certain fibers
of the left ventricle of the normal rat (Fig. 8a) and
all fibers of the left ventricle of the hypothyroid rat
Downloaded from http://circres.ahajournals.org/ by guest on May 12, 2016
Bouvagnet et a/./Isomyosins in Human Atria and Ventricles
(Fig. 8c), assessing that these monoclonal antibodies
reacted with epitopes present only in the V3-type of
rat ventricular myosin. We also observed that the
other 2V hybridomas and all IV hybridomas reacted
more or less intensely with the ventricular fibers of
both rats, but in a single way, suggesting that these
monoclonal antibodies reacted with epitopes common to VI- and V3-types of ventricular myosin (Fig.
8, b and d). We also confirmed the histochemical
observations made by Weisberg et al. (1983) and the
relationship between immunological and histochemical fiber typings. Rare Vl-type myosins already have been detected in human ventricles by
Thornell et al. (1982), using histochemistry, and by
Schiaffino et al. (1983) and Mercadier et al. (1983),
using monoclonal and polyclonal antibodies specific
to the rat and beef Vl-type myosins. The probable
absence of anti-Vl monoclonal antibodies in our
hybridoma set could be an unexpected consequence
of the initial choice of an antigen which did not in
fact contain this Vl-type myosin (Mercadier et al.,
1983). On the basis of the histochemical and immunological information currently available, we
tentatively conclude that human ventricle contains
very few Vl-type myosins (presumably, the V—
fibers with an alkaline-stable ATPase activity) and
two molecular variants of V3-type myosins, one
being relatively rare (the V+ fibers with an alkalinelabile ATPase activity), and the other representing
90-95% of the total myosin amount (the V— fibers
with an alkaline-labile ATPase activity). A presence
of two forms of V3-type myosin in rat ventricles has
already been suggested by Watras (1981). More
recently, Mahdavi et al. (1983) have reported a
similar heterogeneity involving two different primary amino acid sequences within the Vl-type myosins in the rat ventricle.
The most interesting result concerning the anatomical distribution of human heart fibers reported
here is the observation of small zones of myosin
variation scattered but not randomly distributed
within a large area of myocardium whose cellular
distribution of myosin is constant. The subendocardial region of the crista terminalis and the anterior
trabeculated wall of the right atrium are rich in
highly immunoreactive myosins, whereas both auricles are rich in nonimmunoreactive myosins. These
regional variations in myosin distribution could perhaps be related to the tension of the different cardiac
walls as the result of a process of adaptation to the
same internal pressure (Sandier and Dogge, 1963).
Our results in human atria were not the same as
those obtained by Gorza et al. (1982), who have
suggested that certain typical patterns of muscle
fiber in beef atrial myocardium may be specialized
for faster conduction. The origin of the discrepancy
is still unknown, but it is probably related to the
postulated existence of specialized conduction pathways in atrial human myocardium (Anderson et al.,
1981). In the ventricle, a distribution gradient of the
small zones of variation in myosin distribution ap-
803
pears to exist within normal myocardium: highly
immunoreactive myosins are more abundant near
the auriculoventricular valves than in the midportion of the ventricular wall or septum, and are nearly
absent in the apex and conus arteriosus. No significant differences in the regional distribution of different myosin types between the left and right ventricle have been detected, with our approach, as
have been found in beef ventricles (Sartore et al.,
1981). However, it should be emphasized that our
anatomical study was limited to whole normal hearts
of two young boys, and that the antibodies used
probably react only with V3-type myosins. The regional distribution of V— fibers with an alkalinestable ATPase activity is thus not described by the
present results. Our anatomical study is currently
being extended to a variety of other whole human
hearts in different physiological or pathophysiological states, in order to detect any possible variation
in their myosin distributions (Schier and Adelstein,
1982; Mercadier et al, 1983; Leger et al., 1983).
We wish to thank Henri Haton for his excellent technical assistance, Dr. G. Biozzi for the kind gift ofH. R. mice, and Dr. ]. Demaille
for valuable discussion. We also thank Drs. L Gorza, S. Sartore, and
S. Schiaffino for their assistance in establishing optimal conditions
for indirect immunofluorescence. We are grateful to Dr. B. Pau for
his assistance in introducing us to hybridoma technology, and for the
support given by the "Service d'lmmunodiagnostic" of the Clin Midy
firm.
This work was supported in part by the "Ministere de I'lndustrie
et de la Recherche" (DGRST81J0260), by TInstitut National de la
Sante et de la Recherche Medicate" (CRL826030), and by la Fondation
de la Recherche Medkale.
Address for reprints: /. /. Leger, INSERM U249, Institut de Biologie,
Boulevard Henri IV, 34000 Montpellier, France.
Received September 9, 1983; accepted for publication August 22,
1984.
References
Anderson RH, Becker AE, Tranum-Jensen J, Janse MJ (1981)
Anatomico-electrophysiological correlations in the conduction
system: A review. Br Heart J 45: 67-82
Bader D, Masaki T, Fischman DA (1982) Immunochemical analysis of myosin heavy chain during avian myogenesis in vivo and
in vitro. J Cell Biol 95: 763-770
Banerjee SK (1983) Comparative studies of atrial and ventricular
myosin from normal, thyrotoxic, and thyroidectomized rabbits.
Circ Res 52: 131-136
Brooke MH, Kaiser KK (1970) Muscle fiber types: how many and
what kind? Arch Neurol 23: 369-379
Chizzonite RA, Everett AW, Clark WA, Jakovcic S, Rabinowitz
M, Zak R (1982) Isolation and characterization of two molecular
variants of myosin heavy chain from rabbit ventricle. J Biol
Chem 257: 2056-2065
Clark WA, Chizzonite RA, Everett AW, Rabinowitz M, Zak R
(1982) Species correlations between cardiac isomyosins. A comparison of electrophoretic and immunological properties. J Biol
Chem 257: 5449-5454
Cummins P (1982) Transitions in human atrial and ventricular
myosin light chain isoenzymes in response to cardiac pressure
overload induced hypertrophy. Biochem J 205: 195-204
Di Pauli R, Raschke WC (1978) A hybridoma secreting IgM anti
Iglb allotype. In Current Topics in Microbiology and Immunology 81, edited by F Melchers, M Potter, NL Warner. Berlin,
Springer-Verlag, pp 37-39
Downloaded from http://circres.ahajournals.org/ by guest on May 12, 2016
804
Circulation Research/Vol. 55, No. 6, December 1984
Flink IL, Morkin E (1979) Thyroid hormone stimulates synthesis
of a cardiac myosin isozyme. J Biol Chem 254: 3105-3110
Gorza L, Sartore S, Schiaffino S (1982) Myosin types and fiber
types in cardiac muscle. II: Atrial myocardium. J Cell Biol 95:
838-845
Hoh JFY, McGrath PA, Hale PT (1978) Electrophoretic analysis
of multiple forms of rat cardiac myosin: effects of hypophysectomy and thyroxine replacement. J Mol Cell Cardiol 10: 10531076
James TN, Sherf L (1971) Specialized tissues and preferential
conduction in the atria of the heart. Am J Cardiol 28: 414-426
Klotz C Swynghedauw B, Mendes H, Marotte F, Leger JJ (1981)
Evidence for new forms of cardiac myosin heavy chains in
mechanical heart overloading and in ageing. Eur J Biochem
115: 415-421
Klotz C, Leger JJ, Elzinga M (1982) Comparative sequence of
myosin light chains from normal and hypertrophied human
hearts. Circ Res 50: 201-209
Klotz C, Dechesne C, Cardinaud P, Leger JJ (1983) Structural
differences between atrial and ventricular myosins from normal
human hearts. Biochimie 65: 569-578
Laemmli UK (1970) Cleavage of structural proteins during the
assembly of the head of bacteriophage T4. Nature (Lond) 227:
680-685
Leger Joe, Klotz C, Leger JJ (1983) Cardiac myosin heavy chains
and rropomyosin in mechanical heart overloading and ageing.
In Perspectives in Cardiovascular Research: Myocardial Hypertrophy and Failure, vol 7, edited by NR Alpert. New York,
Raven Press, pp 385-392
Litten RZ III, Martin BJ, Low RB, Alpert NR (1982) Altered myosin
isozyme patterns from pressure-overloaded and thyrotoxic hypertrophied rabbits hearts. Circ Res 50: 856-864
Lompre AM, Mercadier JJ, Wisnewski C, Bouveret P, Pantaloni
C, D'Albis A, Schwartz K (1981) Species- and age-dependent
changes in the relative amounts of cardiac myosin isoenzymes
in mammals. Dev Biol 84: 286-290
Mahdavi V, Periasamy M, Nadal-Ginard B (1982) Molecular
characterization of two myosin heavy chain genes expressed in
the adult heart. Nature 297: 659-665
Matus A, Pehling G, Ackermann M, Maeder J (1980) Brain postsynaptic densities: Their relationship to glial and neuronal
filaments. J Cell Biol 87: 346-351
Mercadier JJ, Bouveret P, Gorza L, Schiaffino S, Clark WA, Zak
R, Swynghedauw B, Schwartz K (1983) Myosin isozymes in
normal and hypertrophied human ventricular myocardium.
Circ Res 53: 52-62
Miller DM, Ortiz I, Berliner GC, Epstein HF (1983) Differential
localization of two myosins within nematode thick filaments.
Cell 34: 477-490
Netter FH (1974) A compilation of paintings on the normal and
pathologic anatomy and physiology, embryology, and diseases
of the heart. In The Ciba Collection of Medical Illustrations,
vol 5, edited by FF Yonkman. New Jersey, CIBA, pp 9-13
Rupp H (1982) Polymorphic myosin as the common determinant
of myofibrillar ATPase in different haemodynamic and thyroid
states. Basic Res Cardiol 77: 34-46
Sandier H, Dogge HT (1963) Left ventricular tension and stress
in man. Circ Res 13: 91-96
Sartore S, Gorza L, Pierobon Bormioli S, Dalla Libera L, Schiaffino
S (1981) Myosin types and fiber types in cardiac muscle. I:
Ventricular myocardium. J Cell Biol 88: 226-233
Schiaffino S, Gorza L, Sartore S (1983) Distribution of myosin
types in normal and hypertrophic hearts: an immunocytochemical approach. In Perspectives in Cardiovascular Research:
Myocardial Hypertrophy and Failure, vol 7, edited by NR
Alpert. New York, Raven Press, pp 149-166
Schier JJ, Adelstein RS (1982) Structural and enzymatic comparison of human cardiac muscle myosins isolated from infants,
adults, and patients with hypertrophic cardiomyopathy. J Clin
Invest 69: 816-825
Schwartz K, Lompre AM, Bouveret P, Wisnewsky C, Swynghedauw B (1980) Use of antibodies against dodecyl sulfate denatured heavy meromyosins to probe structural differences
between muscular myosin isoenzymes. Eur J Biochem 104:341 346
Schwartz K, Lecarpentier Y, Martin JL, Lompre AM, Mercadier
JJ, Swynghedauw B (1981) Myosin isoenzymic distribution
correlates with speed of myocardial contraction. J Mol Cell
Cardiol 13: 1071-1075
Swynghedauw B, Delcayre C (1982) Biology of cardiac overload.
Pathobiol Annual 12: 137-183
Thornell LE, Forsgren S (19820 Myocardial cell heterogeneity in
the human heart with respect to myosin ATPase activity.
Histochem J 14: 479-490
Towbin H, Staehelin T, Gordon J (1979) Electrophoretic transfer
of proteins from polyacrylamide gels to nitrocellulose sheets:
Procedure and some applications. Proc Natl Acad Sci USA 76:
4350-4354
Tuchschmid CR, Srihari T, Hirzel HO, Schaub MC (1983) Structural variants of heavy and light chains of atrial and ventricular
myosins in hypertrophied human hearts. In Cardiac Adaptation
to Hemodynamic Overload, Training and Stress. Erwin Riesch
Symposium, Tubingen, Darmstadt, edited by D Steinkopff
Verlag, pp 123-128
Wachsberger P, Lampson L, Pepe FA (1983) Non-uniform staining of myofibril A bands by a monoclonal antibody to skeletal
muscle myosin SI heavy chain. Tissue Cell 15: 341-349
Watras J (1981) Changes in rat cardiac myosin during development and in culture. J Mol Cell Cardiol 13: 1011-1021
Weisberg A, Winegrad S, Tucker M, McClellan G (1982) Histochemical detection of specific isozymes of myosin in rat ventricular cells. Circ Res 51: 802-809
Winckelmann DA, Lowey S, Press JL (1983) Monoclonal antibodies localize changes on myosin heavy chain isozymes during
avian myogenesis. Cell 34: 295-306
INDEX TERMS: Anatomy • Human • Hybridoma • Immunofluorescence • Myosin • Myocardium
Downloaded from http://circres.ahajournals.org/ by guest on May 12, 2016
Fiber types and myosin types in human atrial and ventricular myocardium. An anatomical
description.
P Bouvagnet, J Leger, F Pons, C Dechesne and J J Leger
Circ Res. 1984;55:794-804
doi: 10.1161/01.RES.55.6.794
Circulation Research is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1984 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7330. Online ISSN: 1524-4571
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circres.ahajournals.org/content/55/6/794
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in
Circulation Research can be obtained via RightsLink, a service of the Copyright Clearance Center, not the
Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further information
about this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation Research is online at:
http://circres.ahajournals.org//subscriptions/
Downloaded from http://circres.ahajournals.org/ by guest on May 12, 2016