Download Developmental anatomy of the heart: a tale of

Physiol Genomics 15: 165–176, 2003;
10.1152/physiolgenomics.00033.2003.
review
Developmental anatomy of the heart: a tale of mice and man1
Andy Wessels and David Sedmera
Department of Cell Biology and Anatomy, Cardiovascular Developmental Biology Center,
Medical University of South Carolina, Charleston, South Carolina 29425
Submitted 12 March 2003; accepted in final form 26 August 2003
Wessels, Andy, and David Sedmera. Developmental anatomy of
the heart: a tale of mice and man. Physiol Genomics 15: 165–176, 2003;
10.1152/physiolgenomics.00033.2003.—Because of the increasing availability of tools for genetic manipulation, the mouse has become the most
popular animal model for studying normal and abnormal cardiac development. However, despite the enormous advances in mouse genetics,
which have led to the production of numerous mutants with cardiac
abnormalities resembling those seen in human congenital heart disease,
relatively little comparative work has been published to demonstrate the
similarities and differences in the developmental cardiac anatomy in
both species. In this review we discuss some aspects of the comparative
anatomy, with emphasis on the atrial anatomy, the valvuloseptal complex, and ventricular myocardial development. From the data presented
it can be concluded that, apart from the obvious differences in size, the
mouse and human heart are anatomically remarkably similar throughout development. The partitioning of the cardiac chambers (septation)
follows the same sequence of events, while also the maturation of the
cardiac valves and myocardium is quite similar in both species. The
major anatomical differences are seen in the venous pole of the heart. We
conclude that, taking note of the few anatomical “variations,” the use of
the mouse as a model system for the human heart is warranted. Thus the
analysis of mouse mutants with impaired septation will provide valuable
information on cellular mechanisms involved in valvuloseptal morphogenesis (a process often disrupted in congenital heart disease), while the
study of embryonic lethal mouse mutants that present with lack of
compaction of ventricular trabeculae will ultimately provide clues on the
etiology of this abnormality in humans.
mouse; human; cardiac development; embryo
A FAST-INCREASING NUMBER
of genetically modified mouse
models with structural and functional abnormalities in
the cardiovascular system undoubtedly will contribute
to an improved understanding of molecular and morphological mechanisms that regulate human heart development in health and disease. It becomes more and
more obvious, however, that for proper extrapolation of
findings in the murine model systems it is crucial to
have detailed knowledge of the specific anatomical/
morphological details of developing and adult heart in
mice and humans (2, 33, 35).
Before discussing some of the specific anatomical
features in both species, it is important to consider the
Article published online before print. See web site for date of
publication (http://physiolgenomics.physiology.org).
Address for reprint requests and other correspondence: A. Wessels, Dept. of Cell Biology and Anatomy, Medical Univ. of South
Carolina, 173 Ashley Ave., Basic Science Bldg., Rm. 648, PO Box
250508, Charleston, SC 29425 (E-mail: [email protected]).
1
This review article was based on work originally presented at the
“NHLBI Symposium on Phenotyping: Mouse Cardiovascular Function and Development” held at the Natcher Conference Center, NIH,
Bethesda, MD, on October 10–11, 2002.
dimensions and properties of the full-grown hearts.
The human heart weighs about 250–300 g and beats,
on average, 60–70 times per minute. In contrast, the
mouse heart weighs only about 0.2 g, and has a heartbeat of around 500–600 times per minute. Apart from
their size, the most pronounced difference in external
features between the adult mouse and human heart is
found in their overall shape. The most important factor
determining this shape is the anatomical context of the
adult heart within in thoracic cavity. In human, the
heart “rests” on the diaphragm. This is reflected in its
more pyramidal shape and a flat dorsal (or inferior)
surface. In comparison, the mouse heart, which in the
four-legged mouse does not rest on the diaphragm and
has more room to move freely within the pericardial
cavity, has a more ellipsoidal (“rugby ball” shape).
Another difference in external features is that in the
human heart the atria are very prominent structures,
whereas in the mouse heart the atrial chambers and
their appendages are relatively small.
Developmentally, it is interesting to note that the
gestational window during which the heart develops is
quite different in the mouse and human. In the human
1094-8341/03 $5.00 Copyright © 2003 the American Physiological Society
165
166
REVIEW: ANATOMY OF MOUSE AND HUMAN HEART
it takes about 2 mo (from conception) for the heart to
complete septation, followed by another 7 mo to further
mature until the baby is born and the pulmonary
circulation kicks in. In the mouse, however, it takes
only 2 wk from the time of conception for cardiac
septation to complete. After that, the mouse fetus has
less than 1 wk of prenatal life before birth. Without
going into any detail, it suffices to say that some of the
developmental events that in the human are more or
less completed at birth are still in progress in the
neonatal mouse. There are several publications in literature in which the comparative developmental
stages in mouse and human are listed (28, 39). These
tables are helpful when studying developmental
events. It is important to realize, however, that the
staging protocol used to determine the developmental
stage of mouse embryos is not always the same. For
instance, when using the “vaginal plug” method, most
investigators will mark the morning on which a sperm
plug is observed as embryonic day 0.5 (ED0.5 ; Ref. 34).
Others, however, will consider this ED1.0. This obviously can lead to confusion, and more importantly,
misinterpretation of data. Thus, when describing and
comparing developmental events, it is very important
to be very explicit about the staging process when
presenting developmental data. Moreover, when working with tissues generated by others (previously collected materials, existing collections, collaborations)
where the exact staging protocol that has been followed
is not known, it is equally important to point that out
when reporting on data based on these tissues. As more
and more clinical imaging techniques are now being
adapted for use in studies on mouse models [e.g., MRI,
ultrasound (10)], it is imperative to keep the issues
mentioned above in mind. Thus, although the murine
and human heart are anatomically very similar, it is
crucial to be very careful when extrapolating information obtained by studying the mouse into the human
context.
In this review we will first discuss some of the anatomical characteristics of the human and murine cardiovascular system. It will be demonstrated that these
characteristics do not differ significantly. In the remainder of the review we will, therefore, focus on the
general development of the cardiovascular system in
both species by providing some insight in the mechanisms that lead to the formation of the functional, fully
septated, four-chambered heart. As cardiac development is a very complex process, and it is beyond the
scope of this review to discuss all the aspects that
relate to this event, we will limit ourselves to the
discussion of two topics that relate to congenital malformations frequently seen in humans. First, as many
of the congenital heart defects found in the newborn
human heart include abnormalities of the valvuloseptal complex, we will discuss some of the developmental
events that lead to proper valvuloseptal morphogenesis. We will also, albeit less extensively, describe some
of the events that are involved in myocardial differentiation and maturation.
Physiol Genomics • VOL
15 •
MATERIAL AND METHODS
Most of the images of human and murine specimens used
in the illustrations of this review were generated in the
context of studies published previously (26, 27, 31, 37, 38,
40–43). Some of the scanning electron microscope (SEM)
images are from the collection of late Tomas Pexieder. Details regarding tissue collection, preparation, histology, immunohistochemical staining procedures, characterization
and properties of antibodies, and the preparation of SEM
samples can be found in papers mentioned above and in the
remainder of the text. As most of the photographs in the
respective histological and SEM collections did not contain
scale bars, we could unfortunately not provide scale bars for
the figures in this paper.
THE ANATOMY OF THE POSTNATAL HEART IN MOUSE
AND HUMAN
The basic anatomical features of the postnatal heart
in the human and mouse are very similar (Fig. 1). Thus
in both species the heart has four chambers; two atria,
separated by an interatrial septum (IAS), and two
ventricles, separated by an interventricular septum
(IVS). In addition, located between the IAS and IVS
there is a small “septal segment,” which, as a result of
the offset of the atrioventricular (AV) valves, is known
as the atrioventricular septum (AVS), as it is basically
situated in between the subaortic outlet segment of the
left ventricle from the right atrium. In the human, this
AVS is a thin fibrous structure, known as the membranous septum. In the mouse this structure is relatively
thick and mostly muscular, partly as a result of delayed delamination of the septal leaflet of the tricuspid
valve (35), and partly as a result of myocardialization
of the mesenchymal tissues (13). In the junction situated between the atria and ventricles, i.e., the AV
junction (AVJ), we find two AV valves. The arrangement of the AV valves in mouse and human is comparable (35). In the left AVJ we find a mitral valve, which
has two distinct leaflets (a bicuspid valve), whereas in
the right AVJ a tricuspid valve is located, which has
three distinct leaflets. In the human heart the valves
are at their “tips” in continuity with very pronounced
papillary muscles via relatively long and numerous
tendinous chords (chordae tendineae; Fig. 1A). In the
mouse these chordae are far less prominent (Fig. 2, A
and C). The inner lining of the ventricles is characterized by the presence of numerous myocardial protrusions better known as trabeculae (trabeculae carneae).
In the human, but not so in the mouse, there is a
pronounced difference in the morphology of the trabeculae in the right vs. the left ventricle (35). In the left
ventricle of the human heart, the trabeculae are relatively thin, whereas the trabeculae in the right ventricle have been described as being coarse. In addition to
the trabeculation and the chordae tendineae, we can,
within the apical cavity of the ventricles in both species, also discriminate thin, cordlike, structures that
resemble the tendinous chords attached to the papillary muscle. Because of their tendonlike appearance, in
older anatomy textbooks (e.g., Told-Hochstetter,
Anatomischer Atlas, 1948) these structures are often
www.physiolgenomics.org
REVIEW: ANATOMY OF MOUSE AND HUMAN HEART
167
Fig. 1. Anatomy of the postnatal human
heart. A: a postnatal human heart in 4-chamber view (from Ref. 1, with permission). B:
histological section of a 5-wk old neonatal
heart, sectioned in the same orientation, is
shown. The section was immunohistochemically stained for the presence of atrial myosin
heavy chain (aMHC) as described in Ref. 40.
In addition to the expression of aMHC in the
atria, expression of this MHC isoform is also
observed in the left and right bundle
branches of the atrioventricular conduction
system (arrows). Note the orientation and
thickness of the interventricular septum
(IVS) in relation to the left and right ventricular free wall. Ao, aorta; AVS, atrioventricular septum; IAS, interatrial septum; LA, left
atrium; LBB, left bundle branch; LV, left
ventricle; RA, right atrium; RBB, right bundle branch; RV, right ventricle.
described as trabeculae tendineae. However, despite of
this resemblance, these thin structures are not really
tendons, but are actually extensions of the subendocardial ventricular network of the cardiac conduction system, and are therefore nowadays generally referred to
as “false tendons” (Fig. 2D). As they consist almost
exclusively of Purkinje cells (i.e., specialized myocytes
of the conduction system) within a fibrous sheet, they
can be used to obtain enriched preparations of conduction system related genes (20) or to perform physiological experiments (9). A slight morphological difference
in the overall ventricular anatomy between mouse and
human is found in the relative size and shape of the
muscular ventricular septum and the position of the
aortic outlet in relation to the IVS (cf. Figs. 1 and 2). In
the human heart the muscular IVS is a massive structure, its thickness approaching or exceeding that of the
left ventricular free wall (Fig. 1, A and B). The human
muscular IVS has a very broad base just below the AV
valve attachments. In the mouse, the IVS is not quite
as massive and compact (Fig. 2, A and B), and at the
base it gradually tapers toward the AV septum. The
difference in the angle of the aortic outlet relative to
the axis of the IVS is another remarkable feature (cf.
Fig. 1B and Fig. 2A). The most prominent anatomical
variations between the cardiovascular systems in
mouse and human are probably those found in the
venous components of the atria (Fig. 3). All of these
variations reflect the species-specific differences in development of the venous tributaries connecting to the
inferior atrial wall. Thus, whereas in the human heart
the left atrium receives four pulmonary veins (37), in
the mouse heart, the pulmonary veins join in a pulmonary confluence behind the left atrium (Fig. 3, G and
H), which in turn empties via a single foramen into the
dorsal wall of the left atrium (13, 36). Another anatomical difference in the atrial anatomy relates to the
venous drainage into the right atrium. During the
early stages of cardiac morphogenesis, in both the
Physiol Genomics • VOL
15 •
human and murine embryo, the left and right superior
caval veins (LSCV and RSCV) initially drain into the
sinus venosus. The sinus venosus itself opens via the
sinoatrial foramen into the right atrium (36, 37). As
development progresses, in the human heart the left
caval vein regresses and the remaining proximal portion (with part of the left sinus horn) becomes the
coronary sinus, the remnant of the LSCV becoming the
so-called ligament of Marshall and oblique vein (Fig. 3,
A–C). In the mouse, however, the LSCV does not regress and persists into postnatal life (Fig. 3H). In the
human, persistent LSCV is considered a congenital
malformation (incidence ⬍1:100) frequently associated
with other congenital malformations. It is likely that
these differences in development of the venous tributaries into the atrial chambers are the underlying
reasons for the relatively small size of the atrial components in the murine heart.
Valvuloseptal Development
The first step in cardiac development in the vertebrate heart involves the formation of two bilateral
heart fields of precardiac mesoderm. These heart fields
are situated on opposite sides of the embryonic midline
(6, 24) and contain endocardial as well as myocardial
precursor cells. As the embryo develops, the heart
fields fuse, resulting in the formation of a primary
heart tube. This tubular heart consists of a myocardial
outer mantle with an endocardial inner lining. Between these two concentric epithelial cell layers, an
acellular matrix is found which is generally referred to
as the cardiac jelly (Fig. 4A). During cardiac looping,
the cardiac jelly basically disappears from the future
major chambers of the heart (i.e., atria and ventricles),
whereas in the junction between the atria and ventricles, the atrioventricular junction (AVJ), as well as in
the developing outflow tract (OFT), the cardiac jelly
starts to accumulate. This results in the formation of
www.physiolgenomics.org
168
REVIEW: ANATOMY OF MOUSE AND HUMAN HEART
Fig. 2. Scanning electron micrographs
(SEM) of the postnatal mouse heart. A
and B: SEM images of the posterior (A)
and anterior (B) half of an adult mouse
heart. C: an enlargement of the boxed
area in A, showing the relatively small
leaflets of the murine mitral valve (cf.
Fig. 1A). D: an enlargement of the
boxed area in B and demonstrates
some of the false tendons in the apical
portion of the right ventricle. See legend to Fig. 1 for definitions of abbreviations.
Fig. 3. Development of the venous pole of the heart. A–C: a schematic representation of the development of the
venous pole in human heart (modified from Fig. 7-10 in Human Embryology, William J. Larsen, Churchill
Livingstone, 1993). D–F: immunohistochemically stained section of human embryos stained for ␣-MHC at,
respectively, 5.5 (D), 7 (E), and 8 wk (F) of development. G: a histological section of a nonperfused neonatal mouse
heart. H: is a schematic representation of G (color coding according to A–C). A: in early development (⬃4–5 wk)
the left and right horns of the sinus venosus communicate with the developing (unseptated) atrium. The left sinus
horn (LSH) receives blood from several vessels including the left superior caval vein (LSCV), and the right sinus
horn (RSH) is connected to the right superior caval vein (RSCV). As development progresses, the LSCV regresses
(see B), resulting in the formation of the oblique vein (see C), with the remnant of the LSH becoming the coronary
sinus (CS) emptying into the right atrium (RA; see C). The cartoon also illustrates how during human cardiac
development the confluence of the pulmonary veins (PuV; see A) becomes incorporated into the left atrium (see B
and C). D: at 5.5 wk of development the pulmonary veins connect to the LA through a single communication (note
that the walls of the pulmonary veins are not myocardial at this point). At 7 wk the walls of the confluence of the
pulmonary veins are becoming myocardial (E). The process of incorporation is nearly completed at 8 wk; the walls
of the pulmonary veins are now myocardialized up to the pericardial reflection (arrows). The images in G and H
demonstrate that in the mouse the pulmonary confluence does not become incorporated into the LA. Instead, the
pulmonary veins drain into the LA via a single channel (arrow). G and H also illustrate that, unlike in the human,
persisting LSCV is a normal anatomical situation in the postnatal mouse. ICV, inferior caval vein; PA, pulmonary
artery; SCV, superior caval vein. See legends of previous figures for other abbreviations.
Physiol Genomics • VOL
15 •
www.physiolgenomics.org
REVIEW: ANATOMY OF MOUSE AND HUMAN HEART
the endocardial cushion tissues in the AVJ (see Figs. 4,
E–H, for examples in the mouse and Figs. 6, E and F
for an example in the human) and OFT. An important
developmental event in the subsequent maturation of
these cushions is the endocardial epithelial-to-mesenchymal transformation (EMT; see Ref. 19). During this
EMT, a subset of the endocardial cells delaminate from
Physiol Genomics • VOL
15 •
169
their epithelial context, transform, thereby assuming a
mesenchymal phenotype, and start to migrate into the
extracellular matrix of the cushions (Fig. 4B). This
process, which is, at least partly, regulated by growth
factors that are produced in the underlying myocardium (for a review, see Ref. 7), results in the mesenchymalization of the cushions. The endocardial cushion
www.physiolgenomics.org
170
REVIEW: ANATOMY OF MOUSE AND HUMAN HEART
tissues are extremely important for cardiac morphogenesis. They are the major “building blocks” of, and
provide the “glue” for, virtually all of the septal structures in the heart (38). Dysmorphogenesis of the endocardial cushions and/or failure of proper fusion are
generally thought to play a major role in the etiology of
congenital heart disease in humans, for instance, in
the setting of common AV defects. Gross abnormalities
in formation of endocardial cushion tissue is observed
in a variety of genetically modified mouse models such
as the neurofibromin-1 (16), hyaluronan synthase-2
(Has2) (4), and RXR-␣ knockout mice (8) and is usually
associated with fetal lethality. The spectrum of cushion
abnormalities is very broad. In some genetically altered mice, hardly any cushion tissue is formed,
whereas others develop hypercellularized and/or enlarged cushions. In many cases this appears to be a
result of perturbed endocardial-to-mesenchymal transformation leading to either too many (16) or too few
(14) endocardially derived cells in the cushions tissues.
Although, in general, mice with cushion abnormalities
manifest defects in both the AV as well as the OFT
region, sometimes this is not the case and only the OFT
cushions are affected (NFAT-C and SOX-4; Ref. 44). As
indicated above, endocardial cushions develop in the
AVJs as well as in the OFT. There are many similarities in the fate of endocardial cushions in AVJ and
OFT. The cushion tissues in the AVJ contribute to the
formation of AV septal structures and AV valves, and
the cushion tissues in the OFT take part in septation of
the OFT and in the formation of the semilunar valves
of aorta and pulmonary artery (see, e.g., Ref. 23). In the
next section we will discuss some of the aspects of AV
development in a little more detail.
The AV junction. As briefly outlined above, the endocardial jelly in the AVJ forms the base material for
the AV cushions (21, 38). Initially only two endocardial
cushions develop which face each other on opposing
sides of the common AV canal. These cushions, which
we will refer to as the “major” AV cushions, are known
as the inferior (also known as posteroinferior or dorsal)
AV cushion and the superior (also known as ventral or
anterosuperior) AV cushion (Fig. 4, B and E). The
major AV cushions are the most prominent endocardial
cushion tissues in the heart, and, as development proceeds, the leading edges of the cushions fuse (Fig. 4, C,
F, and G), thereby separating the common AV canal
into a left and right AV orifice (Fig. 4, C, D, and F). At
this point, another set of, relatively small, endocardial
cushions starts to develop in the lateral aspects of the
AVJ. These cushions are generally referred to as the
lateral cushions (Fig. 4, C and F). Interestingly, although the lateral cushions are very important for
valvuloseptal morphogenesis, as they contribute to the
formation of anterosuperior leaflet of the tricuspid
valve (17) and the mural leaflet of the mitral valve, the
mechanisms that are responsible for inducing their
development and differentiation have thus far been
very poorly studied.
Fate of the AV cushions. After fusion, the (major) AV
cushion-derived tissue basically forms a large mesenchymal “bridge” spanning from the posterior wall of the
AV canal to the anterior wall of the heart (37). Posteroinferiorly, this mesenchymal component is in continuity with the dorsal mesenchymal protrusion/dorsal mesocardium and the mesenchymal cap (also endocardial
cushion material) that covers the leading edge of the
forming primary interatrial septum (pIAS) septum (described in detail in Ref. 37). This mesenchymal cap is
anterosuperiorly also in continuity with the remnant of
the superior AVC cushion. The pIAS at this point has
not closed the communication between left and right
atrium yet, and as a result the mesenchymal bridge
spans the primary atrial foramen (interatrial communication) between the dorsal and ventral aspects of the
fused cushions. As development progresses, the mesenchymal cap on the primary septum and the fused
cushions merge completely, thereby closing this primary foramen. The remodeling of this mesenchymal
“crux of the heart,” which at this point basically consists of the fused major AV cushions and the mesenchymal cap of the pIAS and which is situated between
the muscular IVS and the IAS (Fig. 5A), eventually
leads to the formation of the membranous AV septum,
the septal leaflet of the tricuspid valve (17), and the
aortic leaflet of the mitral valve (Fig. 5B). Although the
endocardial cushion tissues are important in the formation of the valves, AV valve morphogenesis also
involves a number of myocardial remodeling steps (Fig.
6). Below, we have summarized (and simplified) the
steps that are involved in the formation of the valves. It
is important to realize that not all the leaflets of the AV
valves develop exactly the same (see Refs. 17 and 21).
However, in the context of this paper it suffices to
describe the general principle, focusing on the fate of
the lateral cushions. First, it is important to determine
the relationship between the cushions and the adjacent
tissues. The cushions have an upper (atrial) boundary
Fig. 4. Development of the atrioventricular cushions in the mouse. A–D: schematic of the development of the AV
cushions in the mouse/human heart. A: the tubular heart stage. The outer myocardial layer and the inner
endocardial cell layer sandwich the cardiac jelly (CJ). B: the stage in which the cardiac jelly has given rise to the
formation of the major AV cushions. The cells within the subendocardial space indicate that endocardial-tomesenchymal transformation has been initiated. C: the mesenchymalized major cushions are fusing, while at the
lateral aspects of the AV canal the lateral AV cushions have formed. In D, the fusion has been completed, and a
wedge of mesenchyme separates the left from the right AV orifice. The specimen shown in E is representative for
the cartoon in B, F shows a specimen that correlates to C, while H demonstrates the fused mesenchymal tissues
as depicted in D. G: a higher magnification of the boxed area in F, showing the fusion of the endocardial layers of
the two major AV cushions. “cap,” mesenchymal cap on leading edge of IAS; endo, endocardium; epi, epicardium;
iAVC, inferior AV cushion; llAVC, left lateral AV cushion; OFT, outflow tract; sAVC, superior AV cushion; myo,
myocardium; rlAVC, right lateral AV cushion. See legends of previous figures for other abbreviations.
Physiol Genomics • VOL
15 •
www.physiolgenomics.org
REVIEW: ANATOMY OF MOUSE AND HUMAN HEART
Physiol Genomics • VOL
15 •
www.physiolgenomics.org
171
172
REVIEW: ANATOMY OF MOUSE AND HUMAN HEART
Fig. 5. Contributions of the major atrioventricular cushions to valvuloseptal morphogenesis in the human heart.
These panels show two serial sections of a human heart at
⬃6 wk of development stained for the presence of atrial
MHC (A) and ventricular MHC (B) as described before
(42). A: this is how the mesenchymal tissues of the two
major cushions and the mesenchymal cap on the IAS have
fused to form the mesenchymal “crux” of the heart. B: this
indicates how the structures indicated in A contribute to
valvuloseptal morphogenesis in the AV junction (AVJ).
epi, epicardially derived mesenchyme. See legends of previous figures for other abbreviations.
and a lower (ventricular) boundary. The bulk of the
cushion tissue is plastered against the AV junctional
and ventricular myocardium (Fig. 6, A, E, and F). The
first myocardial remodeling event leads to the interruption of the continuity between atrial and myocardial ventricular myocardium (Fig. 6, B and C) and is a
result of the fusion between cushion-derived tissues
and epicardially derived mesenchyme (38). Specifically, this process takes place at the lower end of the
AV canal myocardium resulting in the incorporation of
most of the AV canal myocardium into the lower rim of
the atrial chambers (Fig. 6C). This process establishes
the functional separation, and electrical insulation,
between atrial and ventricular working myocardium
(for a detailed description of this process see Refs. 18,
38, and 41). There is only one place where this separation does not take place, and that is where the specialized myocardium of the AV conduction axis (AV node
and proximal His bundle) develops (38). These components of the AV conduction system (AVCS) relay the
cardiac impulse from atria to ventricles. The second
myocardial remodeling event that plays an important
role in the formation of the freely moveable leaflets of
the AV valves is myocardial delamination (17, 21).
During this process the lower part of the AV cushions,
with associated myocardial layer (Fig. 6, B, C, G, and
H), separates from adjacent ventricular myocardium
by a hitherto undetermined process. This results in the
Fig. 6. Atrioventricular valve development in the embryonic human heart. A–D: schematic representations of AV
valves in the developing human heart. E–N: a series of immunohistochemically stained sections of representative
stages in this process. Sections shown in E, J, K, L, and N were stained with the “cushion-tissue antigen” antibody
249-9G9 (anti-CTA; Ref. 38), the sections in F, G, and I were stained with anti-ventricular MHC (43), the section
in H (sister section to G) with anti-mesenchymal cell antigen (anti-MCA; Ref. 38), and the section in M (sister
section to N) was stained with an antibody that recognizes the atrial as well as the ventricular isoform of MHC (38).
The sections shown in E and F are from a heart at ⬃5 wk of development (38). G and H: sister sections from a
9-wk-old heart. I–K: sister sections from a heart at 12–13 wk of development. L–N: sister sections from a heart at
17 wk of development. The panels demonstrate (represented schematically in A–D) how during development the
leaflets of the valves delaminate and detach from the myocardial wall. Initially, the leaflets contain a considerable
amount of myocardium at their ventricular aspect (best seen in G), but as development progresses this myocardium
gradually disappears (cf. G, I, and M). AV myo, atrioventricular myocardium. See legends of previous figures for
other abbreviations.
Physiol Genomics • VOL
15 •
www.physiolgenomics.org
REVIEW: ANATOMY OF MOUSE AND HUMAN HEART
Physiol Genomics • VOL
15 •
www.physiolgenomics.org
173
174
REVIEW: ANATOMY OF MOUSE AND HUMAN HEART
Fig. 7. Development of the compact
ventricular myocardium in mouse development in health and disease. Compaction of the ventricular trabeculae
adds to proportion and thickness of the
compact myocardium. In the human,
this takes place between 12 wk (A) and
18 wk (B). Sagittally dissected parietal
halves of left ventricles are shown.
Scale bars 100 ␮m. (Images are from
the Pexieder collection: B first published in Ref. 26, and A appeared in
Ref. 31). Compaction occurs in the
mouse between ED13 and 14, resulting
in well-formed multilayered compact
layer by ED14.5 in the wild-type animal (black arrows in C and white arrow in E). This process is perturbed in
several knockouts, such RXR-␣ (black
arrows in D and white arrow in F),
resulting in “thin compact myocardium
syndrome” and embryonic lethality.
formation of “immature” or “prevalvar” bilayered leaflets, with the inferior (or “ventricular”) layer consisting
of myocardium and the superior (“atrial”) layer being
cushion derived (Fig. 6, C and G). Toward the ventricular apex, these leaflets are attached to the ventricular
walls and/or IVS through the developing papillary
muscles (Fig. 6, G and H). The myocardium gradually
disappears from the developing leaflets between 10
and 17 wk of development (Fig. 6, I–N). Interestingly,
in the right AVJ of the chick heart, the initial development of the lateral AV leaflet resembles that of the
mouse and human. However, instead of the regression
of muscular tissue, leading to the formation of a fibrous
leaflet, the mesenchymal tissue gets replaced by myocardium, resulting in the formation of the characteristic muscular flap valve in the right AVJ of the chick.
Immunohistochemical studies indicate that the mechanism that is responsible for this process is myocardialization, i.e., the ingrowth of existing myocardium
into flanking mesenchymal tissues. Experimental
studies have revealed that this myocardialization process in the avian right AV valve can be disturbed by
Physiol Genomics • VOL
15 •
surgical left atrial ligation (25). This procedure leads to
the formation of a (albeit dysmorphic) fibrous leaflet,
reminiscent of the situation in mouse and human.
In-depth studies of the molecular mechanisms involved
in the molecular regulation of this myocardialization
process might provide important clues regarding valvuloseptal morphogenesis in general, as myocardial-tofibrous and fibrous-to-myocardial “transformations”
are common steps in normal and perturbed cardiovascular development in the chick, mouse, and human
heart (13, 29, 30).
Myocardial Compaction
Ventricular trabeculation starts to develop in the
apical region of the ventricles soon after looping in both
mouse and human. The trabeculation serves primarily
as a means to increase myocardial oxygenation in absence of coronary circulation. At its peak prior to completion of ventricular septation, the trabeculae can
form as much as 80% of the myocardial mass in the
embryonic human heart (3) and provide most of the
www.physiolgenomics.org
REVIEW: ANATOMY OF MOUSE AND HUMAN HEART
ventricular wall thickness between ED11 and ED16 in
the mouse. Concomitant with ventricular septation,
the trabeculae start to compact at their base adjacent
to the outer compact myocardium, adding substantially to its thickness (reviewed in Ref. 26). This process is fairly rapid (in mouse, between ED13 and ED14,
in human between 10 and 12 wk; Ref. 26), coinciding
with establishment of the coronary circulation. From
ED16 in the mouse and the 4th mo of gestation in
human, the compact layer forms most of the ventricular myocardium, although its structural complexity
continues to develop (12). Noncompaction of the myocardium presents serious functional consequences for
the heart. Several mouse null mutants present with a
complete lack of compaction [e.g., RXR-␣ knockout
mouse (8); see Fig. 7], which results in lethality around
ED14 (reviewed in Ref. 26), emphasizing the importance of the compact myocardium for force generation
in later fetal stages. In humans, ventricular noncompaction is usually localized. Although it can occur in
isolation, it was found in association with heart failure
and sudden cardiac death (32). Even if only a relatively
small proportion of ventricular wall is affected, ventricular dynamics is perturbed (11). This condition can
now be diagnosed by ultrasound and is classified as a
distinct cardiomyopathy. There is a growing body of
evidence that the epicardium and the epicardially derived cells play a crucial role in the development of the
compact layer of the ventricular myocardial walls. Not
only is the “thin-myocardium syndrome” observed in
several mouse models with perturbed epicardial development [e.g., ␣4-integrin (45) and VCAM-1 (15) knockout mice], recent experimental studies in which epicardial development was perturbed in avian embryos also
demonstrated inhibition of ventricular compaction
(22). The mechanism(s) that are responsible for the
regulation of ventricular myocardial differentiation,
growth, and compaction are at present poorly understood, but it is to be expected that further analysis of
mutant mice with “thin compact myocardium,” such as
the RXR-␣ knockout mouse (5, 14), can provide clues on
the cascade of genetically regulated events that lead to
trabecular compaction and hence molecular etiology of
this condition in humans.
CONCLUSIONS
In this review we have demonstrated that, given the
considerable differences in cardiac size and heart rate,
the cardiac anatomy in mouse and human is, with the
exception of some small variations (mainly in the venous pole), remarkably similar. We have also shown
that, at least within the context of the topics discussed,
the developmental events that lead to the formation of
the four-chambered heart are also very comparable.
Consequently, the mouse can serve as a good model
with which to study the human heart. The genetic
modification of the murine genome is resulting in a
growing numbers of mouse models presenting with
cardiac malformations that are very reminiscent of the
heart anomalies seen in patients with congenital heart
Physiol Genomics • VOL
15 •
175
disease. We conclude that mouse molecular genetics
combined with detailed morphological assessment of
generated cardiac malformations in the genetically
modified mouse models will lead to important breakthroughs in the identification of a series of (candidate)
genes, and the elucidation of pathways, involved in the
etiology of human congenital heart disease.
ACKNOWLEDGMENTS
We thank Dr. Tom Trusk for preparing the informative cartoons
in Figs. 3 and 4, and we thank Dr. Steve Kubalak for providing the
illustrations of the RXR-␣ knockout mouse (Fig. 7).
GRANTS
A. Wessels and D. Sedmera are financially supported by National
Institutes of Health Grants PO1-HL-52813 (to A. Wessels), P20-RR16434 (to D. Sedmera), and PO1-HD-39946 (to A. Wessels and D.
Sedmera).
REFERENCES
1. Anderson RH and Becker AE. Cardiac Anatomy: An Integrated Text and Colour Atlas. London: Gower, 1980.
2. Anderson RH, Webb S, and Brown NA. The mouse with
trisomy 16 as a model of human hearts with common atrioventricular junction. Cardiovasc Res 39: 155–164, 1998.
3. Blausen BE, Johannes RS, and Hutchins GM. Computerbased reconstructions of the cardiac ventricles of human embryos. Am J Cardiovasc Pathol 3: 37–43, 1990.
4. Camenisch TD, Spicer AP, Brehm-Gibson T, Biesterfeldt
J, Augustine ML, Calabro A Jr, Kubalak S, Klewer SE, and
McDonald JA. Disruption of hyaluronan synthase-2 abrogates
normal cardiac morphogenesis and hyaluronan-mediated transformation of epithelium to mesenchyme. J Clin Invest 106: 349–
360, 2000.
5. Chen TH, Chang TC, Kang JO, Choudhary B, Makita T,
Tran CM, Burch JB, Eid H, and Sucov HM. Epicardial
induction of fetal cardiomyocyte proliferation via a retinoic acidinducible trophic factor. Dev Biol 250: 198–207, 2002.
6. De Haan RL. Morphogenesis of the Vertebrate Heart. New York:
Holt, Rinehart and Winston, 1965.
7. Eisenberg L and Markwald R. Molecular regulation of atrioventricular valvuloseptal morphogenesis. Circ Res 77: 1–6, 1995.
8. Gruber P, Kubalak S, Pexieder T, Sucov H, Evans R, and
Chien K. RXR alpha deficiency confers genetic susceptibility for
aortic sac, conotruncal, atrioventricular cushion, and ventricular
muscle defects in mice. J Clin Invest 98: 1332–1343, 1996.
9. Han W, Wang Z, and Nattel S. Slow delayed rectifier current
and repolarization in canine cardiac Purkinje cells. Am J Physiol
Heart Circ Physiol 280: H1075–H1080, 2001.
10. Huang GY, Cooper ES, Waldo K, Kirby ML, Gilula NB, and
Lo CW. Gap junction-mediated cell-cell communication modulates mouse neural. J Cell Biol 143: 1725–1734, 1998.
11. Jenni R, Oechslin E, Schneider J, Jost CA, and Kaufmann
PA. Echocardiographic and patho-anatomical characteristics of
isolated left ventricular non-compaction: a step towards classification as a distinct cardiomyopathy. Heart 86: 666–671, 2001.
12. Jouk PS, Usson Y, Michalowicz G, and Grossi L. Threedimensional cartography of the pattern of the myofibres in the
second trimester fetal human heart. Anat Embryol (Berl) 202:
103–118, 2000.
13. Kruithof BPT, Van den Hoff MJB, Wessels A, and Moorman AFM. Cardiac muscle cell formation after development of
the linear heart tube. Dev Dyn 227: 1–13, 2003.
14. Kubalak SW, Hutson DR, Scott KK, and Shannon RA.
Elevated transforming growth factor beta2 enhances apoptosis
and contributes to abnormal outflow tract and aortic sac development in retinoic X receptor alpha knockout embryos. Development 129: 733–746, 2002.
15. Kwee L, Baldwin HS, Shen HM, Stewart CL, Buck C, Buck
CA, and Labow MA. Defective development of the embryonic
www.physiolgenomics.org
176
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
REVIEW: ANATOMY OF MOUSE AND HUMAN HEART
and extraembryonic circulatory systems in vascular cell adhesion molecule (VCAM-1) deficient mice. Development 121: 489–
503, 1995.
Lakkis ML and Epstein JA. Neurofibromin modulation of ras
activity is required for normal endocardial-mesenchymal transformation in the developing heart. Dev Biol 125: 4359–4367,
1998.
Lamers WH, Viragh S, Wessels A, Moorman AF, and
Anderson RH. Formation of the tricuspid valve in the human
heart. Circulation 91: 111–121, 1995.
Lamers WH, Wessels A, Verbeek FJ, Moorman AF, Viragh
S, Wenink AC, Gittenberger-de Groot AC, and Anderson
RH. New findings concerning ventricular septation in the human heart. Implications for maldevelopment. Circulation 86:
1194–1205, 1992.
Markwald RR, Fitzharris TP, and Manasek FJ. Structural
development of endocardial cushions. Am J Anat 148: 85–119,
1977.
Oosthoek PW, Viragh S, Mayen AE, van Kempen MJ, Lamers WH, and Moorman AF. Immunohistochemical delineation
of the conduction system. I. The sinoatrial node. Circ Res 73:
473–481, 1993.
Oosthoek PW, Wenink AC, Vrolijk BC, Wisse LJ, DeRuiter
MC, Poelmann RE, and Gittenberger-de Groot AC. Development of the atrioventricular valve tension apparatus in the
human heart. Anat Embryol (Berl) 198: 317–329, 1998.
Perez-Pomares JM, Phelps A, Sedmerova M, Carmona R,
Gonzalez-Iriarte M, Munoz-Chapuli R, and Wessels A. Experimental studies on the spatiotemporal expression of WT1 and
RALDH2 in the embryonic avian heart: a model for the regulation of myocardial and valvuloseptal development by epicardially derived cells (EPDCs). Dev Biol 247: 307–326, 2002.
Qayyum SR, Webb S, Anderson RH, Verbeek FJ, Brown
NA, and Richardson MK. Septation and valvar formation in
the outflow tract of the embryonic chick heart. Anat Rec 264:
273–283, 2001.
Redkar A, Montgomery M, and Litvin J. Fate map of early
avian cardiac progenitor cells. Development 128: 2269–2279,
2001.
Sedmera D, Pexieder T, Rychterova V, Hu N, and Clark
EB. Remodeling of chick embryonic ventricular myoarchitecture
under experimentally changed loading conditions. Anat Rec 254:
238–252, 1999.
Sedmera D, Pexieder T, Vuillemin M, Thompson RP, and
Anderson RH. Developmental patterning of the myocardium.
Anat Rec 258: 319–337, 2000.
Seo JW, Kim EK, Brown NA, and Wessels A. Section directed
cryosectioning of specimens for scanning electron microscopy: a
new method to study cardiac development. Microsc Res Tech 30:
491–495, 1995.
Sissman NJ. Developmental landmarks in cardiac morphogenesis: comparative chronology. Am J Cardiol 25: 141–148, 1970.
van den Hoff MJ, Kruithof BP, Moorman AF, Markwald
RR, and Wessels A. Formation of myocardium after the initial
development of the linear heart tube. Dev Biol 240: 61–76, 2001.
van den Hoff MJ, Moorman AF, Ruijter JM, Lamers WH,
Bennington RW, Markwald RR, and Wessels A. Myocardialization of the cardiac outflow tract. Dev Biol 212: 477–490,
1999.
Varnava AM. Isolated left ventricular non-compaction: a distinct cardiomyopathy? Heart 86: 599–600, 2001.
Physiol Genomics • VOL
15 •
32. Waller BF, Smith ER, Blackbourne BD, Arce FP, Sarkar
NN, and Roberts WC. Congenital hypoplasia of portions of both
right and left ventricular myocardial walls. Clinical and necropsy observations in two patients with parchment heart syndrome. Am J Cardiol 46: 885–891, 1980.
33. Waller BR, McQuinn T, Phelps AL, Markwald RR, Lo CW,
Thompson RP, and Wessels A. Conotruncal anomalies in the
trisomy 16 mouse: an immunohistochemical analysis with emphasis on the involvement of the neural crest. Anat Rec 260:
279–293, 2000.
34. Waller RB III and Wessels A. Cardiac morphogenesis and
dysmorphogenesis II. An immunohistochemical approach. In:
Developmental Biology Protocols, edited by Lo CW. Totowa, NJ:
Humana, 2000, p. 151–161.
35. Webb S, Brown NA, and Anderson RH. The structure of the
mouse heart in late fetal stages. Anat Embryol (Berl) 194: 37–47,
1996.
36. Webb S, Brown NA, Wessels A, and Anderson RH. Development of the murine pulmonary vein and its relationship to the
embryonic venous sinus. Anat Rec 250: 325–334, 1998.
37. Wessels A, Anderson RH, Markwald RR, Webb S, Brown
NA, Viragh S, Moorman AF, and Lamers WH. Atrial development in the human heart: an immunohistochemical study
with emphasis on the role of mesenchymal tissues. Anat Rec 259:
288–300, 2000.
38. Wessels A, Markman MW, Vermeulen JL, Anderson RH,
Moorman AF, and Lamers WH. The development of the atrioventricular junction in the human heart. Circ Res 78: 110–117,
1996.
39. Wessels A and Markwald R. Cardiac morphogenesis and dysmorphogenesis. I. Normal development. Methods Mol Biol 136:
239–259, 2000.
40. Wessels A, Mijnders TA, de Gier-de Vries C, Vermeulen
JLM, Viragh S, Lamers WH, and Moorman AFM. Expression of myosin heavy chain in neonatal human hearts. Cardiol
Young 2: 318–334, 1992.
41. Wessels A, Vermeulen JL, Verbeek FJ, Viragh S, Kalman
F, Lamers WH, and Moorman AF. Spatial distribution of
“tissue-specific” antigens in the developing human heart and
skeletal muscle. III. An immunohistochemical analysis of the
distribution of the neural tissue antigen G1N2 in the embryonic
heart; implications for the development of the atrioventricular
conduction system. Anat Rec 232: 97–111, 1992.
42. Wessels A, Vermeulen JL, Viragh S, Kalman F, Lamers
WH, and Moorman AF. Spatial distribution of “tissue-specific”
antigens in the developing human heart and skeletal muscle. II.
An immunohistochemical analysis of myosin heavy chain isoform expression patterns in the embryonic heart. Anat Rec 229:
355–368, 1991.
43. Wessels A, Vermeulen JL, Viragh S, Kalman F, Morris GE,
Man NT, Lamers WH, and Moorman AF. Spatial distribution
of “tissue-specific” antigens in the developing human heart and
skeletal muscle. I. An immunohistochemical analysis of creatine
kinase isoenzyme expression patterns. Anat Rec 228: 163–176,
1990.
44. Ya J, Schilham MW, de Boer PA, Moorman AF, Clevers H,
and Lamers WH. Sox-4 deficiency syndrome in mice is an
animal model for common trunk. Circ Res 83: 986–994, 1998.
45. Yang JT, Rayburn H, and Hynes RO. Cell adhesion events
mediated by alpha 4 integrins are essential in placental and
cardiac development. Development 121: 549–560, 1995.
www.physiolgenomics.org