Developmental Biology 368 (2012) 203–213

Contents lists available at SciVerse ScienceDirect

Developmental Biology
journal homepage: www.elsevier.com/locate/developmentalbiology

Growth of the developing mouse heart: An interactive qualitative

and quantitative 3D atlas
Bouke A. de Boer 1, Gert van den Berg 1, Piet A.J. de Boer, Antoon F.M. Moorman, Jan M. Ruijter n
Heart Failure Research Center, Academic Medical Center, Amsterdam, The Netherlands

a r t i c l e i n f o abstract

Article history: Analysis of experiments aimed at understanding the genetic mechanisms of differentiation and growth
Received 11 January 2012 of the heart, calls for detailed insights into cardiac growth and proliferation rate of myocytes and their
Received in revised form precursors. Such insights in mouse heart development are currently lacking. We quantitatively assessed
4 April 2012
the 3D patterns of proliferation in the forming mouse heart and in the adjacent splanchnic mesoderm,
Accepted 3 May 2012
Available online 14 May 2012
from the onset of heart formation till the developed heart at late gestation. These results are presented
in an interactive portable document format (Suppl. PDF) to facilitate communication and under-
Keywords: standing. We show that the mouse splanchnic mesoderm is highly proliferative, and that the
Cardiac development proliferation rate drops upon recruitment of cells into the cardiac lineage. Concomitantly, the
Developmental biology
proliferation rate locally increases at the sites of chamber formation, generating a regionalized
Quantitative reconstructions
proliferation pattern. Quantitative analysis shows a gradual decrease in proliferation rate of the
Proliferation
ventricular walls with progression of development, and a base-to-top decline in proliferation rate in the
trabecules. Our data offers clear insights into the growth and morphogenesis of the mouse heart and
shows that in early development the phases of tube formation and chamber formation overlap. The
resulting interactive quantitative 3D atlas of cardiac growth and morphogenesis provides a resource for
interpretation of mechanistic studies.
& 2012 Elsevier Inc. All rights reserved.

Introduction
A central challenge in cardiac developmental biology is the
understanding of the mechanisms that effectuate the growth of
the heart. Proper formation of the four-chambered heart depends
on a delicate balance between growth and differentiation, which
is highly prone to errors, witness the high incidence of congenital
cardiac malformations (Hoffman and Kaplan, 2002). Insight into
these processes is crucial for the understanding of cardiac
morphogenesis, and the interpretation of cellular lineage analyses
(Buckingham et al., 2005) and molecular mechanistic data that
provide a great momentum to the field of cardiac developmental
biology (Bruneau, 2008; Evans et al., 2010; Vincent and
Buckingham, 2010).
The mouse embryo is currently the most important animal
model for molecular genetics (Rosenthal and Harvey, 2010).
However, most knowledge on cardiac morphogenesis comes from
classic studies of mostly chicken and human embryos (His, 1885;
n
Correspondence to: Academic Medical Center, Department of Anatomy, Embry-
ology & Physiology. Meibergdreef 15, 1105 AZ Amsterdam, The Netherlands.
Fax: þ31 20 697 6177.
E-mail address: j.m.ruijter@amc.uva.nl (J.M. Ruijter).
URL: http://www.hfrc.nl (J.M. Ruijter).
1
These authors contributed equally.
Romanoff, 1960). Also, data on cardiac proliferation are scarce and
rarely placed within a three-dimensional (3D) context, leading to
conflicting conclusions such as either rapid (Thompson et al.,
1990) or slow (Sissman, 1966) proliferation in the early heart
tube. Previously, we analyzed local proliferation rates within the
developing chicken heart and observed that proliferation stops
when myocardium differentiates from rapidly proliferating
splanchnic mesoderm, to be followed by a new phase of rapid
proliferation at the sites of chamber formation (Soufan et al.,
2006; van den Berg et al., 2009). A similar pattern is observed
during human heart formation (Sizarov et al., 2011).
Information on local proliferation rates in the developing
mouse heart is, however, entirely lacking. To generate a general
framework of the 3D architecture and growth of the developing
mouse heart, we performed a 3D analysis of morphogenesis and
proliferation of the developing mouse heart. Our analysis spans
the entire gestational period and the results are presented in an
interactive fashion. In contrast to chicken and human develop-
ment, where clearly separated phases of early heart development
can be distinguished, we show that in early mouse heart devel-
opment the phases of tube formation and chamber formation
overlap. The mouse heart, nevertheless, forms by differentiation
of splanchnic mesoderm into myocardium accompanied by a
slowdown of proliferation, which is followed by rapid prolifera-
tion at the site of chamber differentiation. During later chamber

0012-1606/$ - see front matter & 2012 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.ydbio.2012.05.001
204 B.A. de Boer et al. / Developmental Biology 368 (2012) 203–213
development the proliferation rate decreases progressively, while
proliferation in the trabecules shows a base-to-top decline at
every stage. All reconstructions are supplemented with interac-
tive 3D models, and thus form a comprehensive 3D atlas of the
growth and morphogenesis of the developing mouse heart.
Materials and methods
Animal handling and immunohistochemistry
To visualize proliferating cells, FVB mouse embryos were
exposed to Bromodeoxy-Uridine (BrdU) (Sigma nr. B5002) for
1 h. BrdU is a thymidine analogue which is incorporated into the
nucleus during the S-phase of the cell cycle (during DNA synth-
esis). Prior to embryonic day (E) 9 exposure was by culturing in
0.05 mg BrdU/ml medium (DMEM (Invitrogen), pH 7.4, supple-
mented with 10% foetal calf serum). Older stages were exposed by
peritoneal injection of pregnant mice with 50 mg BrdU/ kg (using
a 10 mg BrdU/ ml 0.9% NaCl solution), as previously described
(Aanhaanen et al., 2009). Immediately after exposure, embryos
were fixed in freshly prepared 4% paraformaldehyde. Embryos
were staged in days of development according to their general
and their cardiac morphological characteristics (Bard et al., 1998;
Kaufman, 1994; Theiler, 1972). It is important to appreciate that
early heart development progresses very fast, especially in a few
hours around day 8 of embryonic development, which means that
the developmental variation within a single mouse litter can
range from cardiac crescent to tubular heart stages. All experi-
mental procedures were in line with institutional and national
regulations for animal welfare.
Standard procedures were used for dehydration and embed-
ding in paraffin (Mommersteeg et al., 2010). Embryos were
serially sectioned at 7 mm. Mounted sections were dewaxed,
rehydrated and equilibrated in phosphate-buffered saline (PBS),
after which antigens were retrieved by pressure cooking for 3 min
in antigen unmasking solution (1:100 Vector laboratories Inc.
H-3300). After a PBS wash, sections were blocked with 2% bovine
serum albumin and exposed overnight to a mixture of primary
antibodies; rat-monoclonal anti-BrdU (1:600, Immunosource
OBT0030CX), rabbit polyclonal anti-cTnI (1:250 HyTest 4T21/2),
monoclonal anti-MHC (Mf20, Hybridomabank, Iowa City, IA, USA)
and goat polyclonal anti-Nkx2.5 (1:200 Santa Cruz 8697). After
washing, secondary antibodies were added for 2 h periods. Firstly,
donkey–anti-goat Alexa-680, and, again after washing, goat–anti-
rat Alexa-405 and goat–anti-rabbit Alexa-568 (Invitrogen,
A-21084, A-31556, A-11077, respectively). After washing, Sytox
Green (1:40,000 Molecular Probes S-7020) was added for 15 min
before a last wash and mounting of a cover slip with either PBS/
glycerol or Vectashield (Vectorlabs H1000).
3D-reconstruction and morphometry
Images of each channel were acquired with a fluorescence
microscope (Leica DM6000) at 10 times magnification, generating
image series with pixels representing 0.9  0.9 mm2 tissue. Upon
visual inspection, damaged sections were replaced by neighbour-
ing sections. Images were loaded into Amira (Mercury Computer
Systems) and, with an automatic module, aligned by translation
and rotation. This procedure was visually inspected and corrected
where necessary. In the aligned stacks of images, anatomical
structures were manually segmented based on either a specific
signal (Mf20, cTnI or Nkx2.5), or on morphological criteria
(mostly using the Sytox Green signal). In E7.75 embryos, Nkx2.5
could be used as an early cardiogenic marker to identify the
heart-forming region in the splanchnic mesoderm. After
segmentation, the myocardium and splanchnic mesoderm labels
were inspected to ensure that all nuclei were included. 3D surface
reconstructions were generated as previously described (Fig. 1,
top row) (van den Berg and Moorman, 2011). For interactive 3D
presentation, 3D-pdf files were generated as previously described
(de Boer et al., 2011).
Detection and quantification of nuclei was performed using a
program in Matlab (The Mathworks). The complete procedure is
illustrated in Fig. 1. In short, background staining was removed
from the Sytox Green image by subtracting the local average from
the image. The resulting local maxima were then thresholded into
binary nuclei. This binary image was masked by the segmented
label of either the myocardium or the splanchnic mesoderm. For
nuclei within these masks, a positive BrdU-signal was defined as a
nucleus with a staining intensity of at least a standard deviation
above its local background intensity (Fig. 1, middle row) resulting
in a 3D representation of the BrdU-positive and BrdU-negative
cardiomyocyte nuclei (Fig. 1, bottom row). From E10 onwards
non-myocytes gradually invade the heart (Zhou et al., 2008) and,
therefore, Nkx2.5-staining was used to identify cardiomyocyte
nuclei from stage E9.5 onwards. These Nkx2.5-positive nuclei
were identified similar to the BrdU-positive nuclei. The BrdU-
positive fraction at each location was then determined by count-
ing the total number of nuclei and the number of BrdU-positive
nuclei in a cubic sampling volume surrounding the location and
by plotting the resulting fraction into the target volume at the
location (Fig. 1, bottom row) (Soufan et al., 2007). The sampling
volume (1053 mm3) contains enough nuclei to ensure reliable
BrdU-positive fractions whereas the target volume (213 mm3) is
small enough to avoid loss of biological resolution. This measure-
ment procedure serves to recover the continuous proliferation
information (Fig. 1, bottom right) from the binary information
(Fig. 1, bottom left). Based on the total number of nuclei in the
sample volume the maximum width of the 95% confidence
interval of the BrdU-positive fraction at each location is
0.5070.07; for higher and lower fractions the confidence range
is even narrower (Soufan et al., 2007).
Proliferation rate in the ventricles was also measured relative
to the trabecular base which is defined as the junction of the
trabecules and the compact myocardium. The trabeculations of
both ventricles were separately segmented. For each position in
the trabecules and the compact myocardium, the BrdU-positive
nuclei were counted and the Euclidean distance to the nearest
trabecular base was determined. For embryos between E10, and
E12.5 this was done in cubic volumes of 7.2  7.2  7 mm3, and for
E14.5 and E17.5 embryos in 13.6  13.6  14 mm3. Subsequently,
the counted nuclei and BrdU-positive nuclei were pooled in
15 mm wide distance categories, which ensured that enough
nuclei were counted in each distance zone to reliably determine
the BrdU-positive fraction.
The original image data on which the 3D reconstructions are
based are available from the authors on request.
Results
Morphometry: number of cardiomyocytes and cardiac volume
To quantify growth during the formation of the mouse heart
we counted nuclei in, and measured the volume of, developing
hearts between E8- and E17.5, thus ranging from the initially
formed cardiac crescent to the prenatal four-chambered heart
(Fig. 2). Because the cardiomyocytes are mono-nuclear until post-
natal development (Liu et al., 2010) the number of nuclei
represents the number of cardiomyocytes. The non-linear loga-
rithmic relation of the number of cells and of the total myocardial
 B.A. de Boer et al. / Developmental Biology 368 (2012) 203–213 205

Reconstruction and quantification process

Reconstruction
Cell detection

20 μm 20 μm

BrdU- and
BrdU+( nuclei
Visualization

0
25

50

75
10
0

BrdU- (red) and Labeling fraction % BrdU-positive nuclei Projection of labeling fraction
BrdU+(green) nuclei on 21 μm3 boxels Section Reconstruction

Fig. 1. Reconstruction and quantification. Overview of the 3D reconstruction and quantification procedures. The upper row shows on the left a section stained for cTnI
expression which is segmented to enable 3D reconstruction. The middle position shows this reconstruction up till the current section; in the left view the reconstruction is
completed. The middle row shows on the left a section on which, based on a Sytox green staining, all nuclei within the myocardium are identified. The middle panel shows
that within and around each nucleus the signal from the BrdU images is determined and compared. When the BrdU signal in the nucleus is at least a standard deviation
higher than the background signal around the nucleus, the nucleus is classified as positively labelled. In the right panel, nuclei that are positive (green) and negative (red)
for BrdU are shown. The third row shows a 3D reconstruction of all detected nuclei, which are either positive (green) or negative (red) for BrdU. Next to this the 213 mm3
target volumes are shown into which the BrdU labelling fractions of the 1053 mm3 sample volume are projected. These fractions can then be projected on the original
section and on the complete 3D reconstruction.

volume with developmental time indicates that the heart grows
relatively fast in the younger stages and that a growth rate slows
down in the oldest stages (Fig. 1). From day 8- in development up
to day 11 the number of cardiomyocytes increases from 700 to
68,000, whereas a similar 100-fold increase is observed in cardiac
volume (from 0.0014 to 0.150 mm3, Table 1). After E11 both
growth rates taper off further, but still show an almost 13-fold
increase approaching 1 million cardiomyocytes and a volume of
almost 2 mm3 in the E17.5 heart. The parallel development of the
number of cells and of total volume suggests that the average
volume of cardiomyocytes is constant (approx. 2150 mm3)
throughout embryonic and foetal development. Note that the
graph gives no information on how the number of cells increase,
nor does it imply that the growth rate is equal for the whole heart
or that all cardiomyocytes have the same size. The quantitative
3D reconstructions presented below indeed show a highly regio-
nalized pattern of proliferation.
Morphology and proliferation pattern
The reader is encouraged to read this text while referring to
the interactive 3D-pdf file, which is available in the online Suppl.
PDF.
Initial heart tube formation (Fig. 3)
Mouse development occurs on the bottom of a deeply invagi-
nated yolk sac (Fig. 3a and d). This is in contrast to early chicken
and human development, which occurs on an almost flat yolk sac.
In chicken, the foregut lengthens more-or-less parallel to the
surface of the yolk sac and has a narrow anterior intestinal portal
(van den Berg et al., 2009). Due to the invagination of the yolk sac
in mouse, the foregut lengthens in a perpendicular orientation to
the yolk sac, and the anterior intestinal portal is wide (Kaufman
and Navaratnam, 1981).
206 B.A. de Boer et al. / Developmental Biology 368 (2012) 203–213

Quantification of number of cells and volume

10,000,000 10

Total volume myocardium

1,000,000 Total # cardiomyocytes 1
500 μm
Number of cells

Volume (mm3)
100,000 0.1

10,000 0.01

1,000 0.001

100 0.0001
7 9 11 13 15 17 19
Embryonic day

Fig. 2. Quantification of number of cells and volume. The number of cardiomyocytes in the developing mouse heart were derived from the number of nuclei observed within
the tissue labelled by the specific staining with a cardiac marker. The number of nuclear profiles per tissue area was converted into number per volume according to
Abercrombie’s equation (Weibel, 1979), thus accounting for section thickness and average nuclear diameter. Myocardial volume was calculated according to the Cavalieri
principle (Howard and Reed, 1998). Note that the data are displayed on the logarithmic Y-axes. Reconstructions of the myocardium ranging from embryonic day 8- to 17.5,
all displayed at the same scale, are added.

Table 1 mesoderm remain in close proximity to the foregut, forming the

Numerical summary of the analyzed individual mouse hearts. For each heart the pericardial back wall with progressive folding (Suppl. Fig. 1e). At
total number of cells (from E9.5 onwards only nkx2.5 positive cells) within the
the site of vessel formation a drop in proliferation rate is seen in
heart, the total myocardial volume and the percentage of BrdU labelled cells
are given. the left splanchnic epithelium (Suppl. Fig. 1f).
At E8- the cardiovascular lumen is formed. Moreover, the
Embryonic day Total #cells Volume (mm3) % BrdU gutters, which form in the lateral aspects of the splanchnic
mesoderm, have differentiated into Mf20-positive myocardium
8- 648 0.0014 33
8- 721 0.0013 35
(Fig. 3e and j). This differentiation is accompanied by a drop in
8 1069 0.0025 32 proliferation rate within the myocardium (Fig. 3j). With further
8 1240 0.0024 35 elongation of the foregut, the medial aspects of the splanchnic
8 1549 0.0031 34 mesoderm grow, forming the pericardial back wall (Suppl. Fig. 2a
8 1321 0.0022 37
and d). Within the heart tube, the endocardium shows several
8.25 1751 0.0033 36
8.25 2071 0.0043 39 finger-like protrusions, adhering to the myocardium (Fig. 3b).
8.5 3332 0.0064 37 With progressive fusion, the lateral myocardial gutters fuse in
9.5 16745 0.0285 33 the ventral midline (Fig. 3f and k) which can be discriminated by a
9.5 21319 0.0487 33
ventral fusion seam. The medial splanchnic mesoderm now forms
10.5 34998 0.0826 27
11 53067 0.1480 28
a longer pericardial back wall, in line with growth of the foregut.
11 68417 0.1493 26 Unlike observations in chicken (van den Berg et al., 2009), but
12.5 135053 0.2941 25 resembling the pattern observed in human development (Sizarov
12.5 162372 0.4573 22 et al., 2011), all splanchnic mesoderm in mouse is fast-proliferat-
14.5 415187 1.0830 23
ing (Fig. 3g–i). The endocardial tubes are entirely suspended in
17.5 919598 1.8611 21
cardiac jelly, as in the human heart (Sizarov et al., 2011). The
endocardium within the heart tube is fused and shows multiple
At E7.75 the cardiogenic part of the splanchnic mesoderm, ventral protrusions. The adjacent ventral part of the heart tube
expressing the transcription factor Nkx2.5, is present as bilateral shows an increase in proliferation rate, especially on the left side
plates of epithelium (Fig. 3d). The foregut just begins to form and (Fig. 3k). The dorsal side of the heart tube is still completely open
can be identified as a ventral protrusion. No cardiovascular lumen and displays a low proliferation rate (Suppl. 3D PDF).
can be discerned yet, which is consistent with previous analyses
of the expression of molecular vessel markers (Wood et al., 1997). Lengthening and looping of the heart tube (Fig. 4)
This flat pre-cardiac mesoderm displays rapid proliferation
(Fig. 3g), as was also observed in chicken (van den Berg et al., Fig. 4 shows 3D reconstructions of mouse hearts ranging from
2009). Shortly later some endothelial cells are observed between E8- to E8.5. Other embryonic structures that provide a spatial
the lateral sides of the splanchnic mesoderm and the endoderm context and indicate developmental progression are shown in the
(Suppl. Fig. 1d–f). The lateral mesoderm forms a gutter (Suppl. interactive Suppl. 3D-pdf. At E8 the ventral fusion seam of the
Fig. 1d). In line with the progression of development and folding, myocardium is no longer discernable (Fig. 4d). Compared to E7.75
the foregut elongates. The medial aspects of the splanchnic the endocardium shows more protrusions adhering to the
 B.A. de Boer et al. / Developmental Biology 368 (2012) 203–213 207

Morphology and proliferation in initial mouse heart-tube formation

Embryonic day 7.75 Embryonic day 8-(a) Embryonic day 8-(b)
a b c
foregut

lateral view ventral view ventral view

d e f
yolk sac
(transporent)
ventral view

ic lef
hn t
c

sp
an

lan
spl

ch
t
righ

nic
Splanchnic mesoderm

g h i
Proliferation

j k
Myocardium

myocardium splanchnic
mesoderm fusion seam
endoderm cardiovascular
lumen
0
25
50
75
10

200 μm
0

% BrdU-positive nuclei

Fig. 3. Morphology and proliferation in initial mouse heart-tube formation. The figure shows different views of three reconstructions of mouse heart development, between
embryonic day (E) 7.75 and 8-. The middle and right column both represent stage E8-, but with a further progressed cardiac development in the right column. Panel (a) and
(d) shows the relation of the forming heart with the endoderm of the yolk sac. Proliferation rate in the splanchnic mesoderm and in the developing myocardum is shown in
panels (g)–(i) and (j) and (k), respectively. Other views and structures as depicted. (Abbreviation: AIP: anterior intestinal portal). Note that for the sake of clarity only part
of the splanchnic mesoderm was included in the reconstructions.

myocardium (Fig. 4a). The foregut (Suppl. Fig. 2c) elongates future atrioventricular canal (AVC). Note that this does not imply
further and the pericardial back wall extends by the concomitant a definite indication of lineage, since these cells will eventually
growth of the medial splanchnic mesoderm (Fig. 4g). Moreover, contribute to the left ventricular free wall (Aanhaanen et al.,
closure of the dorsal mesocardium starts by fusion of the left and 2009). Within the myocardium of the future AVC slower prolif-
right components of the pericardial back wall (Fig. 4g and m). The eration is observed (Fig. 4o), consistent with previous studies
pattern of proliferation resembles that of the previous stage, (Aanhaanen et al., 2009). The lengthening of the heart tube
displaying rapid proliferation in the ventral heart tube (Fig. 4j) and progresses, as is revealed by the grown outflow tract and the
the splanchnic mesoderm (Fig. 4g) and slower proliferation where further bulging of the prospective ventricle and the caudal
the myocardium is connected to the splanchnic mesoderm (Fig. 4m). displacement of its apex (Fig. 4f). The dorsal mesocardium is
A few hours later, at E8.25, the lengthening of the heart tube ruptured completely, leaving the heart tube suspended only by its
and closure of the pericardial back wall continues and fusion of the arterial and venous poles (Fig. 4f and i).
endocardial tubes progresses. The heart tube loops, and its elonga-
tion at the arterial pole becomes apparent by a clearly distinguish- Formation of the four-chambered heart (Figs. 5 and 6)
able outflow tract (Suppl. Fig. 3g). Upstream from the outflow tract,
relative to the blood flow, the future left ventricle forms at the Figs. 5 and 6 show reconstructions of the embryonic mouse
outer curvature of the looped heart. At E8.5, fusion of the hearts from E9.5 through E12.5 and from E14.5 through E17.5,
endocardial tubes at the inflow of the heart tube is completed respectively. At E9.5 the outflow tract has elongated, and shows
and the resulting fused intra-cardiac lumen is slightly displaced to the characteristic bend (Fig. 5d). Only the distal part of the OFT
the left of the embryonic midline (Fig. 4b). The ventricular region (downstream from the bend) shows slow proliferation. In con-
bulges ventrally and caudally accompanied by rapid proliferation trast, closer to the left ventricle the outflow tract shows higher
at the outer curvature of the outflow tract and the forming proliferation rates along with formation of trabecules (Fig. 5g),
ventricle (Fig. 4k); foci with rapid proliferation appear to be indicating differentiation into right ventricular myocardium
associated with endocardial protrusions. The inner curvature still (Zaffran et al., 2004). A clear ventricular septum cannot be
shows slower proliferation and is now completely closed (Fig. 4n). discerned, although an indentation between both ventricles is
At E8.5 the intra-cardiac portion of the inflow vasculature is present. The AVC is still clearly recognizable between the left
located left from the midline (Fig. 4c). The endocardium at the ventricle and the atria. At the inflow, the left and right atria are
inflow shows a constriction (Fig. 4c) that we designate as the formed, only proliferating at the lateral sides (Fig. 5d and j). The
208 B.A. de Boer et al. / Developmental Biology 368 (2012) 203–213

Morphology and proliferation in early looping mouse heart

Embryonic day 8 Embryonic day 8.25 Embryonic day 8.5
a b c
lumen - ventral

od
ds
blo
islan
Morphology
myocardium and mesoderm - ventral

d e f
splanchnic mesoderm - ventral

g h i
Proliferation

j k l
ventral
myocardium

m n
dorsal

o
0
25

50

75
10

splanchnic cardiovascular
0

myocardium mesoderm lumen 200 μm % BrdU-positive nuclei

Fig. 4. Morphology and proliferation in early looping mouse heart. Panels (a)–(o) show different views of three reconstructions of mouse heart development, between
embryonic day 8 and 8.5. Panels (a)–(c) show the intra cardiac vasculature (myocardium shown in transparent grey). Panels (d)–(f) show the myocardium (grey) and the
splanchnic mesoderm (yellow). Proliferation rates in the splanchnic mesoderm and myocardium are shown in panels (g)–(i) and (j)–(o), respectively.

confluence of the systemic veins is shifted towards the right. At
E9.5 the highest rates of proliferation are found in the developing
chambers. One day later, at E11 (Fig. 5h), the indentation between
the ventricles has become more prominent and a clear trabecular
ridge has developed. The AVC is now suspended over this future
interventricular septum enabling a direct connection of the atria
to their respective ventricles (Fig. 5h). At this stage the distal
outflow tract still is almost devoid of BrdU-incorporation (Fig. 4e)
indicating that its growth occurs almost exclusively by addition of
cells (Zaffran et al., 2004).
 B.A. de Boer et al. / Developmental Biology 368 (2012) 203–213 209

Mid gestational stages

Embryonic day 9.5 Embryonic day 11 Embryonic day 12.5

Lumen
a b c

d e f

ium
at r
ht
rig
Ventral

g h i
right right left
right atrium left
left atrium atrium atrium
atrium atrium

avc avc avc

right left
ventricle ventricle

right right
ventricle septum left ventricle left
ventricle septum ventricle
j k l
Dorsal

cardiovascular
lumen
500 µm
0
25
50
75
10
0

% BrdU-positive nuclei

Fig. 5. Morphology and proliferation of mid-gestational embryonic mouse hearts. Reconstructions of the myocardium and lumen of mouse hearts ranging from embryonic day
9.5–12.5 are shown. Panels (a)–(c) show the development of the cardiovascular lumen in relation to the myocardium (shown transparently). Panels (d)–(l) show
proliferation in the myocardium. Panels (a)–(f) show a ventral views, panels (g)–(i) show a ventral surface cut of the same hearts, and panels (j) and (k) show dorsal views.
All reconstructions are also interactively available in the supplements. (Abbreviations: BrdU: BromodeoxyUridine, AVC: atrioventricular canal)
210 B.A. de Boer et al. / Developmental Biology 368 (2012) 203–213

Fetal stages outflow tract gradually differentiates into the right ventricle
Embryonic day 14.5 Embryonic day 17.5 (Fig. 5d–f, Fig. 6c and d), finally generating a fully septated four
chambered heart. At E11 and E12.5 proliferation in the outflow
a b tract remained slow but both ventricles still show BrdU-positive
fractions (Fig. 5e and f) that are similar to those at E8.5 (Fig. 4i and
l). The proliferation rate in the ventricular wall decreases steadily
in later stages (Fig. 6c–h).
The BrdU-positive fractions in both ventricles at E9.5 till
E17.5 were measured in relation to the base of the trabecules
Lumen

(Fig. 7). The youngest stages show the highest proliferation

rate in the ventricular wall close to the trabecular base. This
proliferation rate decreases with increasing distance to the
base. In later stages the BrdU-positive fractions are lower and
c d this gradient becomes less obvious, coinciding with the disap-
pearance of the fast proliferating foci on the surface of the
ventricle (Fig. 6d and h). Although at all stages the BrdU-positive
fractions in the trabecules are about 10% lower than in the
outer wall, the observed decrease in proliferation rate with
increasing distance from the trabecular base remained signifi-
cant at all stages.

Discussion

The comprehensive series of 3D reconstructions that we

e f
Ventral

prepared provides insights into both morphology and prolifera-

right
atrium left atrium left atrium tion pattern in the developing mouse heart and can thus be used
to interpret the growing body of scientific data harvested in
the mouse.
t
right
atrium
Mouse heart development

Whereas in chicken and human the different phases of early

heart development, namely the formation of the tube, the tube
right
right
ventricle
elongation phase and the chamber formation, are clearly sepa-
ventricle left left
septum ventricle
rated in time, our 3D reconstructions show that during early
septum ventricle
mouse heart development these different growth phases overlap.

g h Initial tube formation

In chicken, the differentiation of mesoderm into myocardium
occurs concomitantly with midline fusion of the left and right
lateral mesoderm (OConnor and Runquist, 2008;Stalsberg and de
Haan, 1969). In mouse, due to its wide intestinal portal, the lateral
Dorsal

aspects of the splanchnic mesoderm cannot easily adjoin, and

myocardial differentiation occurs prior to midline fusion
(Kaufman and Navaratnam, 1981). Although this phase of cardiac
development in the mouse is generally known as the cardiac
crescent, if compared to chicken development this stage repre-
cardiovascular
lumen sents a laterally orientated, not yet fused, straight heart tube.
Progressive folding, as seen by lengthening of the foregut, causes
0

500 µm
10
25
50
75
0

% BrdU-positive nuclei
Fig. 6. Morphology and proliferation of foetal mouse hearts. Reconstructions
of the myocardium and lumen of mouse hearts of embryonic day 14.5 and 17.5
are shown. Panels (a) and (b) show the development of the cardiovascular lumen
in relation to the myocardium (shown transparently). Panels (d)–(h) show pro-
liferation in the myocardium. Panels (a)–(f) show a ventral views, panels (e) and
(f) show a ventral surface cut of the same hearts, and panels (f) and (g) show
dorsal views. All reconstructions are also interactively available in the supple-
ments. (Abbreviations: BrdU: BromodeoxyUridine, AVC: atrioventricular canal)
During further development, the overall morphology in the
ventricles and outflow tract remains largely similar, although
volume growth continues. From E12.5 onwards, the ventricular
septum progressively forms and shows a high proliferation rate
except at its top where proliferation is slower (Fig. 5i) as reported
previously (Bakker et al., 2008). The trabeculated part of the
the left and right myocardial components of this early heart tube
to finally join in the midline (Mommersteeg et al., 2010) which
can still be appreciated by the ventral seam seen at E8- (Fig. 3k).
As in chicken (Soufan et al., 2006; van den Berg et al., 2009) the
proliferation rate drops when mesoderm differentiates into car-
diomyocytes. However, the re-initiation of proliferation that was
observed in chicken after looping at the site of chamber formation
(Soufan et al., 2006), already occurs at this stage, most promi-
nently at the left side. The different growth phases observed in
chicken (van den Berg et al., 2009) and human heart development
(Sizarov et al., 2011) are thus represented in this one stage of
mouse heart development. During further development a slowing
down of proliferation is observed when cells pass through the
transition of mesoderm to myocardium, whereas the foci of fast
proliferation at the outer curvature are a sign of the start of
chamber development.
 B.A. de Boer et al. / Developmental Biology 368 (2012) 203–213
Quantifications in ventricles
wall
Right ventricle
Distance from trabecular base (μm)
wall
Left ventricle
Fig. 7. Local proliferation rate in older embryonic mouse hearts. Graphs showing the detailed quantification of proliferation in the ventricular walls and trabeculations.
The y-axes indicate the BrdU-labelled fractions. The x-axes represent the distance from the trabecular base. Error bars are standard deviations derived from the normal
approximation to the binomial distribution and thus reflect the measurement error.
Lengthening and looping of the heart tube
As described in chicken (Moreno-Rodriguez et al., 2006),
closure of the dorsal wall of the heart tube starts in the middle
and proceeds in the cranial and the caudal direction to ‘‘zipper
up’’ the heart tube. In contrast to chicken development, elonga-
tion of the outflow tract in mouse is initiated prior to rupture of
the dorsal mesocardium. However, in both species the distance
between the arterial and venous poles remains constant while the
heart tube rapidly lengthens. In mouse this distance is approxi-
mately 200 mm (Patten, 1922); in chicken at comparable stages
this distance measures about 700 mm (van den Berg et al., 2009).
In chicken the caudal part of the splanchnic mesoderm harbours a
fast proliferating growth centre (van den Berg et al., 2009); in
mice, however, the whole dorsal region displays the highest
proliferation rate and the cranial, slower proliferating region that
is present in chicken does not exist. This overall high proliferation
rate is probably a requirement because of the small size of the
region that needs to provide the cells that are added to both poles
of the heart (Cai et al., 2003; Kelly et al., 2001; Meilhac et al.,
2004; Snarr et al., 2007; Waldo et al., 2001; Zaffran et al., 2004).
The precise driving forces of cardiac looping remain unclear
(review: (Taber, 2006)), although it has long been recognized in
chicken (Patten, 1922) that ventral looping occurs of necessity
because the heart tube lengthens faster than the distance between
its arterial and venous poles. Our reconstructions show that this
principle may also be valid during mouse heart formation.
Chamber formation
The regionalization of proliferation, which can already be seen
in the developing ventricular wall in early stages might be due to
a genetic interaction at the endo- and myo-cardial connections,
which are known sites where Notch signalling mediates, prolif-
eration and trabecular differentiation (Grego-Bessa et al., 2007).
Such endo- and myo-cardial connections are present in the
ventral, fastest proliferating, part of the heart tube. Evidence for
chamber differentiation was previously demonstrated to occur
already at the 3–4 somites stage; using electron microscopy,
formation of gap-junctions was observed in the ventral region of
211
trabeculations
trabeculations
the developing heart, which also showed mitoses and formation
of a multilayered and trabeculated wall (Navaratnam et al., 1986).
Proliferation and morphogenesis
Evidence has accumulated that, after formation of the primary
heart tube, cells from the second heart field are added to the heart
tube at the arterial pole of the heart (Kelly et al., 2001; Mjaatvedt
et al., 2001; Waldo et al., 2001). The primary heart tube would
then originate from a first heart-forming field. It was later
demonstrated that the second heart field also contributed cells
to the venous pole of the heart (Cai et al., 2003; Snarr et al., 2007)
and that two lineages of precursor cells contribute to the heart
(Meilhac et al., 2004).
A separation of the heart-forming region does occur with the
rupture of the dorsal mesocardium, limiting the contact between
the forming heart tube and the pericardial back wall to the
arterial and venous poles. The distance between the arterial and
venous poles is then still only about 200 mm, or a column of about
15–20 precursor cells. Within this small region distinct anterior
and posterior domains of gene expression have been proposed
and prior to rupture of the dorsal mesocardium the medial part of
this region contained the precursors for the left ventricle, or first
heart field (Kelly and Buckingham, 2002). Our analyses of mor-
phogenesis and proliferation cannot address these issues, but
reveal a continuing temporal formation of the heart from pre-
cursors in the dorsal mesoderm, which in all stages shows the
overall highest proliferation rate.
As in human and chicken, proliferation in the mouse myocardium
is highly regionalized, with slower proliferation in the atrioventricular
canal, outflow tract and inner curvature, and rapid proliferation in the
forming chambers. The patterns of clonal growth within the myo-
cardium observed by Meilhac and coworkers (Meilhac et al., 2004;
Meilhac et al., 2004; Meilhac et al., 2003) are in agreement with the
observed proliferation patterns. Meilhac and co-workers observe a
dispersive growth phase, generating cells that intermingle during
recruitment into the heart, followed by a coherent growth phase
leading to clusters of daughter cells (Meilhac et al., 2003). Our current
observations show that the highly proliferative cardiac precursors in
212 B.A. de Boer et al. / Developmental Biology 368 (2012) 203–213
the dorsal mesocardium are separated from the proliferating cham-
ber-forming myocytes by a zone of non-proliferative newly differ-
entiated cardiomyocytes. This indicates that both growth phases
occur simultaneously. The start of the coherent growth phase, at
E8.5 (Meilhac et al., 2003), coincides with the observed re-initiation of
proliferation at the site of chamber formation and can thus explain
the occurrence of most and largest clones in the developing chambers
(Meilhac et al., 2004). The observed variation in cluster size in
different cardiac compartments (Meilhac et al., 2004a; Meilhac
et al., 2004b) may result from the local differences in proliferation
rate thus leading to remodelling of the heart tube.
Formation of the trabecules
During chamber differentiation, the walls of the left and right
ventricles form trabeculated myocardium on the luminal side,
while the remaining outer lining later differentiates into compact
myocardium. The trabeculated myocardium and the outer myo-
cardial wall adopt distinct programs of gene expression, reflecting
both changes in function and differential modes of growth.
Our analysis of proliferation within the myocardial walls and
the trabecules shows the most rapid proliferation at the junction
between the outer wall and the base of the trabecules, while
proliferation rate tapered off towards the top of the trabecules.
This persistent base-to-top gradient, in both ventricles, showed
that the focus of proliferation of cardiomyocytes in the forming
ventricles is at or near the base of the trabecules. Similarly,
retroviral labelling of pre-cardiac mesoderm in chicken showed
elongated clusters of labelled cells in the myocardial wall,
extending into the trabecules (Mikawa et al., 1992). Genetic
lineage analyses in mouse, using randomly generated clones, also
showed orientation of growth along the transmural axis (Meilhac
et al., 2003). In combination with the fast proliferation observed
at the base of the ventricular trabecules, these data suggest
growth of the ventricles by apposition, thereby elongating the
trabecules and causing an outward expansion of the ventricular
chambers (Moorman and Christoffels, 2003). This conclusion is
further supported by the observation that activated Notch signal-
ling, which regulates the proliferation and differentiation of the
trabecules, is most abundant at the base of the trabecules (Grego-
Bessa et al., 2007). Regression of proliferation in the trabecules
has been suggested to imply terminal differentiation into the
conduction system (Sedmera et al., 2003). Our study shows,
however, that both compartments continue to proliferate, albeit
at different rates. Furthermore, both compartments differentiate
into distinct phenotypes. Therefore, a strict inverse relationship
between differentiation and proliferation cannot be assumed.
Taken together, our analyses of proliferation rate in the
developing mouse heart from the initial formation of the heart
up to the late prenatal four-chambered heart forms a valuable
resource for future mechanistic studies. The interactive 3D
reconstructions allow for an independent inspection of the 3D-
architecture in relation to this important parameter of growth.
Source of Funding
This study was funded by the Netherlands Heart Foundation,
Grant no. NHS1996M002 (GvdB), and the EU (CHeartED), Grant
no. Health-F2-2008-223040 (BAdeB).
Acknowledgments
Alexandre Soufan, Stan Hoogcarspel and Tilly Mommersteeg
are acknowledged for their valuable contributions to the initial
experiments. Jaco Hagoort is thanked for creating the interactive
3D-pdfs.
Appendix A. Supporting information
Supplementary data associated with this article can be found in
the online version at http://dx.doi.org/10.1016/j.ydbio.2012.05.001.
References
Aanhaanen, W.T., Brons, J.F., Dominguez, J.N., Rana, M.S., Norden, J., Airik, R.,
Wakker, V., de Gier-de Vries, C., Brown, N.A., Kispert, A., Moorman, A.F.,
Christoffels, V.M., 2009. The Tbx2þ primary myocardium of the atrioventri-
cular canal forms the atrioventricular node and the base of the left ventricle.
Circ. Res. 104, 1267.
Bakker, M.L., Boukens, B.J., Mommersteeg, M.T.M., Brons, J.F., Wakker, V., Moor-
man, A.F.M., Christoffels, V.M., 2008. Transcription factor Tbx3 is required for
the specification of the atrioventricular conduction system. Circ. Res. 102,
1340–1349.
Bard, J.B.L., Kaufman, M.H., Dubreuil, C., Brune, R.M., Burger, A., Baldock, R.A.,
Davidson, D.R., 1998. An internet-accessible database of mouse developmental
anatomy based on a systematic nomenclature. Mech. Dev. 74, 111–120.
Bruneau, B.G., 2008. The developmental genetics of congenital heart disease.
Nature 451, 943–948.
Buckingham, M., Meilhac, S., Zaffran, S., 2005. Building the mammalian heart from
two sources of myocardial cells. Nat. Rev. Genet. 6, 826–837.
Cai, C.L., Liang, X., Shi, Y., Chu, P.H., Pfaff, S.L., Chen, J., Evans, S., 2003. Isl1 identifies
a cardiac progenitor population that proliferates prior to differentiation and
contributes a majority of cells to the heart. Dev. Cell 5, 877–889.
de Boer, B.A., Soufan, A.T., Hagoort, J., Mohun, T.J., van den Hoff, M.J., Hasman, A.,
Voorbraak, F.P., Moorman, A.F., Ruijter, J.M., 2011. The interactive presentation
of 3D information obtained from reconstructed datasets and 3D placement of
single histological sections with the 3D portable document format. Dev. 138,
159–167.
Evans, S.M., Yelon, D., Conlon, F.L., Kirby, M.L., 2010. Myocardial lineage develop-
ment. Circ. Res. 107, 1428–1444.
Grego-Bessa, J., Luna-Zurita, L., del Monte, G., Bolos, V., Melgar, P., Arandilla, A.,
Garratt, A.N., Zang, H., Mukouyama, Y.S., Chen, H., Shou, W., Ballestar, E.,
Esteller, M., Rojas, A., Perez-Pomares, J.M., de la Pompa, J.L., 2007. Notch
signaling is essential for ventricular chamber development. Dev. Cell 12,
415–429.
His, W., 1885. Menschlicher Embryonen. F.C.W. Vogel, Leipzig.
Hoffman, J.I., Kaplan, S., 2002. The incidence of congenital heart disease. J. Am.
Coll. Cardiol. 39, 1890–1900.
Howard, C.V., Reed, M.G., 1998. Unbiased Stereology. Three-Dimensional Mea-
surement in Microscopy. Bios Scientific Publishers, Liverpool, UK.
Kaufman, M.H., 1994. The Atlas of Mouse Development. Academic Press, London.
Kaufman, M.H., Navaratnam, V., 1981. Early differentiation of the heart in mouse
embryos. J. Anat. 133, 235–246.
Kelly, R.G., Brown, N.A., Buckingham, M.E., 2001. The arterial pole of the mouse
heart forms from Fgf10-expressing cells in pharyngeal mesoderm. Dev. Cell 1,
435–440.
Kelly, R.G., Buckingham, M.E., 2002. The anterior heart-forming field: voyage to
the arterial pole of the heart. Trends Genet. 18, 210–216.
Liu, Z., Yue, S., Chen, X., Kubin, T., Braun, T., 2010. Regulation of cardiomyocyte
polyploidy and multinucleation by CyclinG1. Circ. Res. 106, 1498–1506.
Meilhac, S.M., Esner, M., Kelly, R.G., Nicolas, J.F., Buckingham, M.E., 2004a. The
clonal origin of myocardial cells in different regions of the embryonic mouse
heart. Dev. Cell 6, 685–698.
Meilhac, S.M., Esner, M., Kerszberg, M., Moss, J.E., Buckingham, M.E., 2004b.
Oriented clonal cell growth in the developing mouse myocardium underlies
cardiac morphogenesis. J. Cell Biol. 164, 97–109.
Meilhac, S.M., Kelly, R.G., Rocancourt, D., Eloy-Trinquet, S., Nicolas, J.F., Bucking-
ham, M.E., 2003. A retrospective clonal analysis of the myocardium reveals
two phases of clonal growth in the developing mouse heart. Dev. 130,
3877–3889.
Mikawa, T., Borisov, A., Brown, A.M.C., Fischman, D.A., 1992. Clonal analysis of
cardiac morphogenesis in the chicken embryo using a replication-defective
retrovirus. I. Formation of the ventricular myocardium. Dev. Dyn. 193, 11–23.
Mjaatvedt, C.H., Nakaoka, T., Moreno-Rodriguez, R., Norris, R.A., Kern, M.J.,
Eisenberg, C.A., Turner, D., Markwald, R.R., 2001. The outflow tract of the
heart is recruited from a novel heart-forming field. Dev. Biol. 238, 97–109.
Mommersteeg, M.T., Dominguez, J.N., Wiese, C., Norden, J., de Gier-de, V.C., Burch,
J.B., Kispert, A., Brown, N.A., Moorman, A.F., Christoffels, V.M., 2010. The sinus
venosus progenitors separate and diversify from the first and second heart
fields early in development. Cardiovasc. Res. 87, 92–101.
Moorman, A.F.M., Christoffels, V.M., 2003. Cardiac chamber formation: develop-
ment, genes and evolution. Physiol. Rev. 83, 1223–1267.
 B.A. de Boer et al. / Developmental Biology 368 (2012) 203–213
Moreno-Rodriguez, R.A., Krug, E.L., Reyes, L., Villavicencio, L., Mjaatvedt, C.H.,
Markwald, R.R., 2006. Bidirectional fusion of the heart-forming fields in the
developing chick embryo. Dev. Dyn. 235, 191–202.
Navaratnam, V., Kaufman, M.H., Shepper, J.N., Barton, S., Guttridge, K.M., 1986.
Differentiation of the myocardial rudiment of mouse embryos: an ultrastruc-
tural study including freeze-fracture replication. J. Anat. 146, 65–85.
OConnor, W., Runquist, E.A., 2008. Error minimization algorithm for comparative
quantitative PCR analysis: Q-Anal. Anal. Biochem. 378, 96–98.
Patten, B.M., 1922. The formation of the cardiac loop in the chick. Am. J. Anat. 30,
373–397.
Romanoff, A.L., 1960. The Avian Embryo. Structural and Functional Development.
The Macmillan Company, New York.
Rosenthal, N., Harvey, R.P., 2010. Heart Development and Regenaration, 1.
Academic Press, Rome, Sidney.
Sedmera, D., Reckova, M., DeAlmeida, A., Coppen, S.R., Kubalak, S.W., Gourdie, R.G.,
Thompson, R.P., 2003. Spatiotemporal pattern of commitment to slowed
proliferation in the embryonic mouse heart indicates progressive differentia-
tion of the cardiac conduction system. Anat. Rec. 274A, 773–777.
Sissman, J., 1966. Cell multiplication rates during development of the primitive
cardiac tube in the chick embryo. Nature 210, 504–507.
Sizarov, A., Ya, J., de Boer, B.A., Lamers, W.H., Christoffels, V.M., Moorman, A.F.,
2011. Formation of the building plan of the human heart: morphogenesis,
growth, and differentiation. Circulation 123, 1125–1135.
Snarr, B.S., O’Neal, J.L., Chintalapudi, M.R., Wirrig, E.E., Phelps, A.L., Kubalak, S.W.,
Wessels, A., 2007. Isl1 expression at the venous pole identifies a novel role for
the second heart field in cardiac development. Circ. Res. 101, 971–974.
Soufan, A.T., van den Berg, G., Moerland, P.D., Massink, M.M.G., van den Hoff,
M.J.B., Moorman, A.F.M., Ruijter, J.M., 2007. Three-dimensional measurement
and visualization of morphogenesis applied to cardiac embryology. J. Microsc.
225, 269–274.
Soufan, A.T., van den Berg, G., Ruijter, J.M., de Boer, P.A.J., van den Hoff, M.J.B.,
Moorman, A.F.M., 2006. Regionalized sequence of myocardial cell growth and
proliferation characterizes early chamber formation. Circ. Res. 99, 545–552.
213
Stalsberg, H., de Haan, R.L., 1969. The precardiac areas and formation of the
tubular heart in the chick embryo. Dev. Biol. 19, 128–159.
Taber, L.A., 2006. Biophysical mechanisms of cardiac looping. Int. J. Dev. Biol. 50,
323–332.
Theiler, K., 1972. The House Mouse, Development and Normal Stages from
Fertilization to 4 Weeks of Age. Springer-Verlag, Berlin.
Thompson, R.P., Lindroth, J.R., Wong, Y.M.M., 1990. Regional differences in DNA-
synthetic activity in the preseptation myocardium of the chick. In: Clark, E.B.,
Takao, A. (Eds.), Developmental Cardiology: Morphogenesis and Function.
Futura Publishing Co., Mount Kisco, NY, pp. 219–234.
van den Berg, G., Abu-Issa, R., de Boer, B.A., Hutson, M.R., de Boer, P.A., Soufan, A.T.,
Ruijter, J.M., Kirby, M.L., van den Hoff, M.J., Moorman, A.F., 2009. A caudal
proliferating growth center contributes to both poles of the forming heart
tube. Circ. Res. 104, 179–188.
van den Berg, G., Moorman, A.F., 2011. Development of the pulmonary vein and
the systemic venous sinus: an interactive 3D overview. PLoS One 6, e22055.
Vincent, S.D., Buckingham, M.E., 2010. How to make a heart: the origin and
regulation of cardiac progenitor cells. Curr. Top. Dev. Biol. 90, 1–41.
Waldo, K.L., Kumiski, D.H., Wallis, K.T., Stadt, H.A., Hutson, M.R., Platt, D.H., Kirby,
M.L., 2001. Conotruncal myocardium arises from a secondary heart field.
Development 128, 3179–3188.
Weibel, E.R., 1979. Stereological Methods. Academic Press, London.
Wood, H.B., May, G., Healy, L., Enver, T., Morriss-Kay, G.M., 1997. CD34 expression
patterns during early mouse development are related to modes of blood vessel
formation and reveal additional sites of hematopoiesis. Blood 90, 2300–2311.
Zaffran, S., Kelly, R.G., Meilhac, S.M., Buckingham, M.E., Brown, N.A., 2004. Right
ventricular myocardium derives from the anterior heart field. Circ. Res. 95,
261–268.
Zhou, B., Ma, Q., Rajagopal, S., Wu, S.M., Domian, I., Rivera-Feliciano, J., Jiang, D., von,
G.A., Ikeda, S., Chien, K.R., Pu, W.T., 2008. Epicardial progenitors contribute to the
cardiomyocyte lineage in the developing heart. Nature 454, 109–113.