Postnatal isl1+ cardioblasts enter fully differentiated cardiomyocyte lineages.

Abstract

The purification, renewal and differentiation of native cardiac progenitors would form a mechanistic underpinning for unravelling steps for cardiac cell lineage formation, and their links to forms of congenital and adult cardiac diseases. Until now there has been little evidence for native cardiac precursor cells in the postnatal heart. Herein, we report the identification of isl1+ cardiac progenitors in postnatal rat, mouse and human myocardium. A cardiac mesenchymal feeder layer allows renewal of the isolated progenitor cells with maintenance of their capability to adopt a fully differentiated cardiomyocyte phenotype. Tamoxifen-inducible Cre/lox technology enables selective marking of this progenitor cell population including its progeny, at a defined time, and purification to relative homogeneity. Co-culture studies with neonatal myocytes indicate that isl1+ cells represent authentic, endogenous cardiac progenitors (cardioblasts) that display highly efficient conversion to a mature cardiac phenotype with stable expression of myocytic markers (25%) in the absence of cell fusion, intact Ca2+-cycling, and the generation of action potentials. The discovery of native cardioblasts represents a genetically based system to identify steps in cardiac cell lineage formation and maturation in development and disease.

Full text

Page 1

Insect Mol. Biol. 12, 433–440 (2003).
16. El-Kifl,A.H.MorphologyoftheadultTriboliumconfusumDuv.anditsdifferentiationfromTribolium
(stene) castaneum Herbst. Bull. Soc. Fouad 1er Entom. 37, 173–249 (1953).
17. deCelis, J.F.,Barrio,R.& Kafatos,F.C. Agenecomplexactingdownstreamofdpp in Drosophilawing
morphogenesis. Nature 381, 421–424 (1996).
18. Sanchez-Salazar, J. et al. The Tribolium decapentaplegic gene is similar in sequence, structure, and
expression to the Drosophila dpp gene. Dev. Genes Evol. 206, 237–246 (1996).
19. Grimm, S. & Pflugfelder, G. O. Control of the gene optomotor-blind in Drosophila wing development
by decapentaplegic and wingless. Science 271, 1601–1604 (1996).
20. Gomez-Skarmeta, J. L., Diez del Corral, R., de la Calle-Mustienes, E., Ferre-Marco, D. & Modolell, J.
araucan and caupolican, two members of the novel iroquois complex, encode homeoproteins that
control proneural and vein-forming genes. Cell 85, 95–105 (1996).
21. Wheeler,S.R.,Carrico,M.L.,Wilson,B.A.,Brown,S.J.&Skeath,J.B.Theexpressionandfunctionof
the achaete-scute genes in Tribolium castaneum reveals conservation and variation in neural pattern
formation and cell fate specification. Development 130, 4373–4381 (2003).
22. Campuzano,S. et al.Moleculargenetics of the achaete-scutegene complexof D.melanogaster.Cell40,
327–338 (1985).
23. Sturtevant, M. A., Biehs, B., Marin, E. & Bier, E. The spalt gene links the A/P compartment boundary
to a linear adult structure in the Drosophila wing. Development 124, 21–32 (1997).
24. de Celis, J. F. & Barrio, R. Function of the spalt/spalt-related gene complex in positioning the veins in
the Drosophila wing. Mech. Dev. 91, 31–41 (2000).
25. Abouheif, E. & Wray, G. A. Evolution of the gene network underlying wing polyphenism in ants.
Science 297, 249–252 (2002).
26. Beeman, R. W., Brown, S. J., Stuart, J. J. & Denell, R. E. in Molecular Insect Science (eds Hagedorn,
H. H., Hildebrannd, J. G., Kidwell, M. G. & Law, J. H.) 21–29 (Plenum, New York, 1990).
27. Beeman, R. W. & Stauth, D. M. Rapid cloning of insect transposon insertion junctions using
‘universal’ PCR. Insect Mol. Biol. 6, 83–88 (1997).
28. Quennedey,A.&Quennedey,B.Morphogenesisof thewinganlagenin themealworm beetleTenebrio
molitor during the last larval instar. Tissue Cell 22, 721–740 (1990).
Supplementary Information accompanies the paper on www.nature.com/nature.
Acknowledgements We thank G. Bucher, M. Weber and M. Klingler for the pu11 enhancer-trap
line, R.Whitefor theFP6.86antibody,DSHBfor the4D9antibodyandG.Pflugfelderforsharing
information about omb degenerate primers. We thank K. Leonard for maintaining beetle stocks,
T. Shippy for discussion and reading and S. Brown, R. Beeman, S. Haas and all the Manhattan
beetle/insect laboratory members for discussion and comments. Y.T. thanks A. Sato and
T. Yamaguchi for discussion. This work was supported by the international Human Frontier
Science Program Organization (Long-term Fellow) and the National Science Foundation.
Competing interests statement The authors declare that they have no competing financial
interests.
CorrespondenceandrequestsformaterialsshouldbeaddressedtoY.T.(tomoyasu@ksu.edu).The
sequences are available in GenBank under the following accession numbers. Tc-sal: AY600513;
Tc-iro: AY600514; Tc-omb: AY600516.
..............................................................
Postnatal isl11cardioblasts enter
fully differentiated cardiomyocyte
lineages
Karl-Ludwig Laugwitz1*, Alessandra Moretti1*, Jason Lam1*,
Peter Gruber3, Yinhong Chen1, Sarah Woodard1, Li-Zhu Lin1,
Chen-Leng Cai1, Min Min Lu4, Michael Reth5, Oleksandr Platoshyn2,
Jason X.-J. Yuan2, Sylvia Evans1& Kenneth R. Chien1
1Institute of Molecular Medicine and2Department of Medicine, University of
California, San Diego, School of Medicine, La Jolla, California 92093, USA
3Children’s Hospital of Philadelphia, Cardiac Center, Philadelphia, Pennsylvania
19104, USA
4Cardiovascular Division, Department of Medicine, University of Pennsylvania,
Philadelphia, Pennsylvania 19104, USA
5Max-Planck Institut fu ¨r Immunbiologie, Universita ¨t Freiburg, Biologie III,
Abteilung Molekulare Immunologie, Freiburg 79108, Germany
*These authors contributed equally to this work
.............................................................................................................................................................................
The purification, renewal and differentiation of native cardiac
progenitors would form a mechanistic underpinning for un-
ravelling steps for cardiac cell lineage formation, and their
links to forms of congenital and adult cardiac diseases1–3. Until
now there has been little evidence for native cardiac precursor
cells in the postnatal heart4. Herein, we report the identification
of isl11cardiac progenitors in postnatal rat, mouse and human
myocardium. Acardiac mesenchymal feeder layer allows renewal
of the isolated progenitor cells with maintenance of their capa-
bility to adopt a fully differentiated cardiomyocyte phenotype.
Tamoxifen-inducible Cre/lox technology enables selective mark-
ing of this progenitor cell population including its progeny, at a
defined time, and purification to relative homogeneity. Co-
culture studies with neonatal myocytes indicate that isl11cells
represent authentic, endogenous cardiac progenitors (cardio-
blasts)thatdisplayhighlyefficientconversiontoamaturecardiac
phenotype with stable expression of myocytic markers (25%) in
theabsenceofcellfusion,intactCa21-cycling,andthegeneration
of action potentials. The discovery of native cardioblasts rep-
resents a genetically based system to identify steps in cardiac
cell lineage formation and maturation in development and
disease.
In skeletal muscle, satellite cells are considered specialized
endogenous muscle precursors (that is, myoblasts) that are pre-
programmed to enter the skeletal muscle lineage5,6. Despite inten-
sive inquiry, there has been no clear evidence for a native cardiac
progenitor population with similar characteristics in the myocar-
dium. Various cell surface markers have been used to purify
precursor cells from heart muscle including c-kit and sca-1
(refs 7, 8), but the precise role of these cells in in vivo cardiogenesis
is unclear and the efficiency of conversion to fully differentiated
myocytes, in the absence of fusion, is relatively low9–11. Taking
advantage of a developmental lineage marker for undifferentiated
cardiogenic precursor cells as a requirement for a heart-specific
origin, we have identified in the postnatal heart a novel cardiac cell
type. The LIM-homeodomain transcription factor islet-1 (isl1)
marks a cell population that makes a substantial contribution to
the embryonic heart, comprising most cells in the right ventricle,
both atria, the outflow tract and also specific regions of the left
ventricle12.
Asubset of isl1þundifferentiated precursors remains embedded
in the embryonic mouse heart after its formation and their
number decreases progressively from embryonic day 12.5
(ED12.5) to ED18.5 (Fig. 1a–d). After birth, relatively few isl1þ
cardioblasts were still detectable, averaging 500 to 600 in the
myocardium of a 1–5-day-old rat (Fig. 1e, f). Their organ
distribution matched the defined contributions of isl1þembryonic
precursors (Fig. 1i), suggesting these cells are developmental
remnants of the fetal progenitor population. Clusters of isl1þ
cardioblasts were observed in both atria, whereas in the ventricles
they occurred mostly as single cells (Fig. 1e, f). The localization of
isl1þprogenitors in specific cardiac segments (atrial muscle wall,
intra-atrial septum, conus muscle, right ventricle) was conserved
among diverse species: mouse, rat and human (Fig. 1g, h;
Supplementary Table 1).
Using conditional genetic marking techniques in the mouse, we
performed Cre-recombinase-triggered cell lineage tracing experi-
ments to irreversibly mark isl1-expressing cells as well as their
differentiated progeny during embryonic development (Fig. 2).
Isl1-IRES-Cre mice13werecrossed into the conditional Crereporter
strain R26R14, in which Cre-mediated removal of a stop sequence
results in the ubiquitous expression of the lacZ gene under the
control of the endogenous Rosa26 promoter (Fig. 2a). In neonatal
mice bearing both alleles a high proportion of right ventricular
myocardium expressed b-galactosidase (b-gal) detected by
5-bromo-4-chloro-3-indolyl-b-D-galactoside (X-gal) staining
(Fig.2a).Around30–40%ofcardiacmyocytesisolatedfromdouble
heterozygous hearts stained positive for X-gal and displayed co-
expression ofb-galandsarcomeric a-actinin (Fig.2b), demonstrat-
ing that a significant proportion of myocytes originate from isl1þ
cardiac progenitors.
To achieve temporal and spatial control of Cre expression
letters to nature
NATURE|VOL 433|10 FEBRUARY 2005|www.nature.com/nature
647
© 2005 Nature Publishing Group

Page 2

required for tracing the fate of isl1þprogenitors resident in the
postnatal heart, we generated mice that harbour a knockin of the
tamoxifen-dependent Cre recombinase (isl1-mER-Cre-mER) into
the isl1 locus (Fig. 3). The presence of two mutated oestrogen-
receptors (mERs) results in the sequestration of Cre in the cyto-
plasm15. Tamoxifen induces a rapid nuclear translocation of the
mER-Cre-mER protein, which permits Cre-mediated recombina-
tion16,17. As a readout for Cre activity, we again used the reporter
strain R26R14. Thus, application of tamoxifen to double hetero-
zygous isl1-mER-Cre-mER/R26R mice leads to b-gal expression in
cells expressing isl1 at the time of exposure (Fig. 3a). Experiments
with isl1-mER-Cre-mER/R26R mice were performed to assess the
specificity of cell labelling. b-galþcells are found only within the
looping heart in double heterozygous animals after tamoxifen
injection at ED7.5 and X-gal staining at ED9.5 (Fig. 3b). Double
immunofluorescent analysis confirmed the co-expression of isl1
and Cre in undifferentiated cardiac progenitors and demonstrated
that nuclear or cytosolic Cre localization was dependent on tamox-
ifenexposure(Fig.3c).AtED12,sectionsofrightventricleandatrial
tissue, heart regions that have been shown to originate from isl1þ
precursors12, exhibited expression of a-myosin heavy chain but
were negative for isl1 and Cre (Fig. 3d). Thus, tamoxifen injection
leads to a highly specific marking of isl1-expressing progenitorsand
Cre protein is not detectable ,54h after the promoter has been
inactivated. After birth, b-galþcells can be detected with highest
frequency in the conus muscle, the right atrium and the right
ventricle by X-gal stain in isl1-mER-Cre-mER/R26R double
heterozygous neonatal hearts after tamoxifen injection at ED17
Figure 1 Isl1þprogenitors in the late embryonic and postnatal heart. a–d, Isl1þ
progenitors in mouse sections at ED12.5 (a, atrial septum; c, ventricular tissue) and 18.5
(b,freeatrialwall;d,ventriculararterialregion).Scalebar,20mm(a,b,d),50mm(c).EV,
eustachian valve; RA, right atrium; Ao, aorta; LMCA, left main coronary artery; Pa,
pulmonary artery. e, f, Cluster of isl1þcardiac precursors embedded in right atrial tissue
and in the right ventricle of a postnatal day 1 rat heart. Scale bars, 10mm (e), 15mm (f).
g, h, Isl1þprogenitors in human right atrial tissue from an 8-day-old patient and intra-
atrial septum from a 2-day-old patient. Scale bars, 75mm (g), 150mm (h).
i, Quantification and localization of isl1þcells in postnatal day 1 rat hearts. Mean values
^s.e.m. from three hearts. Arrows designate isl1þcells in different species and insets
represent a magnification of the areas of interest.
letters to nature
NATURE|VOL 433|10 FEBRUARY 2005|www.nature.com/nature
648
© 2005 Nature Publishing Group

Page 3

(Fig. 3e), and represent a rare subset of cells in the adult right
atrium and outflow tract (Supplementary Fig. 1).
Followingcellisolationfrompostnatalisl1-mER-Cre-mER/R26R
mouseheartsandreporter geneinduction invitro,mostb-galþcells
were found in the cardiac mesenchymal cell (CMC) fraction
(Fig. 4a). Mesenchymal cells isolated from postnatal rat hearts
also contain isl1þprogenitors (Supplementary Fig. 2). The CMC
culture environment maintains isl1 expression in the progenitor
population and promotes their expansion in culture without
differentiation (Fig. 4b and Supplementary Fig. 2).
b-galþcells can be purified from the mesenchymal cells using
fluorescence-activated cell sorting (FACS) after labelling with a
fluorogenic, lipophilic b-gal substrate (C12FDG) (Fig. 4c, d). Isl1þ
progenitors were isolated as a distinct population presenting seven
to ten times higher C12FDG fluorescence than background at a
purity of 90–95%. They corresponded to 0.5% of the mesenchymal
cellfractionwhen FACS-sorted during an earlystage in culture(4–6
days) and reached ,8% after 15–18 days culture (Fig. 4c, d). As
phenotypic characteristics of an undifferentiated state, b-galþpre-
cursors expressed isl1 and early specification markers for cardiac
mesoderm,Nkx2.5andGATA-4,whilelackingtranscriptsofmature
myocytes (Fig. 4e).
FACSanalysisoftheFDGþcellfraction,expandedinvitroover10
days, revealed negative staining for sca-1 and no correlation
between b-gal expression and the efflux of the DNA-binding dye
Hoechst 33342, demonstrating that isl1þcardioblasts are distinct
from cardiac side population cells (Fig. 4f–h and Supplementary
Table 2)8,18. In addition, isl1þcells did not express c-kit.
After exposure to 4-hydroxytamoxifen (4-OH-TM) in vitro, a
rare subset of b-galþcells coexpressing cardiac troponin T was
observed at day 4 (n ¼ 51 ^ 12), day 7 (n ¼ 81 ^ 29) and day 10
(n ¼ 157 ^ 10) in the myocyte fraction (Fig. 4i, j), resulting in
,75% of the total b-galþcells. The fact that isl1 is repressed as
soon as the cells acquire a differentiated phenotype12and the lack of
b-gal expression in the absence of 4-OH-TM (Fig. 3b) strongly
suggest that isl1þcardioblasts can spontaneously acquire myocytic
characteristics when dispersed and cultured in the presence of
differentiated myocytes.
To assess the potential of b-galþprogenitors to adopt a
myocytic phenotype, we performed co-culture experiments of
FACS-sorted precursors with neonatal cardiac myocytes. After 3
days in co-culture, b-galþcells were found within the myocytic
syncytium and expressed cardiac specific proteins, that is, a-
sarcomeric actinin or troponin T in a partially or completely
organized pattern (Fig. 4k, l). Moreover, a subset of b-galþ
cardioblasts showed contractile activity and intact electromecha-
nical coupling to neighbouring cells, as documented by expression
of the gap-junctional protein connexin 43 (data not shown). The
differentiation capacity of b-galþprecursors did not significantly
differ between cells maintained in culture for only a few days
(early stage) and cells expanded for 15–18 days (late stage)
(Fig. 4l).
We addressed the role of cell fusion in the progenitor-to-cardiac
myocyte transition, by performing co-cultureexperiments inwhich
neonatal myocytes were pre-fixed with paraformaldehyde to stabil-
ize cellular components and render the cells refractory to cell
fusion19. Paraformaldehyde treatment disrupted the exclusion of
propidium iodide by the myocytes, indicating that these cells were
non-viable (Fig. 4m). FACS-sorted b-galþand b-gal2cells were
labelledwithsemiconductorred-fluorescentnanocrystals(QD655)20
and co-cultured with pre-fixed myocytes in the presence of
conditioned medium from viable myocytes. Under these con-
ditions, within 5 days around 30% of the QD655-marked b-galþ
progenitors started to express a-actinin (Fig. 4m), indicating
that cardioblasts enter the myocytic program independently
from cell fusion. In contrast, all b-gal2cells stained negative for



Figure 2 Genetic marking of isl1þprogenitors and myocytic cell fate. a, Mice carry one
isl1-IRES-CrealleleandoneR26Rreportergene.Creexpressioncatalysesexcisionofthe
stop cassette, resulting in selective lacZ expression and genetic marking of isl1-
expressing cells and their differentiated progeny. RTV, retroviral integration sequence.
Shownareb-galþcardiomyocytesintherightventricle(middlepanel;insetdemonstrates
a-actinin expression; scale bar, 180mm) and after cell isolation (right panel; scale bar,
20mm) from double heterozygoushearts of 4-month- or 1-day-old animals, respectively.
b, Immunocytochemistryforb-galand sarcomerica-actinin inisolatedcardiacmyocytes
from animals carrying both alleles. Scale bar, 15mm.
letters to nature
NATURE|VOL 433|10 FEBRUARY 2005|www.nature.com/nature
649
© 2005 Nature Publishing Group

Page 4

a-actinin, as well as b-galþprecursors when cultured inthe absence
of pre-fixed myocytes, demonstrating that differentiation is cell-
specific and requires secreted and membrane-bound factors
(Fig. 4m).
To characterize whether the b-galþcardioblasts could adopt a
fully differentiated cardiac phenotype, C12FDG-labelled cells in co-
culture were analysed by real-time intracellular calcium [Ca2þ]i
imaging. Differentiated C12FDGþprogenitors showed periodic
[Ca2þ]ioscillations similar to and synchronized with those in
adjacentmyocytes (Fig. 5a, b;Supplementary Video). Isoproterenol
Figure 3 Specificity of recombination and distribution of b-galþcells in isl1-mER-Cre-
mER/R26R hearts. a, Tamoxifen injection of isl1-mER-Cre-mER/R26R double
heterozygousmiceoradministrationof4-OH-TMincultureresultsinheritableexpression
of lacZ. Labelling with C12FDG allows FACS purification of isl1þcells. b, X-gal stain in
whole-mount double heterozygous embryos with or without tamoxifen injection into
pregnant mothers at ED7.5. c, Cre localization in isl1þcells from isl1-mER-Cre-mER/
R26R embryos is cytosolic without tamoxifen and nuclear with tamoxifen injection. Scale
bar, 25mm. d, Downregulation of isl1 and Cre in sequential sections of right ventricular
and atrial tissue (left panels; scale bar, 80mm) but not of dorsal mesocardium (right
panels;scalebar,30mm).e,Sectionsfromheartsof1-day-olddoubleheterozygousmice
injected with tamoxifen at ED17, after X-gal stain and haematoxylin counterstain. LMCA,
left main coronary artery; Ao, aorta; RA, right atrium; RV, right ventricle. Scale bar,
150mm.Arrowsdesignateb-galþcells andinsetsrepresentamagnificationof theareas
of interest.
letters to nature
NATURE|VOL 433|10 FEBRUARY 2005|www.nature.com/nature
650
© 2005 Nature Publishing Group

Page 5

stimulationinduceda30%increaseoftheCa2þtransientamplitude
indifferentiatedb-galþcells,suggestingtheexistenceofafunctional
b-adrenergic signalling pathway21(Fig. 5c). The [Ca2þ]ioscillations
responded to increasing frequencies of external field stimulation
(Fig. 5d). Of the C12FDG-labelled cardioblasts in the co-culture
2.3%showedactiveintracellularCa2þtransients.Functionalmatur-
ity was confirmed by analysing action potential characteristics in
QD655-marked progenitors and neighbouring myocytes. Beating
differentiated cardioblasts displayed continuous electrical activity
or action potentials characterized by upstroke velocity (dV/dtmax),
amplitude (APA), duration (APD) and maximum diastolic
potential (MDP) comparable to those of neonatal cardiac myocytes
(Fig. 5e, f). Repeated impalements of individual QD655-
marked differentiated progenitors revealed reproducible action
potentials with distinct morphologies, nodal-like and atrial-like,
supportingtheconceptthatthesecellsareabletodifferentiateintoa
distinct subset of cardiac lineages, including conduction system
myocytes22.
We demonstrate that the mammalian heart harbours an
organ-specific progenitor cell that can be localized, purified,
expanded and differentiated into mature cardiac myocytes
(Fig. 5g). It will be interesting to investigate whether isl1þ





Figure 4 Amplification, characterization and myocytic differentiation of isl1þ
cardioblasts in vitro. a, b, Mesenchymal cell fractions from isl1-mER-Cre-mER/R26R
hearts at 10 days in culture after 4-OH-TM treatment. Arrows point to b-galþ
cardioblasts detected by X-gal stain (a) and isl1 expression (b). Scale bars, 15mm.
c, d, Histograms of FACS-sorted b-galþ-C12FDG-labelled cells in cardiac
mesenchymal fractions after 5 days (c) and 14 days (d) in culture. e, RT–PCR
analysis for myocytic and progenitor markers in FACS-sorted progenitors (P) and
neonatal myocytes (M). f–h, Flow cytometry profile of Hoechst 33342 dye efflux in
cardiac mesenchymal cell fractions from double heterozygous animals after C12FDG
and sca-1 labelling. i, j, Neonatal myocytes isolated from isl1-mER-Cre-mER/R26R
mice at 4 days in culture. Arrows designate b-galþcells coexpressing troponin T and
inset shows the same cells before immunocytochemistry. Scale bar, 15mm (i).
Frequency of b-galþmyocytes and non-myocytes over time (j, mean value ^s.e.m.,
n ¼ 3). Exposure to 1mM 4-OH-TM was performed at the first day in culture.
k, Images of differentiated b-galþprogenitors in co-culture with wild-type neonatal
myocytes. Arrows mark progenitor-derived b-galþmyocytes. Scale bar, 15mm.
l, Quantification of differentiation events over time in co-culture of b-galþprogenitors
FACS-sorted at day 10–14 (left panel) and comparison at 5 days co-culture between
b-galþprogenitors expanded for 5 days (early) or 14 days (late) in culture (right
panel). Mean values ^s.e.m. from six experiments (n ¼ 1,200 cells per group).
m, Cell fusion independent cardioblast-myocyte conversion. Propidium-iodide stain in
pre-fixed myocytes (scale bar, 25mm). Arrows point to QD655-labelled precursors
expressing a-actinin in b-galþ, but not in b-gal2cells during co-culture with pre-
fixed myocytes. Scale bar, 15mm. The diagram represents a quantification of fusion-
independent myocytic transition. Grey columns indicate the percentage of b-galþ
cells expressing a-actinin in absence of pre-fixed myocytes. Mean values ^s.e.m.
(n ¼ 3).
letters to nature
NATURE|VOL 433|10 FEBRUARY 2005|www.nature.com/nature
651
© 2005 Nature Publishing Group

Page 6

precursors can contribute to the cardiac lineage in the absence of
fusion after cellular transplantation in vivo, and to analyse
functionally on a single-cell level the differentiated progeny23.
Cardioblasts may represent an in vitro system to study signalling
pathways for cardiac cell lineage differentiation, which will
ultimately give new insights into key steps that govern entry
into specific contractile and conduction system lineages; recent
work has now documented defects in later stages of ventricular
lineage specification and forms of human congenital heart
disease24. Furthermore, this study suggests the feasibility of
isolating isl1þcardiac progenitors from mammalian embryonic
stem cell systems during cardiogenesis.
A
Methods
Mice
Isl1-IRES-Cre were provided by T. M. Jessel and have been previously described13. An
IRES-Cre SV40pA and a pgk-neomycin cassette were inserted into the exon encoding the
second LIM homeodomain of isl1. The conditional Cre reporter mouse line R26R was
generatedbyP.Soriano14.Theisl1-mER-Cre-mERtargetingconstructwasgeneratedbyan
in-frame insertion of a mER-Cre-mER SV40pA cassette15along with a neo-selectable
marker flanked by FRTsites into exon 1 of the genomic isl1 locus. The generation of
isl1-mER-Cre-mER knockin mice will be described in detail elsewhere. Isl1-IRES-Cre/
R26R and isl1-mER-Cre-mER/R26R double heterozygous mice were generated by
crossing single heterozygous mice. Mice are in a mixed 129 £ C57Bl/6 background. The
isl1-mER-Cre-mER line showed exclusively a tamoxifen-dependent expression of Cre.
Tamoxifen (Sigma) was dissolved in corn oil and injected twice intraperitoneally in the
pregnant mothers at a dose of 75mgg21body weight at ED17 and ED18, unless specified
otherwise25. Animal care and all procedures were approved by the Institutional Animal
Care Committee at UCSD.
Isolation of mouse cardiac progenitors and cell culture conditions
For cell isolation, we used 40–60 1–5-day-old pups that were double heterozygous for
isl1-mER-Cre-mER and R26R alleles. Hearts were cut into four pieces, washed
repetitively in ice-cold Hank’s balanced salt solution without Ca2þ(HBSS) and
predigested overnight in 0.5mgml21trypsin-HBSS solution at 48C under constant
shaking, to remove blood and dead cells. Cardiac cells were obtained by four rounds of
10min digestions with 240Uml21collagenase type II (Worthington) in HBSS at 378C.
The mesenchymal cell fraction containing most b-galþprogenitor cells was separated
from myocytes and endothelial cells by two rounds of 1h differential plating on plastic in
DMEM/M199 (4:1 ratio) medium containing 10% horse serum and 5% fetal bovine
serum at 378C in the presence of 5% CO2. Under these conditions preferentially
mesenchymal and b-galþprecursor cells attach on plastic. After rigorous washing to
remove non-adherent cells, CMC and cardioblasts were cultured for 48h in DMEM
containing penicillin (100Uml21), streptomycin (100mgml21), HEPES (25mM),
Figure 5 Real time [Ca2þ]itransients and action potentials in FACS-sorted b-galþ
precursors after myocytic differentiation in co-culture. a, b, Identification of b-galþcells
by C12FDG labelling, before cell loading with the Ca2þindicator fluo-4. Pseudo-colour
images show minimal (Ca2þ
circles indicate the Ca2þmeasuring areas as outlined in the fluo-4 intensity traces in b.
Eighteen measured cells: n ¼ 11 differentiated progenitors, n ¼ 7 neonatal myocytes.
c, Increase in [Ca2þ]itransient amplitude of a b-galþcell under isoproterenol (1027M)
stimulation. d, Calcium transients of differentiated C12FDGþprecursor in response to
electrical pacing at different frequencies after cell loading with the Ca2þindicator fura-2.
min) and maximal (Ca2þ
max) fluo-4 fluorescence intensity and
Six measured C12FDGþcells. e, QD655-marked differentiated progenitors stained for
a-actinin (left panel; scale bar, 20mm) and impaled by an intracellular electrode to
measure electrical activity (right panel). Inset represents the QD655
potentials in a differentiated progenitor and a neonatal cardiomyocyte. Summarized data
(means ^ s.e.m.) showing dV/dtmax(the maximum rate of rise of the action potential
stroke), APA, APD90(action potential duration at 90% of repolarization) and MDP (the
maximum diastolic potential). We analysed 160 action potentials from four different cells
for each cell type. g, Working model of isl1þcardiac progenitors for self-renewal and
cardiomyocytic differentiation.
þcell. f, Action
letters to nature
NATURE|VOL 433|10 FEBRUARY 2005|www.nature.com/nature
652
© 2005 Nature Publishing Group

Page 7

glutamine (2mM), 10% newborn calf serum and 5% fetal bovine serum (FBS). Culture
medium was exchanged to DMEM/F12 containing B27 supplement, 2% FBS and
10ngml21EGF on the second day in culture when the CMC reached confluency, to
stimulate isl1 progenitor growth. 4-OH-TM (stock solution 1mM in ethanol; Sigma) was
applied in culture one day after cell plating at a concentration of 1mM and maintained for
48h.
Methods regarding flow cytometry analysis, co-culture experiments,
immunocytochemistry and lacZ staining, polymerase chain reaction with reverse
transcription (RT–PCR), detection of [Ca2þ]itransients and electrophysiological action
potential recordings are described in the Supplementary Methods section.
Received 12 October; accepted 22 November 2004; doi:10.1038/nature03215.
1. Chien, K. R. Stem cells: lost in translation. Nature 428, 607–608 (2004).
2. Epstein, J. A. Developing models of DiGeorge syndrome. Trends Genet. 17, S13–S17 (2001).
3. Chien, K. R. & Olson, E. C. Converging pathways and principles in heart development and disease:
CV@CSH. Cell 110, 153–162 (2002).
4. Rosenthal, N. Prometheus’s vulture and the stem cell promise. N. Engl. J. Med. 346, 267–274
(2003).
5. Bischoff, R. in Myogenesis (eds Engel, A. G. & Franszini-Armstrong, C.) 97–118 (McGraw-Hill, New
York, 1994).
6. Seale,P.&Rudnicki,M.A.Anewlookattheorigin,function,and“stem-cell”statusofmusclesatellite
cells. Dev. Biol. 218, 115–124 (2000).
7. Beltrami, A. P. et al. Adult cardiac stem cells are multipotent and support myocardial regeneration.
Cell 114, 763–776 (2003).
8. Oh, H. et al. Cardiac progenitor cells from adult myocardium: Homing, differentiation, and fusion
after infarction. Proc. Natl Acad. Sci. USA 100, 12313–12318 (2003).
9. Balsam, L. B. et al. Haematopoietic stem cells adopt mature haematopoietic fates in ischemic
myocardium. Nature 428, 668–675 (2004).
10. Murry, C. E. et al. Hematopoietic stem cells do not transdifferentiate into cardiac myocytes in
myocardial infarcts. Nature 428, 664–668 (2004).
11. Nygren, J. M. et al. Bone marrow-derived hematopoietic cells generate cardiomyocytes at a low-
frequency through cell fusion, but not transdifferentiation. Nature Med. 10, 487–493 (2004).
12. Cai,C.-L.et al. Isl1identifies acardiacprogenitor population that proliferatesprior to differentiation
and contributes a majority of cells to the heart. Dev. Cell 5, 877–889 (2003).
13. Srinivas, S. et al. Cre reporter strains produced by targeted insertion of EYFP and ECFP into the
ROSA26 locus. BMC Dev. Biol. 1, 4 (2001).
14. Soriano,P.GeneralizedlacZexpressionwiththeROSA26Crereporterstrain.NatureGenet.21,70–71
(1999).
15. Verrou, C., Zhang, Y., Zurn, C., Schamel, W. W. & Reth, M. Comparison of the tamoxifen regulated
chimeric Cre recombinases MerCreMer and CreMer. Biol. Chem. 380, 1435–1438 (1999).
16. Danielian, P. S., Muccino, D., Rowitch, D. H., Michael, S. K. & McMahon, A. P. Modification of gene
activity in mouse embryos in utero by tamoxifen-inducible form of Cre recombinase. Curr. Biol. 8,
1323–1326 (1998).
17. Schwenk,F.,Kuhn,R.,Angrand,P.O.,Rajewsky,K.&Stewart,A.F.Temporallyandspatiallyregulated
somatic mutagenesis in mice. Nucleic Acids Res. 26, 1427–1432 (1998).
18. Martin, C. M. et al. Persistent expression of the ATP-binding cassette transporter, Abcg2, identifies
cardiac SP cells in the developing and adult heart. Dev. Biol. 265, 262–275 (2004).
19. Wurmser, A. E. et al. Cell fusion-independent differentiation of neural stem cells to the endothelial
lineage. Nature 430, 350–356 (2004).
20. Voura, E. B., Jaiswal, J. K., Mattoussi, H. & Simon, S. M. Tracking metastatic tumor cell extravasation
with quantum dot nanocrystals and fluorescence emission-scanning microscopy. Nature Med. 9,
993–998 (2004).
21. Maltsev,V.A.,Ji,G.J.,Wobus,A.M.,Fleischmann,B.K.&Hescheler,J.Establishmentofb-adrenergic
modulation of L-type Ca2þcurrent in the early stages of cardiomyocyte development. Circ. Res. 84,
136–145 (1999).
22. He, J.-Q., Ma, Y., Lee, Y., Thomson, J. A. & Kamp, T. J. Human embryonic stem cells develop into
multiple types of cardiac myocytes. Circ. Res. 93, 32–39 (2003).
23. D’Amour, K.A. & Gage, F. H. Aresomaticstem cellspluripotent or lineage-restricted? Nature Med.8,
213–214 (2002).
24. Pashmforoush, M. et al. Nkx2.5 pathways and congenital heart disease: loss of ventricular myocyte
lineagespecificationleadstoprogressivecardiomyopathyandcompleteheartblock.Cell117,373–386
(2004).
25. Hayashi,S.&McMahon,A.P.Efficientrecombinationindiversetissuesbytamoxifen-inducibleform
of Cre: a tool for temporally regulated gene activation/inactivation in the mouse. Dev. Biol. 244,
305–318 (2002).
Supplementary Information accompanies the paper on www.nature.com/nature.
Acknowledgements We thank W. Giles for discussions and comments on the real-time Ca2þ
measurements and electrophysiology, the National Center for Microscopy and Imaging Research
at UCSD for the use of the high-speed multi-photon laser-scanning microscope (Bio-Rad
RTS2000)andcustomplug-insfortheImageJsoftware(H.Hakozaki,S.ChowandM.Ellisman),
the UCSD Cancer Center Digital Imaging Shared Resource for deconvolution microscopy
(S. McMullen and J. Feramisco). We would like especially to acknowledge D. Young for his
expertise and help in the FACS sorting analysis (UCSD Cancer Center Flow Cytometry
Laboratories).ThisworkwassupportedbytheNIHandtheJeanLeDucqFoundation.K.-L.L.isa
Heisenberg Scholar of the German Research Foundation.
Competing interests statement The authors declare that they have no competing financial
interests.
Correspondence and requests for materials should be addressed to K.R.C. (kchien@partners.org)
or to S.E. (syevans@ucsd.edu).
..............................................................
The BRCA2 homologue Brh2
nucleates RAD51 filament formation
at a dsDNA–ssDNA junction
Haijuan Yang1, Qiubai Li1, Jie Fan1, William K. Holloman2
& Nikola P. Pavletich1,3
1StructuralBiologyProgram,MemorialSloan-KetteringCancerCenter,NewYork,
New York 10021, USA
2Department of Microbiology and Immunology, Hearst Microbiology Research
Center,CornellUniversityWeillMedicalCollege,NewYork,NewYork10021,USA
3Howard Hughes Medical Institute, Memorial Sloan-Kettering Cancer Center,
New York, New York 10021, USA
.............................................................................................................................................................................
The BRCA2 tumour suppressor1is essential for the error-free
repair of double-strand breaks (DSBs) in DNA by homologous
recombination2,3. This is mediated by RAD51, which forms a
nucleoprotein filament with the 3
DNA (ssDNA) of the resected DSB, searches for a homologous
donor sequence, and catalyses strand exchange with the donor
DNA4. The 3,418-amino-acid BRCA2 contains eight ,30-amino-
acid BRC repeats that bind RAD51 (refs 5, 6) and a ,700-amino-
acidDBDdomainthatbindsssDNA7.TheisolatedBRCandDBD
domains have the opposing effects of inhibiting8,9and stimulat-
ing recombination7, respectively, and the role of BRCA2 in repair
has been unclear. Here we show that a full-length BRCA2
homologue (Brh2) stimulates Rad51-mediated recombination
at substoichiometric concentrations relative to Rad51. Brh2
recruits Rad51 to DNA and facilitates the nucleation of the
filament, which is then elongated by the pool of free Rad51.
Brh2 acts preferentially at a junction between double-stranded
DNA (dsDNA) and ssDNA, with strict specificity for the 3
overhang polarity of a resected DSB. These results establish a
BRCA2 function in RAD51-mediated DSB repair and explain the
loss of this repair capacity in BRCA2-associated cancers.
Isolated BRC repeats can block the assembly of the RAD51
nucleoprotein filament in vitro when present in excess over
RAD51 (refs 9, 10), and can also interfere with DSB repair in a
dominant-negative fashion in vivo8. These effects have been
suggested to reflect a negative control mechanism of BRCA2 over
RAD51 (ref. 11). In contrast, the findings that the DBD domain
(BRCA2DBD) binds ssDNA and has a putative helix–turn–helix
(HTH) dsDNA-binding motif7, coupled with the demonstration
that a stoichiometric amount of the BRCA2DBDcan stimulate
RAD51-mediated strand exchange in vitro7, indicated that
BRCA2 might facilitate RAD51-mediated recombination and
might have a role at the dsDNA–ssDNA (ds–ss) junction of the
resected DSB7.
One model for reconciling these opposing biochemical activities
isthatBRCA2helpstonucleatethefilamentbyrecruitingRAD51to
the DNA, and possibly to the ds–ss junction. Once there, RAD51
coulddissociatefromBRCA2andpartitionontossDNA,nucleating
filament formation by the pool of free RAD51. Alternatively,
RAD51 could remain bound to the BRC repeat and still nucleate
the filament. This is a possibility because in the RAD51-BRC crystal
structure10the BRC repeat blocks only one of the two polymeriza-
tion ends of RAD51, and the second end is in principle available to
bind an additional RAD51 molecule and elongate the filament in a
unidirectional manner (Supplementary Fig. 1).
To test thismodel, we purified thefull-length BRCA2homologue
Brh2fromthefungusUstilagomaydis.LikeBRCA2,Brh2isrequired
for recombination, DSB repair and genomic stability12. The 1,075-
amino-acid Brh2 has homology to the DBD of human BRCA2
(ref. 13) and contains one consensus BRC repeat12. The number of
0overhanging single-stranded
0
letters to nature
NATURE|VOL 433|10 FEBRUARY 2005|www.nature.com/nature
653
© 2005 Nature Publishing Group