Supplemental Material can be found at:
http://www.jbc.org/cgi/content/full/M213102200/DC1
THE JOURNAL OF BIOLOGICAL CHEMISTRY
© 2003 by The American Society for Biochemistry and Molecular Biology, Inc.
Vol. 278, No. 18, Issue of May 2, pp. 15927–15934, 2003
Printed in U.S.A.
Conditional Expression of SV40 T-antigen in Mouse Cardiomyocytes
Facilitates an Inducible Switch from Proliferation to
S
Differentiation*□
Received for publication, December 23, 2002, and in revised form, January 27, 2003
Published, JBC Papers in Press, February 17, 2003, DOI 10.1074/jbc.M213102200
Igor I. Rybkin, David W. Markham‡, Zhen Yan‡, Rhonda Bassel-Duby, R. Sanders Williams‡,
and Eric N. Olson§
From the Departments of Molecular Biology and Internal Medicine, University of Texas Southwestern Medical Center,
Dallas, Texas 75390 and ‡Duke University Medical School, Durham, North Carolina 27710
Studies of cardiac development have been hampered by the
lack of immortalized cell lines capable of proliferation and
* This work was supported by grants from the National Institutes of
Health, Bridging Project for the Alliance for Cellular Signaling
(NIGMS, National Institutes of Health), the Donald W. Reynolds Foundation, and the Texas Advanced Technology Program. The costs of
publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked “advertisement”
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
□
S The on-line version of this article (available at http://www.jbc.
org) contains Tables S1 and S2.
§ To whom correspondence should be addressed: University of Texas
Southwestern Medical Center, 5323 Harry Hines Blvd., Dallas, TX
75390-9148. Tel.: 214-648-1187; Fax: 214-648-1196; E-mail: Eric.
Olson@utsouthwestern.edu.
This paper is available on line at http://www.jbc.org
differentiation. The major obstacle to deriving such cell lines is
the phenomenon of permanent withdrawal of mammalian cardiac muscle cells from the cell cycle shortly after birth (1, 2).
Although a small fraction of adult mammalian cardiomyocytes
can re-enter the cell cycle and replicate DNA upon physiological or pathological stimulation in vivo, there is no significant
contribution to cardiac repair by hyperplasia of cardiac cells
following damage (i.e. myocardial infarction) (3). Neonatal or
embryonic cardiac muscle cells will progress through limited
rounds of cell division in cell culture, but they ultimately withdraw permanently from the cell cycle, and cultured adult cardiomyocytes will not divide and eventually die.
There have been several attempts to establish cardiac muscle cell lines. Cardiomyocytes have been derived from embryonic stem cells (4), P19 cells (5), and hematopoietic stem cells
(6). During the course of differentiation these cells differentiated into different cell types and therefore did not represent a
homogeneous cell population. To overcome this problem the
cardiomyocytes derived from embryonic stem cells were subjected to enrichment by using a selectable marker driven by a
cardiac-specific promoter (7, 8). Nonetheless, all of these approaches failed to provide a sustainable homogeneous cardiac
cell line.
Other cell lines, QCE-6 and H9c2, derived from precardiac
mesoderm of quail (9) or embryonic rat myocardium (10), respectively, provided useful models for studying early cardiac
fate specification and cardiac ion channel function. However,
upon induction of differentiation, the QCE-6 cell line transformed into a mixed population of cells and failed to differentiate into mature cardiomyocytes. The H9c2 cell line was shown
to express a number of muscle-specific channels but displayed
few muscle structural proteins.
Many investigators have used ectopic expression of oncogenes to transform embryonic or adult cardiac muscle cells into
immortalized cell lines. Rat embryonic ventricular cardiomyocytes infected with recombinant retrovirus expressing v-myc
and v-H-ras resulted in cells exhibiting some myocyte characteristics; however, these cells do not form sarcomeres or contract (11).
Promising results have come from studies utilizing simian
virus 40 (SV40) T-antigen (TAg)1 as a transforming factor in
murine and human primary cells (12). Cell lines derived by
1
The abbreviations used are: TAg, simian virus 40 large T-antigen;
BrdUrd, 5-bromo-2⬘-deoxyuridine; Ad, adenoviruses; Ad-Cre, recombinant adenovirus expressing Cre recombinase; Ad-lacZ, recombinant
adenovirus expressing ␤-galactosidase; BMP, bone morphogenetic protein; FBS, fetal bovine serum; PE, phenylephrine; NE, norepinephrine;
TGF-␤1, transforming growth factor ␤1; IGF-I, insulin-like growth factor I; RT, reverse transcriptase; CaMKI, calcium/calmodulin-dependent
protein kinase I; HDAC, histone deacetylases.
15927
Downloaded from www.jbc.org at VIVA, Univ of Virginia on May 19, 2009
Studies of cardiac muscle gene expression and signaling have been hampered by the lack of immortalized
cardiomyocyte cell lines capable of proliferation and
irreversible withdrawal from the cell cycle. With the
goal of creating such cell lines, we generated transgenic
mice using cardiac-specific cis-regulatory elements
from the mouse Nkx2.5 gene to drive the expression of a
simian virus 40 large T-antigen (TAg) gene flanked by
sites for recombination by Cre recombinase. These
transgenic mice developed tumors within the ventricular myocardium. Cells isolated from these tumors expressed cardiac markers and proliferated rapidly during serial passage in culture, without apparent
senescence. However, they were unable to exit the cell
cycle and failed to exhibit morphological features of
terminal differentiation. Introduction of Cre recombinase to these cardiac cell lines by adenoviral delivery
resulted in the elimination of TAg expression, accompanied by rapid cessation of cell division, and increase in
cell size without an apparent induction of cellular differentiation. Incubation of cells lacking TAg in serumdeficient media with various pharmacological agents
(norepinephrine, phenylephrine, or bone morphogenetic
protein-2/4) or constitutively active calcium/calmodulindependent protein kinase I and/or calcineurin led to the
formation of sarcomeres and up-regulation of cardiac
genes involved in excitation-contraction coupling. The
combination of TAg expression under the control of an
early cardiac promoter and Cre-mediated recombination
allowed us to derive an immortal cell line from the ventricular myocardium that could be controllably withdrawn from the cell cycle. The conditional expression of
TAg in this manner permits propagation and regulated
growth termination of cell types that are otherwise unable to be maintained in cell culture and may have applications for cardiac repair technologies.
15928
Generating Conditional Cardiomyocyte Cell Lines
EXPERIMENTAL PROCEDURES
Generation of Transgenic Mice—DNA constructs, Nk-TAg (Fig. 1A)
and NkL-TAg (Fig. 4A), were used to generate transgenic mice expressing SV40 large TAg under the control of the mouse Nkx2.5 gene heartspecific enhancer (⫺9435 through ⫺7353 bp) and its proximal promoter
(⫺265 through ⫹262 bp) (27). NkL-TAg DNA was constructed by inserting 34-bp loxP sequences on either side of the TAg of Nk-TAg DNA.
The orientation of the loxP recognition sequences was confirmed by
sequencing. Nk-TAg and NkL-TAg DNA fragments were excised from
the pBluescript plasmid using XhoI and XbaI and were gel-purified
prior to microinjection of the DNA into pronuclei of fertilized B6C3F1
oocytes. Genotyping of F0 mice was performed by Southern blot and
PCR analysis using genomic DNA. The probe used for Southern blot
analysis was a 1209-bp BamHI fragment excised from the Nk-TAg
construct. PCR genotyping was performed using primers for TAg, generating a 500-bp band. Primer sequences are as follows: TAg forward,
5⬘-CGCCAGTATCAACAGCCTGTTTGGC-3⬘; TAg reverse, 5⬘CATATCGTCACGTCGAAAAAGGCGC-3⬘.
Cell Culture—Cells were isolated from subendocardial tumor-like
structures in the hearts of transgenic mice as described previously (28)
with some modifications. Briefly, dissected tissues were minced and
dissociated using an enzyme mix containing 0.2% collagen type II
(Worthington) and 0.6 mg/ml pancreatin (Sigma). Cardiac cells derived
from Nk-TAg and NkL-TAg mice were maintained in Dulbecco’s modified Eagle’s/F-12 media supplemented with 100 IU/ml penicillin, 100
␮g/ml streptomycin, 2 mM L-glutamine, and 10% fetal bovine serum
(FBS) on plates coated with 12.5 ␮g/ml fibronectin and 0.1% gelatin.
Cloning cylinders were used to harvest cell clones. Cell growth curves
were generated by plating 1.5 ⫻ 105 or 2 ⫻ 105 cells (as indicated) in
growth medium onto 100-mm plates and counting total cell number
every 24 h for 5 days. Differentiation medium contains Dulbecco’s
modified Eagle’s medium/F-12 supplemented with 100 IU/ml penicillin,
100 ␮g/ml streptomycin, 2 mM L-glutamine, 5% heat-inactivated horse
serum, 10 ␮g/ml insulin, 5.5 ␮g/ml transferrin, and 6.7 ng/ml sodium
selenite.
Histochemical Analysis—Heart tissue was fixed in 10% phosphatebuffered formalin and stained with hematoxylin-eosin. Nk-TAg and
NkL-TAg cells were cultured on coverslips and fixed for 10 min with
either ⫺20 °C methanol or 4% paraformaldehyde for immunohistochemistry. Blocking was performed by incubating fixed cells with 1.5%
bovine serum albumin and 10% normal goat serum in phosphate-buffered saline for 20 min. Primary antibodies were incubated for 30 – 60
min in 1.5% bovine serum albumin in phosphate-buffered saline as
follows: monoclonal anti-myosin (smooth) (1:100, Sigma), polyclonal
anti-myosin (skeletal) (1:100, Sigma), monoclonal anti-␣-smooth muscle
actin (1:100, Sigma), monoclonal anti-skeletal myosin (slow) (1:100,
Sigma), polyclonal anti-connexin-43 (1:100, Sigma), monoclonal anti-␣sarcomeric actin clone 5C5 (1:100, Sigma), monoclonal anti-desmin
(1:100, Sigma), monoclonal anti-calponin (1:100, Sigma), monoclonal
anti-␣-actinin (sarcomeric) clone EA-53 (1:200, Sigma), and monoclonal
anti-SV40 T-antigen (1:100, Santa Cruz Biotechnology). Secondary antibodies conjugated to either fluorescein isothiocyanate or Texas Red
(1:200, Vector Laboratories) were diluted in phosphate-buffered saline
and incubated for 30 min at room temperature. In some cases, nuclei
were co-stained with 4,6-diamidino-2-phenylindole (10 ␮g/ml) for 1 min.
Coverslips were mounted with Vectashield (Vector Laboratories), and
fluorescence or confocal images were captured using Leica DMRXE or
Zeiss 3.95 microscopes, respectively.
Measurements of Contractility and Calcium Transients in Response
to Electrical Stimulation—Excitation-contraction coupling function was
assessed as described previously (29, 30) by measuring cell contractility
and calcium transients. Briefly, cells were grown in 35-mm tissue
culture dishes for 2–3 days until 70 – 80% confluent and loaded with 2
␮M fura 2-AM (Molecular Probes) in minimum Eagle’s medium containing 0.5 mM probenecid for 15 min at 37 °C. A platinum electrode ring
was placed in the tissue culture dish. Myocyte contractions and calcium
transients were elicited by field stimulation at 1.5 Hz (Ion Optix) with
current pulses of 4-ms duration and voltages of 40 V. The polarity of the
stimulating electrodes was alternated at every pulse to prevent accumulation of electrochemical byproducts. Myocyte contractions were imaged and scanned at a rate of 240 Hz (Ion Optix). Calcium transients
were observed by exciting the fura 2-AM-loaded cells with alternating
wavelengths of 340 and 380 nm and recording the emission intensity at
510 nm. Contraction and calcium transient data for each myocyte were
recorded from a minimum of 12 consecutive stimuli.
Drug Treatment—Drugs were added into cell culture media as indicated at the following concentrations: 100 ␮M phenylephrine (PE), 10
␮M norepinephrine (NE), 10 ng/ml recombinant human transforming
growth factor-␤1 (TGF-␤1) (R & D Systems), 1 ␮M dynorphin-␤ (Peninsula Laboratories), 1 ␮M trans- or cis-retinoic acid (Sigma), 15 ng/ml
bone morphogenetic protein (BMP)-2/4 (Genetics Institute, Cambridge,
MA), 1 ␮M angiotensin II (R & D Systems), 20 nM endothelin-1 (R & D
Systems), 100 ng/ml insulin-like growth factor I (IGF-I) (Roche Molecular Biochemicals), 5-azacytidine (Sigma), and 10 ␮g/ml mitomycin C
(Sigma).
Viral Infection—Cells were infected with recombinant adenoviruses
(Ad) at a multiplicity of infection of 100 for 3–12 h. The medium was
replaced with growth medium, and NkL-TAg cells were cultured for the
indicated times before analyses. Recombinant adenoviruses were obtained from the following sources: Ad-Cre recombinase (Ad-Cre) was
provided by Dr. Frank Graham (McMaster University) (25); GATA4
(Ad-GATA4), Nkx2.5 (Ad-Nkx2.5), MEK6 (Ad-MEK6), and green fluorescent protein (Ad-GFP) were generated using the “Easy-Track” system as described (31); antisense HDAC4 and HDAC5 (Ad-HDAC4 or -5),
expressing coding regions of HDAC genes in reverse orientation, were
provided by Dr. Chun Zhang; MEF2C (Ad-MEF2C) was provided by Dr.
Rebekka Nicol (HDAC 4/5 and MEF2C viruses were generated using
the pAC-CMV vector); constitutively active calcineurin (Ad-CnA), IGF-I
receptor (Ad-IGFI), constitutively active CaMKI (Ad-CaMKI), and ␤-galactosidase (Ad-LacZ) were generated by Dr. Robert Gerard (University
of Texas Southwestern) and were constructed using an in vitro Cre-lox
recombination system (32, 33).
DNA Synthesis Assay—NkL-TAg cells were infected with Ad-Cre
virus as described above and cultured in growth medium for 2 more
days followed by incubation with BrdUrd for 2 h. DNA synthesis was
Downloaded from www.jbc.org at VIVA, Univ of Virginia on May 19, 2009
expression of TAg in cardiac, skeletal, and smooth muscle
showed rapid proliferation and, in the case of conditional expression, retained some degree of differentiation (13–20). In
particular, the AT-1 cell line derived from atrial tumors of
transgenic mice expressing SV40 TAg driven by the atrial
natriuretic factor promoter was the first cardiac cell line that
was able to maintain contractility in vitro after being passaged
several times (21, 22). The main drawback of this line is that it
must be propagated as ectopic grafts in syngeneic mice and
cannot be passaged in vitro or frozen. In contrast, the HL-1 cell
line derived from the original AT-1 line can be passaged in
conventional cell culture and frozen, although it must be maintained in a proprietary media of unknown composition supplemented with growth factors and hormones such as insulin,
retinoic acid, and norepinephrine (23). Neither AT-1 nor HL-1
cells become quiescent by serum deprivation. These drawbacks
significantly limit their use in signal transduction and biochemical studies.
The Nkx2.5 gene encodes a homeodomain transcription factor that is among the earliest known markers of cardiogenesis
in the vertebrate embryo (24). Recently, a cardiac cell line that
expresses TAg under the control of the proximal cardiac enhancer of the Nkx2.5 gene was derived (13). This cardiac cell
line exhibited a cardiac embryonic like phenotype. However, it
was devoid of striations or contractile properties and unable to
exit the cell cycle.
In the present study, we developed immortalized cardiac cell
lines from hearts of transgenic mice carrying cardiac-specific
regulatory sequences from the mouse Nkx2.5 gene to drive the
expression of the TAg coding region flanked by loxP sites. We
implemented the Cre-lox system as an efficient method to
inactivate genes permanently (25, 26), allowing for Cre recombinase-mediated deletion of the TAg gene. We show that these
cell lines proliferate rapidly until they are infected with an
adenovirus encoding Cre recombinase, at which time they
cease expressing TAg and exit the cell cycle. Cessation of cell
division in these cells is accompanied by an alteration of cell
morphology, assembly of an organized stress fiber network, and
changes in gene expression characteristic of differentiated
myocytes. This strategy for generating reversibly transformed
cell lines may be widely applicable for the generation of cell
lines from a variety of tissues that are otherwise unable to be
maintained in proliferative or differentiated states in culture.
Generating Conditional Cardiomyocyte Cell Lines
15929
determined based on BrdUrd incorporation using a BrdUrd assay kit
(Roche Molecular Biochemicals) according to the manufacturer’s
instructions.
Western Blot Analysis—Cells were lysed in 10 mM Tris-HCl (pH 6.8),
100 mM NaCl, 1% SDS, 1 mM EDTA, 1% EGTA buffer. Total protein (50
␮g) from lysate was resolved on a 10% SDS-PAGE gel and transferred
to a polyvinylidene difluoride membrane. Membranes were incubated
with primary antibodies recognizing GATA4 (Santa Cruz Biotechnology) and TAg (Santa Cruz Biotechnology) at 1:1000 dilution followed by
incubation with secondary antibodies against rabbit IgG or mouse IgG
(1:7500) conjugated to horseradish peroxidase (Santa Cruz Biotechnology). Signal was detected using ECL reagent (Amersham Biosciences).
RT-PCR and Northern Blot Analysis—RNA was isolated from cells
using Trizol Reagent (Invitrogen) and used in Northern blot analysis
with a probe generated from the coding region of Nkx2.5 or TAg.
RT-PCR was performed using the Superscript II kit (Invitrogen).
Primers used for amplification are listed in the Supplemental Material
Table S1.
Microarray Hybridizations—Microarray analyses for NkL-TAg cells
versus NIH3T3 cells were performed using the InCyte Genomics Mouse
cDNA microchip. Microarray analyses for NkL-TAg cells infected with
Ad-Cre or Ad-lacZ were performed at the University of Texas Southwestern Core Facility using Affymetrix mouse oligo microchip
(MU74A). NkL-TAg cells were infected either with adenovirus expressing Cre recombinase (Ad-Cre) or ␤-galactosidase (Ad-LacZ) at 100 multiplicity of infection per cell for 3 h. Cells were incubated for 4 days in
differentiation media (described above) containing 100 ␮M PE followed
by RNA isolation using Trizol reagent (Invitrogen). Data were analyzed
using GeneSpring and Affymetrix Suite software.
RESULTS
Generation of Nkx2.5-TAg Transgenic Mice with Cardiac
Tumors—The Nkx2.5 cis-regulatory sequences consisting of the
early cardiac-specific enhancer region (⫺9435 to ⫺7353 bp)
fused to the endogenous promoter (⫺265 to ⫹262 bp) was
linked to the coding region for SV-40 large T-antigen (Fig. 1A)
and used to generate transgenic mice. Founder transgenic mice
appeared normal at birth and showed no abnormalities during
the neonatal period. However, one female (Fo) transgenic
mouse died spontaneously at 5 weeks of age. An autopsy
showed that the heart was grossly enlarged with multiple
sessile masses protruding into the left ventricular chamber
from the interventricular septum and anterior surface of the
ventricular free wall (Fig. 1B). Histological analysis confirmed
that the masses were localized subendocardially and consisted
of small poorly differentiated, spindle shaped cells with small
eosin-rich cytoplasm without apparent striation (Fig. 1C).
FIG. 2. Characterization of Nk-TAg cardiac cell lines. A, two
independent Nk-TAg cell lines (lines 5 and 20) were plated at a low
density (1.5 ⫻ 105 cells/100-mm plate) and were counted for 5 subsequent days. Cells reached confluence on day 4 but failed to undergo
growth arrest. B, expression of Nkx2.5 by Northern blot analysis and
T-antigen and GATA4 by Western blot analysis in isolated Nk-TAg cell
lines. All of the established cell lines expressed Nkx2.5, GATA4, and
TAg genes as detected by RT-PCR (data not shown), although the level
of expression varied for each cell line.
Many loci of myocardial hyperplasia were noted, none of which
involved the endocardium. The architecture of the remaining
myocardium was preserved although many cardiomyocytes
had excessively large hematoxylin-rich nuclei as a possible sign
of polyploidy. It was not possible to determine the cause of
death of the animal; however, it is plausible that either outflow
obstruction or ventricular arrhythmia led to sudden death.
Isolation of Immortalized Cardiac Cells—A heart harvested
from a 3.5-week-old (F0) transgenic Nkx2.5-TAg mouse displaying mild cyanosis exhibited protruding masses in the left ventricle. These tumors were excised from the myocardium and
dissociated into single cells. The cells were seeded at low density onto fibronectin/gelatin-coated plates and grown for 10
days. During this period several well defined cell colonies
emerged consisting of small spindle-shaped cells growing on
top of each other. They showed no contractile activity. Cell
colonies were cloned using cloning cylinders and independently
subcultured in 24-well plates. Twenty one individual colonies
were isolated, although only 18 survived subsequent passages.
These cell lines were named Nk-TAg lines.
Following dissociation of the colonies, the plated cells grew
as a monolayer. Different growth rates were observed for individual clones (Fig. 2A). Upon reaching confluence the cells
continued to proliferate, showing no evidence of contact inhibition. When cells were plated at 1.5 ⫻ 105 cells per 10-cm plate
and grown in medium containing 20% FBS, we observed a
doubling time of less than 24 h. Decreasing the serum content
to 15% slightly reduced the doubling time indicating that NkTAg cells respond to factors in serum (data not shown).
Gene Expression Profile of Nk-TAg Cells—All of the Nk-TAg
cell lines expressed TAg, but expression levels of TAg varied in
the different cell lines (Fig. 2B). A correlation was seen between
the expression level of TAg and growth rate of cells. However,
there was no relation seen between expression levels of TAg
Downloaded from www.jbc.org at VIVA, Univ of Virginia on May 19, 2009
FIG. 1. Nk-TAg transgene and cardiac tumors in Nk-TAg transgenic mouse. A, diagram of the transgenic construct Nk-TAg containing the Nkx2.5 cis-regulatory sequences consisting of the early cardiacspecific enhancer region (⫺9435 to ⫺7353 bp), fused to the endogenous
promoter (⫺265 to ⫹262 bp), and linked to the SV-40 large T-antigen
coding region. B, cross-section of a heart of a 3.5-week-old mouse expressing the Nk-TAg transgene showing multiple tumors protruding
into the left ventricular chamber (arrowheads). C, hematoxylin-eosin
staining of a histological section of a Nk-TAg transgenic heart. The
cardiac tumor (indicated by arrow) is localized subendocardially and
consists of poorly differentiated eosin-rich spindle-shaped cells.
15930
Generating Conditional Cardiomyocyte Cell Lines
TABLE I
Genes expressed in Nk-TAg cellsa
Line 20
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫺
⫹
⫹
⫹
⫺
⫺
⫺
⫺
⫺
⫹
⫺
⫺
⫺
⫹
⫹
⫹
⫹
⫺
⫹
⫹
⫹
⫹
⫺
⫹
⫹
⫹
⫺
⫺
⫺
⫹
⫹
⫹
⫺
⫺
⫺
a
Detected by Northern blot analysis, RT-PCR, and/or Western blot
analysis.
b
MHC is myosin heavy chain; ANF is atrial natriuretic factor.
FIG. 4. Analysis of NkL-TAg transgenic mouse cell lines. A,
diagram of the NkL-TAg transgenic construct containing the Nkx2.5
cis-regulatory sequences consisting of the early cardiac-specific enhancer region (⫺9435 to ⫺7353 bp), fused to the endogenous promoter
(⫺265 to ⫹262 bp), and linked to the SV-40 large T-antigen coding
region flanked by loxP sites (red diamonds). B, immunocytochemistry
using TAg antibody of NkL-TAg cells (control) and NkL-TAg cells infected with Ad-Cre. An equivalent number of cells was placed on plates.
All the cells stained for TAg in the absence of Cre recombinase (left
panel), whereas less than 1% of cells stained for TAg after addition of
Cre recombinase (right panel); C, BrdUrd incorporation was measured
in NkL-TAg cells (control) and NkL-TAg cells infected with Ad-Cre.
Equivalent numbers of cells were counted at ⫻20 magnification. D,
NkL-TAg (control) or NkL-TAg cells infected with Ad-LacZ or Ad-Cre
were plated at 2 ⫻ 105 cells per 100-mm plate. Cells were counted for 5
subsequent days. NkL-TAg cells stop proliferating in response to excision of the TAg gene.
FIG. 3. Mitomycin C induces expression of ␣-actinin in Nk-TAg
cells. Nk-TAg cells treated with mitomycin C were stained with 4,6diamidino-2-phenylindole (A) and with anti-␣-actinin antibody (B and
C). Images were taken at ⫻40 magnification (A and B) and ⫻100
magnification (C).
and cardiac-enriched gene expression. The majority of Nk-TAg
cell lines expressed transcripts encoding proteins characteristic
of cardiomyocytes, such as the Nkx2.5, GATA4, and MEF2C
transcription factors (Fig. 2B and see Table I for a list of other
genes). However, many structural proteins that are essential
for contraction of cardiac myocytes, including titin and sodium,
calcium-exchanger (Ncx-1), were not detected even by RT-PCR
(Table I).
Immunohistochemistry analysis of Nk-TAg cells showed a
low level of expression of ␣-actinin, a prototypical Z-line protein
in cardiomyocytes (data not shown). Growing the cells on plates
coated with laminin or type II collagen at normal (10%) or low
serum content (2 or 5%) did not increase expression of ␣-actinin
(data not shown). However, addition of mitomycin C to the
growth medium abated proliferation of the Nk-TAg cells and
enhanced expression of ␣-actinin in the cytoplasm as unassembled Z-lines (Fig. 3). This finding suggested that inhibition
of DNA synthesis in Nk-TAg cells promoted a cardiogenic phenotype. Addition of hypertrophic agents including endothelin-1, PE, and angiotensin II did not further induce the assembly of sarcomeres in Nk-TAg cells (data not shown).
Generation of Conditional TAg-transformed Cardiomyocyte
Cell Line—Based on the finding that mitomycin C enhanced
␣-actinin expression in Nk-TAg cells, we postulated that the
lack of growth control is counterproductive to establishing a
cardiomyocyte cell line. Therefore, in an effort to induce cellular differentiation, we sought to generate cardiac cell lines
capable of terminating cell division. We chose the Cre-lox system as an efficient method to permanently remove and thereby
inactivate TAg.
Two loxP sites were added to flank the TAg-encoding region
in the Nk-TAg DNA construct (Fig. 4A), and this construct,
NkL-TAg, was used to generate transgenic mice. Dissection of
the hearts of NkL-TAg transgenic mice at 3.5 weeks of age
revealed one mouse with gross cardiomegaly due to multiple
ventricular myocardial tumors similar to those found in Nk-TAg
mice (data not shown, see Fig. 1). Histological analysis of the
transgenic heart revealed a phenotype similar to that observed in
the hearts of Nk-TAg transgenic animals (data not shown).
Downloaded from www.jbc.org at VIVA, Univ of Virginia on May 19, 2009
T-antigen
Nkx-2.5
GATA4
Myocardin
MEF2C
Oracle
CHAMP
Tropomyosin
␤-MHCb
␣-MHC
Troponin I
␣-Cardiac Actin
Connexin 45
Dystrophin
Myoglobin
MURF2
Calsarcin
MLC2v
KvLQT
Titin
ANF
Ncx-1
Line 5
Generating Conditional Cardiomyocyte Cell Lines
Cells were dissociated from the tumor regions of the NkLTAg transgenic hearts and plated onto fibronectin/gelatincoated dishes at a low density. Following 10 days in growth
medium containing 10% FBS, the NkL-TAg cells showed dif-
FIG. 6. Effect of different treatments on the morphology of NkL-TAg
cells before and after deletion of TAg.
Uninfected NkL-TAg cells (control) or
NkL-TAg cells infected with Ad-Cre were
cultured in differentiation media containing PE, NE, BMP2/4, Ad-CaMKI, or Adcalcineurin. Immunocytochemistry was
performed using anti-␣-actinin (sarcomeric) antibody. Magnification at ⫻40.
ferent characteristics than the Nk-TAg cells. In contrast to the
Nk-TAg cells, NkL-TAg cells did not form well defined colonies
and did not grow on top of each other. Remarkably, some of the
plated cells maintained contractile activity during the initial
passages, although this characteristic gradually disappeared after a few cell passages. Immunohistochemistry confirmed TAg
expression in the NkL-TAg cells and also showed that some cells
expressed ␣-actinin in a non-striated pattern (data not shown).
NkL-TAg Cells Exit the Cell Cycle following Cre-recombinase
Expression—NkL-TAg cells grew without apparent senescence
even after 50 serial cell passages. Immunocytochemistry using
TAg antibody showed that all NkL-TAg cells expressed TAg
protein regardless of the passage number. Infection of NkLTAg cells with recombinant adenovirus expressing Cre recombinase (Ad-Cre) effectively removed the TAg gene from the
cellular genome as confirmed by PCR (data not shown) and
immunocytochemistry using TAg antibody (Fig. 4B). Deletion
of the TAg gene was followed by a dramatic decrease in BrdUrd
incorporation (Fig. 4C) and eventual cessation of cell growth of
the NkL-TAg cells (Fig. 4D). In contrast, infection of NkL-TAg
cells with a recombinant adenovirus expressing ␤-galactosidase (Ad-LacZ) initially caused a decrease in the cell growth
rate, but ultimately resulted in no change in cell proliferation
rate when compared with non-infected cells (Fig. 4D).
Characterization of NkL-TAg Cells before and after Removal
of Tag—Although NkL-TAg cells showed a gene expression
pattern similar to Nk-TAg cells (Table I), additional cardiacspecific genes were expressed in NkL-TAg cells including myoglobin, ␣-myosin heavy chain, FHL2/DRAL, MCIP1, MEF2D,
and atrial natriuretic factor. Immunocytochemistry revealed
expression of desmin and cadherin (Fig. 5, A and B) as well as
connexin-43 (data not shown) that was localized to the cell
junctions.
Downloaded from www.jbc.org at VIVA, Univ of Virginia on May 19, 2009
FIG. 5. Immunocytochemistry of NkL-TAg cells. A, NkL-TAg
cells were immunostained with antibodies recognizing desmin or cadherin. B, NkL-TAg cells non-infected (control) or infected with Ad-Cre
were immunostained with antibody recognizing ␣-smooth muscle actin.
Magnification at ⫻40.
15931
15932
Generating Conditional Cardiomyocyte Cell Lines
Deletion of the TAg and subsequent withdrawal from the cell
cycle resulted in a dramatic increase in cell surface area (up to
10-fold and dependent on cell density). Immunocytochemistry
showed that NkL-TAg cells depleted of TAg exhibited pronounced stress fibers that were immunoreactive for smooth
muscle actin (Fig. 5, C and D). After removal of TAg, many of
the NkL-TAg cells become binucleated. However, depletion of
TAg had no apparent effect on the level of expression of selected
FIG. 8. Microarray analysis of gene expression profiles of the NkL-TAg cell line. Scatter plot of microarray analysis using RNA isolated from
uninfected NkL-TAg cells (control) and NkL-TAg cells infected with Ad-lacZ (A); uninfected NkL-TAg cells (control) and NkL-TAg cells infected with
Ad-Cre (B). C, NkL-TAg cells infected with Ad-LacZ and NkL-TAg cells infected with Ad-Cre. Each dot represents a gene on the microarray slide. Black
dots denote genes that are expressed in both samples. Green dots represent genes exclusively expressed in NkL-TAg cells infected with Ad-lacZ (A),
NkL-TAg cells infected with Ad-Cre (B), and NkL-TAg cells infected with Ad-Cre (C). Blue dots represent genes exclusively expressed in A, and
NkL-TAg cells (control) in B, and NkL-TAg cells infected with Ad-lacZ in C. Diagonal lines on the graphs show 2-, 5-, and 10-fold differences.
Downloaded from www.jbc.org at VIVA, Univ of Virginia on May 19, 2009
FIG. 7. Measurement of calcium transients in NkL-TAg cell
lines. A, NkL-TAg cells infected with Ad-Cre were subjected to an
electrical stimulation at 1.5 Hz (Ion Optix) with current pulses of 4-ms
duration and voltages of 40 V. Calcium transients were observed by
exciting the fura 2-AM-loaded cells with alternating wavelengths of 340
and 380 nm and recording the emission intensity at 510 nm. Calcium
transient data for each myocyte were recorded from a minimum of 12
consecutive stimuli. B, similar measurements were performed on
mouse adult cardiomyocytes. Pulses are indicated with arrowheads.
cardiac-specific genes as tested by RT-PCR and/or immunocytochemistry (data not shown).
Induction of Differentiation of NkL-TAg Cells by Various
Stimuli—We surveyed the effect of different drugs, hormones,
and overexpression of signaling proteins in attempts to induce
further differentiation of NkL-TAg cells. Differentiation was
assessed using immunocytochemistry with an antibody against
sarcomeric ␣-actinin. Prior to depletion of TAg, addition of
caffeine, potassium chloride, 5-azacytidine, low oxygen, PE,
NE, IGF-I, cis- or trans-retinoic acids, dynorphin, TGF-␤1,
ionomycin, basic fibroblast growth factor, or recombinant adenovirus containing various transcripts (CMV-LacZ, CMV-Calsarcin1, CMV-GATA4, CMV-Nkx2.5, CMV-IGFI, CMVHDAC5, CMV-CaMKI, CMV-MEK6, and CMV-calcineurin A)
had no effect on inducing sarcomere formation as assessed by
␣-actinin immunocytochemistry. However, expression of ␣-actinin was increased in NkL-TAg cells upon removal of TAg and
switching the medium to contain low serum (5% horse serum),
insulin, transferrin, and selenium (Fig. 6). Moreover, addition
of PE, BMP2/4, or NE, to the media led to further induction of
sarcomere formation in some cells (Fig. 6). A similar induction
was seen when NkL-TAg cells depleted of TAg were grown in
differentiation media and infected with recombinant adenovirus expressing constitutively active CaMKI or constitutively
active calcineurin A (Fig. 6), known effectors of cardiomyocyte
hypertrophy (34, 35).
NkL-TAg Cells Exhibit Calcium Transients in Response to
Electrical Stimulation—Calcium current and cell contractility
were analyzed to determine whether NkL-TAg cells were excitable. NkL-TAg cells expressing TAg showed no response to
electrical stimulation. However, when NkL-TAg cells depleted
of TAg and cultured for 7 days in differentiation medium containing PE were subjected to electrical stimulation, calcium
transients were readily detected in ⬃30% of the cells examined
(Fig. 7A). This implies that the sarcoplasmic reticulum of NkLTAg cells releases calcium in response to electrical stimulation.
Despite this fact, cells failed to contract. In comparison to
freshly isolated adult cardiac myocytes (Fig. 7B), NkL-TAg
cells showed a diminished response. In addition, calcium transients in NkL-TAg cells were elicited in response to every other
stimulus at the 50-Hz stimulation frequency, whereas freshly
isolated cardiac myocytes responded to each electrical stimulus. The inability of NkL-TAg cells to respond to every stimulus
may be attributable to delayed restoration of excitability because intracellular calcium in these cells may return to the
basal level more slowly than in normal adult myocytes.
Microarray Analysis of Nk-TAg and NkL-TAg Cell Lines—
Gene expression profiles were examined using microarray
analysis comparing RNA transcripts of NkL-TAg cells with
NIH/3T3 fibroblasts. Scatter plot analysis of the microarray
Generating Conditional Cardiomyocyte Cell Lines
DISCUSSION
The results of this study demonstrate that cardiomyocyte cell
lines capable of proliferation and irreversible withdrawal from
the cell cycle can be generated using a conditional cardiacspecific TAg transgene. Through the use of Cre recombinase
delivered via adenoviral infection, such cell lines can be propagated in culture and rapidly switched from a proliferative to a
terminally quiescent state. This strategy should have applications for the creation of diverse types of cell lines, as well as for
tissue repair and regeneration.
We used Nkx2.5 regulatory sequences, which are active from
the initial stages of cardiogenesis, to direct the expression of
TAg. The rationale for this strategy was to derive an early
cardiac precursor cell line, which upon appropriate stimulus
can be induced to differentiate. Although it is clear that NkLTAg cells exhibit a cardiac phenotype, we cannot be certain
whether these cells are arrested at a specific early stage in the
cardiac developmental pathway or whether they progress partially toward a more mature stage of cardiac development. Our
gene expression data as well as the data of Brunskill et al. (13)
support the notion that cells derived from hearts transformed
by the Nkx2.5 promoter-driven TAg tend to express genes present in both embryonic and adult cardiomyocytes.
It is interesting that only a fraction of the cardiomyocytes
from the Nk-TAg transgenic mice appear to be transformed,
thus forming subendocardial tumors and allowing the remainder of the heart to develop and function normally. We believe
the most likely explanation for this heterogeneity is mosaicism
of transgene expression in founder transgenic mice, because it
would seem that embryonic lethality would result from mice
with fully transformed cardiac cells lacking contractile machinery. Consistent with this notion, we have not observed transmission of the transgene to the F1 generation. Other explanations for observed mosaicism include the following: 1)
additional sources of cardiac progenitors from outside the early
heart-forming region in which the Nkx2.5 transgene is expressed; 2) a “second hit” in cells that form the TAg-transformed tumors; or 3) an expression gradient of the Nkx2.5
transgene from the endocardium into the ventricular walls.
Our observation that NkL-TAg cells permanently withdraw
from the cell cycle upon removal of TAg disputes the second hit
theory and implies that expression of TAg is sufficient and
necessary for propagating these cardiac cells. It also indicates
that NkL-TAg cells even after multiple rounds of proliferation
retain the cell cycle arrest mechanism that is characteristic of
cardiomyocytes. These findings as well as increased expression
of cell cycle inhibitor genes (data not shown) indicate that these
cells could be used as a reagent to study cell cycle events in
cardiomyocytes.
Expression of TAg is compatible with the expression of many
early cardiac lineage markers but is incompatible with growth
arrest and formation of mature sarcomeres. The use of a
“floxed” TAg transgene circumvents these effects and allows for
a rapid and reversible switch in the proliferative properties of
cardiac muscle cells. To our knowledge, this type of conditional
Cre-lox approach has not been utilized previously to expand
and control the differentiation of specific populations of progenitor cells in vitro. It may be useful in the future to use similar
Cre-lox methods to expand and study many progenitor or mature cell lines depending on the promoter or transforming
agent applied.
The “floxed” TAg approach proved to be highly efficient and
efficacious for the control of proliferation of the derived NkLTAg cell lines. However, withdrawal of NkL-TAg cells from the
cell cycle was not sufficient to induce differentiation as assessed by sarcomere formation. It required cultivation of TAgdepleted NkL-TAg cells in low serum media with addition of
several factors or overexpression of genes to induce formation
of sarcomeres. This was accompanied by dramatic changes in
gene expression as revealed by the microarray data. Subpopulation of these cells exhibited Ca2⫹ current upon electrical
stimulation. Although TAg-depleted NkL-TAg cells do not contract spontaneously or upon electrical stimulation, further
studies may focus on this process, perhaps utilizing different
culture strategies.
In summary, we have created cardiomyocyte cell lines using
the Nkx2.5 promoter/enhancer and TAg. These cell lines provide interesting opportunities for the study of cardiomyocytes
particularly because the NkL-TAg cell line can be expanded in
culture and induced to a degree of differentiation with expression of Cre recombinase. In general, the Cre-lox method we
have utilized may be useful for the development of different cell
lines capable of proliferation followed by differentiation. NkLTAg cells may serve as a useful tool for studies of cardiac gene
regulation and cellular signaling, as well as for novel gene
discovery.
Acknowledgments—We thank John McAnally for generating transgenic mice; Dr. Ralph Shohet and Teresa Gallardo for assistance with
microarray data analysis; Dr. Robert Gerard for providing adeno-
Downloaded from www.jbc.org at VIVA, Univ of Virginia on May 19, 2009
results revealed three distinct groups of genes. The first and
largest group contains genes that are equally represented in
both of the cell lines, consisting primarily of housekeeping
genes. The second group contains genes that are predominantly
or exclusively expressed in NIH/3T3 cells and consists of nonmuscle genes. The third group of genes is predominantly or
exclusively expressed in NkL-TAg cells and consists mainly of
cardiac-specific genes, supporting the premise that the NkLTAg cell line is a cardiomyocyte cell line (see Supplemental
Material, Table S2). Notably, many of the genes revealed by
microarray analysis, such as those encoding calponin, smooth
muscle actin, skeletal actin, and atrial natriuretic factor, are
characteristic of embryonic cardiomyocytes. Furthermore,
analysis of the microarray data revealed several uncharacterized ESTs expressed in NkL-TAg cells. These ESTs were shown
to be selectively expressed in heart by Northern blot and in situ
hybridization (data not shown).
Microarray analysis was also performed on RNA transcripts
isolated from NkL-TAg cells (grown in differentiation media
with PE) expressing or depleted of TAg. Infection of NkL-TAg
cells with Ad-LacZ showed an up-regulation of 298 genes; 197
of these genes were also up-regulated in NkL-TAg cells infected
with Ad-Cre, and 232 genes were down-regulated, including 74
genes that were down-regulated in Ad-Cre-infected cells (Fig.
8A). Further studies were not done on these genes, viewing
them as a cellular response to adenovirus infection. Deletion of
the TAg gene by Cre-recombinase led to significant alterations
of gene expression in NkL-TAg cells; 313 genes were downregulated (⬎2-fold), and 214 genes were up-regulated (⬎2-fold)
(Fig. 8, B and C). Specifically, the majority of genes downregulated in TAg-deleted NkL-TAg cells were involved in cell
cycle regulation, including those encoding cyclin A2, cyclin E2,
cyclin B1, cyclin B2, cdc45, cdc46, cdc6, cdc7, Mcm2, cdc25, and
cdk2. In contrast, many of the transcripts that were up-regulated upon deletion of TAg were cardiac-associated genes including those encoding slow/cardiac troponin C; ZASP, a cardiac-specific Z-band protein; skeletal muscle actin; brain
natriuretic peptide; fatty acid transport protein 4; myotonic
dystrophy protein kinase; and sarcoglycan, a dystrophin-associated glycoprotein. Expression of some of these genes was
confirmed with immunohistochemistry on NkL-TAg cells in the
absence or presence of Ad-Cre (data not shown).
15933
15934
Generating Conditional Cardiomyocyte Cell Lines
viruses; Dr. Robert Hammer for providing DNA encoding SV-40
T-antigen; Dr. James Richardson for assistance with histological analyses; and Alisha Tizenor for assistance with images.
REFERENCES
Downloaded from www.jbc.org at VIVA, Univ of Virginia on May 19, 2009
1. Soonpaa, M. H., and Field, L. J. (1997) Am. J. Physiol. 272, H220 –H226
2. MacLellan, W. R., and Schneider, M. D. (2000) Annu. Rev. Physiol. 62,
289 –319
3. Rumyantsev, P. P. (1977) in International Review of Cytology (Bourne, G. H.,
and Danielli, J. F., eds) pp. 187–273, Academic Press, New York
4. Doetschman, T. C., Eistetter, H., Katz, M., Schmidt, W., and Kemler, R. (1985)
J. Embryol. Exp. Morphol. 87, 27– 45
5. Edwards, M. K., Harris, J. F., and McBurney, M. W. (1983) Mol. Cell. Biol. 3,
2280 –2286
6. Makino, S., Fukuda, K., Miyoshi, S., Konishi, F., Kodama, H., Pan, J., Sano,
M., Takahashi, T., Hori, S., Abe, H., Hata, J., Umezawa, A., and Ogawa, S.
(1999) J. Clin. Invest. 103, 697–705
7. Muller, M., Fleischmann, B. K., Selbert, S., Ji, G. J., Endl, E., Middeler, G.,
Muller, O. J., Schlenke, P., Frese, S., Wobus, A. M., Hescheler, J., Katus,
H. A., and Franz, W. M. (2000) FASEB J. 14, 2540 –2548
8. Klug, M. G., Soonpaa, M. H., Koh, G. Y., and Field, L. J. (1996) J. Clin. Invest.
98, 216 –224
9. Eisenberg, C. A., and Bader, D. M. (1996) Circ. Res. 78, 205–216
10. Kimes, B. W., and Brandt, B. L. (1976) Exp. Cell Res. 98, 367–381
11. Engelmann, G. L., Birchenall-Roberts, M. C., Ruscetti, F. W., and Samarel,
A. M. (1993) J. Mol. Cell. Cardiol. 25, 197–213
12. Manfredi, J. J., and Prives, C. (1994) Biochim. Biophys. Acta 1198, 65– 83
13. Brunskill, E. W., Witte, D. P., Yutzey, K. E., and Potter, S. S. (2001) Dev. Biol.
235, 507–520
14. Jahn, L., Sadoshima, J., Greene, A., Parker, C., Morgan, K. G., and Izumo, S.
(1996) J. Cell Sci. 109, 397– 407
15. Jat, P. S., Noble, M. D., Ataliotis, P., Tanaka, Y., Yannoutsos, N., Larsen, L.,
and Kioussis, D. (1991) Proc. Natl. Acad. Sci. U. S. A. 88, 5096 –5100
16. Gu, W., Schneider, J. W., Condorelli, G., Kaushal, S., Mahdavi, V., and NadalGinard, B. (1993) Cell 72, 309 –324
17. Miller, C., Rulfs, J., Jaspers, S. R., Buckholt, M., and Miller, T. B., Jr. (1994)
Mol. Cell. Biochem. 136, 29 –34
18. Morgan, J. E., Beauchamp, J. R., Pagel, C. N., Peckham, M., Ataliotis, P., Jat,
P. S., Noble, M. D., Farmer, K., and Partridge, T. A. (1994) Dev. Biol. 162,
486 – 498
19. Mouly, V., Edom, F., Decary, S., Vicart, P., Barbert, J. P., and Butler-Browne,
G. S. (1996) Exp. Cell Res. 225, 268 –276
20. Tedesco, D., Caruso, M., Fischer-Fantuzzi, L., and Vesco, C. (1995) J. Virol. 69,
6947– 6957
21. Kline, R. P., Sorota, S., Dresdner, K. P., Steinhelper, M. E., Lanson, N. A., Jr.,
Wit, A. L., Claycomb, W. C., and Field, L. J. (1993) J. Cardiovasc. Electrophysiol. 4, 642– 660
22. Steinhelper, M. E., Lanson, N. A., Jr., Dresdner, K. P., Delcarpio, J. B., Wit,
A. L., Claycomb, W. C., and Field, L. J. (1990) Am. J. Physiol. 259,
H1826 –H1834
23. Claycomb, W. C., Lanson, N. A., Stallworth, B. S., Egeland, D. B., Delcarpio,
J. B., and Izzo, N. J. (1998) Proc. Natl. Acad. Sci. U. S. A. 95, 2979 –2984
24. Lints, T. J., Parsons, L. M., Hartley, L., Lyons, I., and Harvey, R. P. (1993)
Development 119, 419 – 431
25. Anton, M., and Graham, F. L. (1995) J. Virol. 69, 4600 – 4606
26. Kawamoto, S., Niwa, H., Tashiro, F., Sano, S., Kondoh, G., Takeda, J., Tabayashi, K., and Miyazaki, J. (2000) FEBS Lett. 470, 263–268
27. Lien, C. L., Wu, C., Mercer, B., Webb, R., Richardson, J. R., and Olson, E. N.
(1999) Development 126, 75– 84
28. Paradis, P., MacLellan, W. R., Belaguli, N. S., Schwartz, R. J., and Schneider,
M. D. (1996) J. Biol. Chem. 271, 10827–10833
29. Cyran, S. E., Ditty, S. E., Baylen, B. G., Cheung, J., and LaNoue, K. F. (1992)
J. Mol. Cell. Cardiol. 24, 1167–1177
30. De Young, M. B., and Scarpa, A. (1989) Am. J. Physiol. 257, C750 –C758
31. He, T. C., Zhou, S., Da Costa, L. T., Yu, J., Kinzler, K. W., and Vogelstein, B.
(1998) Proc. Natl. Acad. Sci. U. S. A. 95, 2509 –2514
32. Aoki, K., Barker, C., Danthinne, X., Imperiale, M. J., and Nabel, G. J. (1999)
Mol. Med. 5, 224 –231
33. Ng, P., Parks, R. J., Cummings, D. T., Evelegh, C. M., Sankar, U., and
Graham, F. L. (1999) Hum. Gene Ther. 10, 2667–2672
34. Passier, R., Zeng, H., Frey, N., Naya, F. J., Nicol, R. L., McKinsey, T. A.,
Overbeek, P., Richardson, J. A., Grant, S. R., and Olson, E. N. (2000)
J. Clin. Invest. 105, 1395–1406
35. Molkentin, J. D., Lu, J. R., Antos, C. L., Markham, B., Richardson, J., Robbins,
J., Grant, S. R., and Olson, E. N. (1998) Cell 93, 215–228