Mature Vascular Endothelium Can Give Rise to Smooth
Muscle Cells via Endothelial-Mesenchymal Transdifferentiation
In Vitro Analysis
Maria G. Frid, Vishakha A. Kale, Kurt R. Stenmark
Downloaded from http://circres.ahajournals.org/ by guest on June 17, 2017
Abstract—Though in the past believed to be a rare phenomenon, endothelial-mesenchymal transdifferentiation has been
described with increasing frequency in recent years. It is believed to be important in embryonic vascular development, yet less
is known regarding its role in the adult vasculature. Using FACS and immunomagnetic (Dynabeads) purification techniques
(based on uptake of DiI-acetylated low-density lipoproteins and/or PECAM-1 expression) and double-label indirect
immunostaining (for endothelial and smooth muscle [SM] markers), we demonstrate that mature bovine vascular endothelium
contains cells of an endothelial phenotype (defined by VE-cadherin, von Willebrand factor, PECAM-1, and elevated uptake
of acetylated low-density lipoproteins) that can undergo endothelial-mesenchymal transdifferentiation and further differentiate
into SM cells (as defined by expression of ␣-SM-actin, SM22␣, calponin, and SM-myosin). “Transitional” cells, coexpressing
both endothelial markers and ␣-SM-actin, were consistently observed. The percentage of cells capable of endothelialmesenchymal transdifferentiation within primary endothelial cultures was estimated as 0.01% to 0.03%. Acquisition of a SM
phenotype occurred even in the absence of proliferation, in ␥-irradiated (30 Gy) and/or mitomycin C–treated primary cell
cultures. Initiation of transdifferentiation correlated with disruption of cell-cell contacts (marked by loss of VE-cadherin
expression) within endothelial monolayers, as well as with the action of transforming growth factor-␤1. In conclusion, our in
vitro data show that mature bovine systemic and pulmonary endothelium contains cells that can acquire a SM phenotype via
a transdifferentiation process that is transforming growth factor-␤1– and cell-cell contact– dependent, but
proliferation-independent. (Circ Res. 2002;90:1189-1196.)
Key Words: endothelial cell 䡲 transforming growth factor-␤ 䡲 smooth muscle myosin 䡲 mesenchymal cell
U
ntil recently, endothelial cells (ECs) and smooth muscle
cells (SMCs) were believed to arise from separate
precursor cells. However, experiments in a murine embryonic
stem cell model demonstrated that a common precursor cell
can differentiate toward either endothelial or SM lineage.1
Furthermore, a progenitor cell at different steps of a given
differentiation pathway can switch its phenotype, a phenomenon called transdifferentiation. Studies of embryonic avian
aortic and human pulmonary vascular development demonstrated that the endothelium itself could be the source of at
least some SMCs in the arteries.2–7 However, whether transdifferentiation of ECs into SMCs can occur in adult vasculature is not clear. Several in vivo and in vitro studies have
demonstrated that, in the adult, ECs can transform at least
into mesenchymal cells.8 –10 Yet, a more advanced differentiation of endothelial-derived mesenchymal cells into SMCs
has, to our knowledge, not been described.
We hypothesized that endothelial-mesenchymal transdifferentiation can take place in the adult vasculature, and that
endothelium-derived mesenchymal cells could further differentiate toward a SM phenotype. Utilizing an in vitro ap-
proach, we demonstrated that alterations in endothelial phenotype, indicative of transdifferentiation toward a SM
phenotype, occurred in the purified primary cultures of
bovine aortic and/or pulmonary artery endothelium. We then
estimated the time course of this process and sought mechanisms by which this phenomenon is controlled, including a
role of proliferation, transforming growth factor-␤1 (TGF-␤1),
and integrity of cell-cell contacts. Our data demonstrate that
mature bovine vascular endothelium contains cells that, in
vitro, can acquire a SMC phenotype via a transdifferentiation
process that is TGF-␤1– and cell-cell contact– dependent, but
proliferation-independent. The observation that endothelialsmooth muscle transdifferentiation can take place in primary
cell cultures derived from the adult vasculature suggests that
endothelium could serve as a potential source of at least some
SMCs in vascular neointimal lesions.
Materials and Methods
Isolation of Primary EC Cultures
Endothelium from adult bovine aortas and/or main pulmonary
arteries (n⫽37) was isolated as intact monolayers, using brief
Original received November 7, 2001; resubmission received March 15, 2002; revised resubmission received April 29, 2002; accepted April 29, 2002.
From the Developmental Lung Biology Research Laboratory, Department of Pediatrics, University of Colorado Health Sciences Center, Denver, Colo.
Correspondence to Maria Frid, Developmental Lung Biology Research Laboratory, Dept of Pediatrics, University of Colorado Health Sciences Center,
4200 E 9th Ave, Denver, CO 80262. E-mail [email protected]
© 2002 American Heart Association, Inc.
Circulation Research is available at http://www.circresaha.org
DOI: 10.1161/01.RES.0000021432.70309.28
1189
1190
Circulation Research
June 14, 2002
pretreatment with Dispase (Becton Dickinson) followed by a single
scrape with a flexible plastic scraper (Nalge/NUNC). Cell monolayers were plated onto at least 10 culture dishes (33 mm, diameter),
either plastic or covered with 1% denatured type I collagen (gelatin,
Sigma Chemical) in complete DMEM (Sigma) with 10% calf serum
(HyClone Laboratories). Primary cultures were confirmed to be free
of contamination with SMCs by thorough screening of at least 2
culture dishes subjected to immunostaining for SM markers. Contamination with subendothelial nonmuscle cells11 was ruled out
based on the following: first, when subendothelial cells are isolated
in primary culture, they appear as individual spindle-shaped cells
sparsely incorporated within extracellular matrix (Figure 3B), in
contrast to morphologically cobblestone endothelium isolated as
monolayers (Figure 3A). Second, nonmuscle cells in subendothelial
layer of bovine arteries do not express EC markers employed in this
study (not shown).
Antibodies and Cytokines
Downloaded from http://circres.ahajournals.org/ by guest on June 17, 2017
The following antibodies (Abs) were used: monoclonal antibodies
(mAbs) against ␣-SM-actin (clone 1A4) (1:100, Sigma), rabbit
anti–von Willebrand factor Abs (1:100, DAKO), goat anti–vascular
endothelial (VE)– cadherin, and rabbit anti–platelet-endothelial cell
adhesion molecule-1 (PECAM-1) (CD-31) Abs (1:10, Santa Cruz
Biotechnology). Anti-SM22␣ mAbs (clone 1B8) were a generous
gift of Dr S. Sartore (University of Padua, Italy). mAbs against SM
myosin heavy chains (SM-MHCs) and calponin were previously
described.11,12 Human recombinant TGF-␤1 and neutralizing chicken
Abs against active TGF-␤1 were from R&D Systems (No. AF-101NA, lot No. FS08). Neutralizing TGF-␤1 Abs were added every day
in fresh medium.
Fluorescence-Activated Cell Sorting (FACS) and
Immunomagnetic (Dynabeads) Purification Techniques
Primary EC cultures were metabolically labeled with DiI-acetylated
low-density lipoproteins (DiI-Ac-LDL) (Biomedical Technologies)
for 4 hours at 10 ␮g/mL or for 16 hours at 5 ␮g/mL13 and subjected
to FACS analysis (sorter model Moflow by Cytomation). Cell
fraction exhibiting elevated uptake of DiI-Ac-LDL was re-plated in
culture and subsequently re-sorted by FACS. As negative controls,
bovine SMCs and adventitial fibroblasts were used and have been
confirmed (as described elsewhere13) to be incapable of metabolizing
Ac-LDL (not shown).
For FACS-purification based on PECAM-1 expression, primary
(d1) ECs were dispersed with Accutase (Innovative Cell Technologies) and split in 2 portions to incubate one with anti–PECAM-1
antibodies and another with nonimmune rabbit IgG as a negative
control. After 2-hour incubation on ice, gently shaking, cells were
incubated with anti-rabbit IgG conjugated with Alexa-488 fluorescent dye (1:250, Molecular Probes) and subjected to FACS analysis.
Cell fraction positive for PECAM-1 expression (Figure 1, F-2) was
re-plated in culture. Purification of primary (d1) ECs was also
performed using superparamagnetic polystyrene beads (Dynabeads)
conjugated with anti–PECAM-1 mAbs (Dynal Biotech) as described
elsewhere.14
Immunofluorescence Analysis
Cells were fixed in 1% paraformaldehyde (1 minute), followed by
absolute methanol (⫺20°C, 10 minutes) and incubated with primary
antibodies for 1 hour at room temperature. Secondary antibodies
were either conjugated to Alexa-488 fluorescence dye (Molecular
Probes) or to biotin (DAKO). For the latter, streptavidin-Alexa-594
(Molecular Probes) was used. Stained cells were embedded in
VectaShield mounting medium with DAPI (Vector Laboratories) and
examined with a Nikon Optiphot epifluorescence microscope with
the DEI-470 Optronics imaging system.
Inhibition of Mitogenesis
Gamma (␥)-irradiation (30 Gy from a cesium source) and mitomycin
C treatment (2 ␮g/mL, 2 hours) were used to inhibit cell replication
in primary EC cultures maintained in DMEM with 10% serum.
Results
Primary EC Cultures Give Rise to SMCs
Primary (day 1) arterial EC cultures were first confirmed to
be free of contamination with SMCs and/or subendothelial
nonmuscle cells (see Materials and Methods). Cultures were
further purified by FACS and/or immunomagnetic (Dynabeads) techniques based on 2 endothelial-specific markers:
elevated uptake of DiI-Ac-LDL (FACS-based purification)
and/or PECAM-1 expression (both FACS and Dynabeads
immunopurification). In primary endothelial cultures metabolically labeled with DiI-Ac-LDL, the cell fraction exhibiting elevated DiI-Ac-LDL uptake was selected by FACS
(Figure 1A), re-plated onto several dishes, and subsequently
reassessed for DiI-Ac-LDL uptake at different days in culture. As shown in Figure 1, an EC fraction, selected as
positive for high DiI-Ac-LDL uptake (Figure 1A), over time
gave rise to cells incapable of metabolizing DiI-Ac-LDL
(Figures 1B through 1E). Based on sequential FACS analysis,
these latter cells were estimated to comprise approximately
0.01% to 0.03% of the total number of ECs on day 3 in
culture (Figure 1B), 0.09% on day 5 (not shown), 1.46% on
day 11 (Figure 1C), and 8.95% on day 17 (Figure 1D). By the
second passage in subculture, they constituted approximately
44% of the total number of cells (Figure 1E). The cells
incapable of DiI-Ac-LDL uptake appeared with time in
primary cultures as very elongated, spindle-shaped cells
forming a network on top of the EC monolayer (Figure 1G)
and, with time, piling on each other. Colonies of these
mesenchymal-like cells were selectively isolated using
Teflon cloning rings, as well as by FACS based on negative
DiI-Ac-LDL uptake, and assessed for expression of SM
markers by indirect immunofluorescence analysis. These data
demonstrated a progressive mesenchymal-to-smooth muscle
differentiation of these cells in culture. During the first days
after isolation, cells maintained their spindle morphology and
weakly expressed ␣-SM-actin in a diffuse, cytoplasmic pattern (Figure 2A). With time in culture, cells began to express
SM22␣, calponin, and, in some cells, SM-MHCs, all in a
diffuse cytoplasmic pattern (Figures 2C, 2E, and 2G). Progressively, the cells became larger, and all 4 SM proteins
were expressed in characteristic stress-fiber patterns (Figures
2B, 2D, 2F, and 2H). At this time, no endothelial-related
markers were detected.
Importantly, if primary EC cultures were allowed to
expand for more than 5 days and only then subjected to FACS
based on DiI-Ac-LDL uptake, the postsorted DiI-Ac-LDL–
positive EC fraction did not yield any mesenchymal cells.
This suggests that the process of endothelial-mesenchymal
transdifferentiation is initiated during first days after isolation
of endothelial monolayers in culture.
In addition, purification of primary EC cultures was
performed using another endothelial-specific marker,
PECAM-1, by both FACS and immunomagnetic (Dynabeads) sorting techniques. As stated, before utilizing these
techniques, primary EC cultures were confirmed to be free of
contamination with non-ECs by other methods. After FACS
and/or Dynabeads sorting based on PECAM-1 expression, the
obtained EC fractions were evaluated over time in culture.
Frid et al
Endothelium Can Give Rise to Smooth Muscle Cells
1191
Downloaded from http://circres.ahajournals.org/ by guest on June 17, 2017
Figure 1. Primary (day 1 after isolation) endothelial cultures, purified by FACS, over time
give rise to mesenchymal cells. FACSorting of
ECs was based on either elevated DiI-Ac-LDL
uptake (A through E, G), or positive PECAM-1
expression (F-2) (see Materials and Methods).
In FACSorting based on DiI-Ac-LDL uptake, the
cell fraction exhibiting high uptake of DiI-AcLDL (A) was re-plated in culture and subsequently analyzed for the capability to uptake
DiI-Ac-LDL over time in culture (B through E).
B, Day 5; C, Day 11; D, Day 17; and E, subculture at second passage. G, Mesenchymal-like
cells (marked by arrowheads), incapable of
Ac-LDL uptake (marked by an arrowhead in C),
appeared as very elongated spindle cells on
top of EC monolayer. In FACS purification
based on PECAM-1 expression, a PECAM-1–
positive postsorted cell fraction (F-2) was isolated, re-plated in culture, and assessed for
expression of SM markers over time. As a negative control, ECs were incubated with nonimmune rabbit IgG (F-1).
The results were similar to those obtained with DiI-Ac-LDL
FACS analysis. That is, primary ECs, selected on day 1 for
PECAM-1 expression by either FACS (Figure 1F) or Dynabeads (not shown), over time yielded mesenchymal cells,
which further differentiated into SMCs (similar to that shown
in Figure 2). Again, as mentioned, if EC purification based on
PECAM-1 expression was performed after day 5 in primary
culture, no mesenchymal or SMCs were observed.
Time Course of Acquisition of a SM Phenotype in
Primary EC Cultures
Endothelium was isolated as “intact” cell monolayers (Figure
3A and 3C), which were plated onto plastic culture dishes.
Endothelial monolayers attached within 18 to 24 hours
(Figure 3D) and acquired a typical cobblestone appearance.
ECs within monolayers were positive for all endothelial
markers tested (VE-cadherin, PECAM-1, von Willebrand
factor, high uptake of Ac-LDL) (not shown). No cells
expressing SM markers (␣-SM-actin, SM22␣, calponin, SMMHCs) were detected on day 1 in culture (Figure 3E, right
panel). The expression of EC and SMC markers in these
cultures was assessed daily.
On day 1 after isolation, cell-cell contacts in endothelial
monolayers appeared tight and were defined by vivid expression of VE-cadherin (Figure 3E, left panel). However, by day
2, cell-cell contacts became disrupted (Figure 3F, left panel),
and cells were observed that expressed ␣-SM-actin in a
diffuse, cytoplasmic pattern (Figures 3F and 3H, right panels). Notably, these cells still exhibited an epithelioid morphology similar to other ECs within the monolayer. Some
cells coexpressed ␣-SM-actin and endothelial markers (von
Willebrand factor antigen, VE-cadherin, and increased uptake
of Ac-LDL) (Figure 3F, VE-cadherin, others not shown). By
day 4 in culture, more ECs expressed ␣-SM-actin (Figure 3I).
By days 5 and 6, endothelial monolayers were no longer
compact, especially at the monolayer border (Figure 4A),
where many ECs had lost their rounded appearance and
acquired an elongated, mesenchymal-like morphology (Figures 3J, 4A, and 4B). These cells no longer expressed
VE-cadherin (Figure 4B), and some of them now expressed
1192
Circulation Research
June 14, 2002
the initial process of ␣-SM-actin acquisition appeared similar
to that in cultures plated on plastic. That is, during the first 2
to 3 days in culture, some cells within the monolayers began
to express ␣-SM-actin in a cytoplasmic pattern (not shown).
Again, certain cells coexpressing ␣-SM-actin and VEcadherin were observed. However, approximately by day 5,
cell monolayers had reestablished their cobblestone appearance with tight cell-cell contacts (Figure 4E), and very few
␣-SM-actin-positive cells were detected at that time. Over
time, endothelial monolayers grown on denatured collagen
did not yield spindle-shaped cells expressing ␣-SM-actin.
Acquisition of a SM Phenotype Occurs in the
Absence of Proliferation
Downloaded from http://circres.ahajournals.org/ by guest on June 17, 2017
Figure 2. Endothelial-derived mesenchymal cells undergo profound SM differentiation in culture. During the first days in subculture (passage 1 after selective isolation), these cells appeared
small and spindle-shaped and expressed some SM proteins in a
diffuse cytoplasmic pattern (A, C, E, and G). With time in culture
(7 to 10 days at the same passage), they became larger, and
SM proteins were expressed in characteristic stress-fiber patterns (B, D, F, and H). Similar results were obtained for primary
EC cultures purified by FACS an/or Dynabeads based on
PECAM-1 expression. A and B, ␣-SM-actin; C and D, SM22␣; E
and F, calponin; G and H, SM-MHCs.
␣-SM-actin in a diffuse, cytoplasmic pattern (Figures 3J and
4C). Expression of SM-MHCs was not noted at that time. By
days 11 to 14, ␣-SM-actin was expressed in a microfilamentous, stress-fiber pattern, and some ␣-SM-actin–positive cells
now expressed SM-MHCs in a diffuse cytoplasmic pattern
(similar to that shown in Figure 2D).
Phenotypic Differences Between EC Cultures
Plated Onto Plastic and/or Denatured Collagen
When the freshly isolated endothelial monolayers were plated
onto denatured collagen, rather than onto plastic, marked
differences were observed with regard to time-related
changes in cell morphology and expression of endothelial and
SM markers. In contrast to endothelial monolayers plated
onto plastic (described in “Time Course of Acquisition . . . ,”
Figures 4A through 4C), those plated onto denatured collagen
maintained over time a more compact appearance (Figure
4D) and more defined cell-cell contacts (Figure 4E). In these
cultures (also negative for SM marker expression on day 1),
To test whether the acquisition of a SM phenotype by select
ECs within the endothelial monolayers required DNA synthesis, day 1 EC cultures, maintained in DMEM with 10%
serum, were either ␥-irradiated (30 Gy) or treated with
mitomycin C. Inhibition of DNA synthesis was confirmed by
the absence of nuclear BrdU uptake (not shown). These
endothelial cultures, initially negative for ␣-SM-actin expression, by day 4 yielded cells that expressed ␣-SM-actin in a
diffuse cytoplasmic pattern (Figure 5B). By day 7, cells were
observed that exhibited an elongated, often stellate morphology and expressed ␣-SM-actin in a microfilamentous stressfiber pattern (Figure 5C). Cells that coexpressed both endothelial and SM-related markers (von Willebrand factor [vWf]
and ␣-SM-actin, respectively) were routinely detected by
double immunofluorescence labeling (Figures 5D through
5G). Some cells positive for ␣-SM-actin concurrently expressed SM-MHCs (not shown); however, no cells were
observed to concurrently express vWf and SM-MHCs.
TGF-␤ Is Involved in Inducing
Endothelial-Mesenchymal Transdifferentiation
We used a dual approach to test the hypothesis that TGF-␤1
may play an important role in inducing endothelialmesenchymal transdifferentiation in our system. First, primary EC cultures, plated onto denatured collagen (in which
case acquisition of SM phenotype was not normally observed), were treated with TGF-␤1 (0.25 to 1 ng/mL). Second,
primary EC cultures, plated onto plastic (in which case
acquisition of SM phenotype was normally observed), were
treated with neutralizing antibodies against active TGF-␤1.
In the first set of experiments, treatment of endothelial
monolayers plated on denatured collagen with TGF-␤1 resulted in profound disruption of cell-cell contacts. Cells
acquired a spindle-shaped morphology and, within the first
week in culture, some of them began to express ␣-SM-actin,
SM22␣, calponin, and SM-MHCs (not shown). Control
primary EC cultures, grown on denatured collagen, maintained a compact cobblestone appearance and did not yield
SMCs over time. As an additional control, a commercially
available CPAE endothelial cell line (American Type Culture
Collection, Manassas, Va), as well as ECs derived by late
(after day 5 in primary culture) FACS purification were
treated with 0.25 to 1 ng/mL of TGF-␤1. In these cell cultures,
␣-SM-actin was detected in some cells, whereas SM-MHC
Frid et al
Endothelium Can Give Rise to Smooth Muscle Cells
1193
Downloaded from http://circres.ahajournals.org/ by guest on June 17, 2017
Figure 3. Cells within “intact” endothelial monolayers can undergo endothelial-mesenchymal phenotypic switch in primary culture. A
through F, Loss of VE-cadherin and subsequent acquisition of ␣-SM-actin expression. A and C, Endothelium was isolated as intact
monolayers (shown floating in culture medium immediately after isolation). B, As a control, subendothelial cells were isolated by a
deeper, more intensive scrape; note, that they appear as individual cells of spindle-shaped or stellate morphology sparsely incorporated within the extracellular matrix. D, Within 24 hours of isolation, endothelial monolayers attached, spread, and acquired a typical
cobblestone appearance. E and F, Double-label immunofluorescence staining with anti–VE-cadherin (left) and anti–␣-SM-actin (right)
antibodies. E, Eighteen hours after isolation. Well-established cell-cell contacts are defined by VE-cadherin staining (left); ␣-SM-actin is
not expressed at this time (right). F, Two days after isolation. Cell-cell contacts are disrupted; VE-cadherin expression is lost in ECs
that detached from each other (left); several ECs now express ␣-SM-actin (right). Arrowheads mark cells that coexpress VE-cadherin
and ␣-SM-actin. G through J, Time course of acquisition of ␣-SM-actin expression by cells within endothelial monolayers in primary
culture. G, Day 1 after isolation; H, Day 2; I, Day 4; and J, Day 6. Left, DAPI staining showing all cell nuclei. Right, ␣-SM-actin expression. Note a peak of ␣-SM-actin expression around day 4 (I). Within endothelial monolayers, cells that begin to express ␣-SM-actin
around day 2 maintain a rounded morphology, and ␣-SM-actin is expressed in a cytoplasmic pattern. By day 6, ␣-SM-actin–positive
cells alter their morphological appearance toward an elongated, spindle-shaped form (J); at this time, ␣-SM-actin is still expressed in a
cytoplasmic pattern. Photomicrographs G and I are taken with a 10⫻ objective, whereas H and J, with a 20⫻ objective.
expression was not detected even after a 20-day treatment
(not shown; other SM proteins were not assessed).
In a second set of experiments, ECs plated onto plastic were
incubated with neutralizing TGF-␤1 antibodies (Abs) (5 and 10
␮g/mL). As expected, untreated endothelial monolayers grown
on plastic over time gave rise to spindle-shaped cells expressing
SM proteins in a stress-fiber pattern (Figure 6A, ␣-SM-actin is
shown). In cultures treated with neutralizing TGF-␤1 Abs at 5
␮g/mL, ␣-SM-actin–positive cells were still observed, yet
␣-SM-actin was expressed in a cytoplasmic pattern, and cells
maintained an epithelioid morphology over time (Figure 6B).
Transition to spindle-shaped morphology with a stress-fiber
pattern of ␣-SM-actin did not occur, and SM-MHC expression
was not detected in these cultures (SM22␣ and calponin were
not assessed). In endothelial monolayers treated with 10 ␮g/mL
of neutralizing Abs, no ␣-SM-actin–positive cells were observed
over time (Figure 6C). Similar strategy was used for ␥-irradiated
(30 Gy) primary EC cultures (data not shown). Treatment with
neutralizing Abs (10 ␮g/mL) suppressed appearance of ␣-SMactin–positive cells, whereas untreated ␥-irradiated primary cultures yielded ␣-SM-actin–positive cells similar to that described
(not shown).
1194
Circulation Research
June 14, 2002
Downloaded from http://circres.ahajournals.org/ by guest on June 17, 2017
Figure 4. Marked phenotypic differences are observed between primary endothelial cultures plated onto plastic (left, A through C) and
those plated onto denatured collagen (right, D through F) (day 7 cultures are shown). Photomicrographs B and C and E and F are the
images of the same field of double-label cultures shown in (A) and (D),
respectively. A through C, Endothelial monolayer integrity is gradually
lost when endothelium is cultured on plastic (A); ECs lose their initially
tight (see Figure 3E, left) cell-cell contacts as defined by the loss of
VE-cadherin (B); some ECs now express ␣-SM-actin (C). D through F,
Endothelial monolayers plated onto denatured collagen maintain a
more compact cobblestone appearance (D) and more defined cell-cell
contacts (marked by VE-cadherin expression [E]); ␣-SM-actin expression is not detected (F).
Discussion
Endothelial-to-mesenchymal transdifferentiation is increasingly
being viewed as an important biological process in development
and disease. Studies of embryonic avian aortic and human
pulmonary vascular development have demonstrated that vascular endothelial cells could undergo an endothelial-mesenchymal
phenotypic switch and migrate into the subendothelial media,
suggesting that at least some SMCs in the embryonic arteries
arise from the endothelium.2–7 Such phenotypic “plasticity”
could be maintained by at least some ECs in the neonatal period,
as recently demonstrated in human umbilical vein ECs
(HUVECs) that were shown capable of transdifferentiating into
beating cardiomyocytes.15 Whether any cells within the mature
vascular endothelium retain this capability has remained unclear.
Interestingly, this possibility was suggested nearly 60 years ago
by R. Altschul16 who, in histological studies of human atherosclerosis, described that some aortic ECs underwent an endothelial-mesenchymal phenotypic switch and invaded the subendothelial space. This concept was recently revived in speculations
that endothelial-derived mesenchymal cells may be involved in
development of neointimal lesions in transplant atherosclerosis
and restenosis.17–19 In addition, transdifferentiation of ECs into
mesenchymal cells in the adult is suggested by findings in
Figure 5. Acquisition of SM phenotype by select cells within the
endothelial monolayers occurs in the absence of mitogenesis.
Endothelial monolayers, maintained in DMEM with 10% CS,
were ␥-irradiated (30 Gy) on day 1 after isolation. Lack of proliferation was confirmed by the absence of BrdU incorporation
(not shown). A through C, left, DAPI staining showing all cell
nuclei. A through C, right, ␣-SM-actin expression. A, ECs within
endothelial monolayers do not express ␣-SM-actin on day 1
after isolation. B, By day 4, some ECs express ␣-SM-actin in a
diffuse cytoplasmic pattern (note that these cells still maintain
rounded, epithelioid morphology). C, By day 7, cells were
observed that had an elongated, often stellate morphology and
expressed ␣-SM-actin in a microfilamentous, stress-fiber pattern. D through G, Coexpression of an endothelial marker, von
Willebrand factor (vWf), and SM-related marker, ␣-SM-actin, in
␥-irradiated primary endothelial cultures (day 7 after isolation).
D, DAPI staining showing all cell nuclei; E, ␣-SM-actin; F, vWf;
and G, an overlay of images (E and F) demonstrates the cell
(indicated by an arrow) that coexpresses vWf and ␣-SM-actin.
experimental wound repair where capillary ECs converted into
interstitial connective tissue cells in granulation tissue and by
findings that microvascular ECs transdifferentiated into mesenchymal spindle-shaped cells on chronic inflammatory stimuli.8,20 –23 Further, in mature ECs, the capacity for transdifferentiation is supported by findings that adult bovine aortic ECs,
when treated with TGF-␤1, convert to spindle-shaped cells
expressing ␣-SM actin,9 and that certain murine endothelial-like
cell lines irreversibly transform into mesenchymal cells on
overgrowth in culture.10 Our findings complement and extend
these previous observations and provide rigorous in vitro evi-
Frid et al
Endothelium Can Give Rise to Smooth Muscle Cells
Downloaded from http://circres.ahajournals.org/ by guest on June 17, 2017
Figure 6. TGF-␤–neutralizing antibodies (Abs) suppress acquisition
of ␣-SM-actin in primary EC cultures in a dose-dependent manner.
Left, DAPI staining showing all cell nuclei. Right, ␣-SM-actin
expression (day 10). A, Untreated primary EC cultures yield cells
that express ␣-SM-actin and acquire an elongated, spindleshaped, or stellate morphology. B, Primary EC cultures treated
with neutralizing TGF-␤1 Abs at 5 ␮g/mL yield few cells that
express ␣-SM-actin in a cytoplasmic pattern and maintain epithelioid morphology. C, No ␣-SM-actin expression is observed in EC
cultures treated with 10 ␮g/mL of neutralizing TGF-␤1 Abs.
dence that mature vascular endothelium contains cells that have
the potential to undergo endothelial–smooth muscle
transdifferentiation.
An important and novel observation of the present study was that
primary EC cultures confirmed to be free of contamination with
non-ECs and further purified by FACS and/or magnetic beads
(Dynabeads) techniques (based on elevated uptake of DiI-Ac-LDL
and/or PECAM-1 expression) gave rise to mesenchymal and even
SM cells. Both FACS and Dynabeads purifications were performed
soon after initiating primary culture, ie, on day 1 after EC isolation.
When the cell sorting was performed on day 5 after cell isolation or
later, no mesenchymal cells were observed to appear in these
cultures because the endothelial-mesenchymal phenotypic switch
had probably already occurred and the mesenchymal cells had been
selectively sorted out. FACS and Dynabeads sorting techniques are
commonly used by investigators to obtain pure endothelial cultures
(ie, to specifically eliminate mesenchymal “contaminant”); however, they are usually applied after cells expand in culture for some
time. In studies of microvascular endothelial cells, sorting techniques are often applied at the time of cell isolation; however, the
need for subsequent sorting is noted as an essential requirement
for maintaining purity of EC cultures.13,14,24 This is an interesting
detail because, for example, dermal microvascular ECs are
described as very susceptible to transdifferentiation into mesenchymal cells.21 Thus, our data would support the assumption that
the mesenchymal cells, often observed in primary EC cultures,
are not the result of contamination but are often the result of
transdifferentiation.
Additional experimental evidence supports our contention that
the appearance of SM-like cells in primary endothelial cultures is
due to transdifferentiation. First, the primary endothelial monolayers
were used for analysis, which displayed typical cobblestone morphology and were composed of endothelial cells only (see Results).
1195
Several endothelial-specific markers were expressed by these cells,
specifically, VE-cadherin was expressed by all cells within the
monolayer (Figure 3E, left). No staining for SM-related markers
was detected at day 1 after isolation. The presence of subendothelial
nonmuscle cells was also ruled out (these cells, isolated as controls
by a deeper scrape, appeared as individual cells of spindle-shaped
morphology sparsely incorporated within the extracellular matrix
(Figure 3B) and did not express endothelial markers (not shown).
Furthermore, “transitional” cells coexpressing both endothelial
(VE-cadherin, von Willebrand factor) and SM-related (␣-SM-actin)
markers were consistently observed. Endothelial-mesenchymal
phenotypic switch was accompanied by changes in cell shape
toward mesenchymal, spindle-shaped, which continued to evolve as
the cells expressed SM phenotype. Finally, the transdifferentiation
process could be experimentally manipulated. It was attenuated by
maintaining primary ECs on denatured collagen, which could be
reversed by addition of TGF-␤1. Moreover, the process was facilitated by plating primary EC cultures on plastic, which was inhibited
by addition of neutralizing TGF-␤1 antibodies. Importantly, studies
in epithelial cells have pointed out the requirement for TGF-␤ for a
change in the matrix environment, for a breakdown of E-cadherin
expression with subsequent alteration of cell morphology, and for
increased expression of ␣-SM-actin.25–32 Other studies have emphasized that myogenesis is controlled by the cell’s shape.33 Thus, it
seems likely that endothelial- and epithelial-to-mesenchymal transdifferentiation processes share many common molecular
mechanisms.
Our findings are the first, to our knowledge, to describe the
acquisition of a more differentiated SM phenotype (defined by
expression of SM-MHCs rather than just that of ␣-SM-actin) in
the course of endothelial–smooth muscle transdifferentiation in
the cells from mature vessels. The data of the present study
suggest a critical role for TGF-␤1 in the transdifferentiation of
ECs. Although TGF-␤1 is known to promote ␣-SM-actin expression in nonmuscle cells (ECs and fibroblasts derived from
various tissues9,34), it is currently thought not to be sufficient to
induce expression of late SM differentiation markers, such as
SM-MHCs, in cells of non-SMC lineage.34 Thus, it is possible
that in our system TGF-␤1 is necessary for initiation of transdifferentiation, yet it may not be sufficient for driving the process
to the higher SM differentiation level described (defined by
expression of SM-MHCs). If so, the other factors/mechanisms
operating to cause transdifferentiation of ECs into differentiated
SMCs need to be identified.
Because only a small percentage (0.01% to 0.03%) of cells
within the endothelium appear capable of undergoing endothelial–smooth muscle transdifferentiation, questions arise regarding the origin of these cells. In the adult, bone marrow– derived
progenitor cells from circulating blood can incorporate into
endothelial monolayer, differentiate into ECs, and constitute a
significant proportion of cells within the vascular endothelium.35,36 Furthermore, bone marrow– derived progenitor cells
have been described that have the potential to differentiate into
endothelial/mesothelial and SM-like cells.37,38 Recently,
Kaushal et al39 have demonstrated that acellular vascular grafts
seeded with blood-borne endothelial progenitor cells later become populated with contractile SMCs, thus raising a possibility
of gradual transformation of endothelial progenitor cells to
vascular ECs to medial SMCs. Taken together, these observa-
1196
Circulation Research
June 14, 2002
tions and our data raise the possibility that endothelial-like cells
within the mature endothelium, which are capable of transdifferentiating into SMCs, may have originated from a circulating
pool of bone marrow– derived endothelial precursors. If this
were the case, mature vascular endothelium, having a constant
supply of blood-borne multipotent endothelial-like cells, could
itself serve as a potential source of at least some SM-like cells in
vascular neointimal lesions.
Acknowledgments
This study was supported by NIH grants SCOR HL-57144, PPG
HL-14985. The authors gratefully acknowledge Stephen Hofmeister
for technical assistance, Dr John T. Reeves for valuable discussion,
and Marcia McGowan for help in preparation of the manuscript.
References
Downloaded from http://circres.ahajournals.org/ by guest on June 17, 2017
1. Yamashita J, Itoh H, Hirashima M, Ogawa M, Nishikawa S, Yurugi T,
Naito M, Nakao K, Nishikawa S-I. Flk1–positive cells derived from
embryonic stem cells serve as vascular progenitors. Nature. 2000;408:
92–96.
2. Arciniegas E, Servin M, Arguello C, Mota M. Development of the aorta in
the chick embryo: structural and ultrastructural study. Atherosclerosis. 1989;
76:219–235.
3. DeRuiter MC, Poelmann RE, VanMunsteren JC, Mironov V, Markwald RR,
Gittenberger-deGrott AC. Embryonic endothelial cells transdifferentiate into
mesenchymal cells expressing smooth muscle actins in vivo and in vitro. Circ
Res. 1997;80:444–451.
4. Hall SM, Hislop AA, Pierce CM, Haworth SG. Prenatal origins of human
intrapulmonary arteries: formation and smooth muscle maturation. Am J
Respir Cell Mol Biol. 2000;23:194–203.
5. Majesky MW, Schwartz SM. An origin for smooth muscle cells from endothelium? Circ Res. 1997;80:601–603.
6. Arciniegas E, Ponce L, Hartt Y, Graterol A, Carlini RG. Intimal thickening
involves transdifferentiation of embryonic endothelial cells. Anat Rec. 2000;
258:47–57.
7. De Reeder EG, Girard N, Poelmann RE, Van Munstern JC, Patterson DF,
Gittenberger-DeGroot AC. Hyaluronic acid accumulation and endothelial
cell detachment in intimal thickening of the vessel wall: the normal and
genetically defective ductus arteriosus. Am J Pathol. 1988;132:574–585.
8. Sarkisov DS, Kolokolchikova EG, Kaem RI, Paltsyn AA. Vascular changes
in maturing granulation tissue. Bull Exp Biol Med. 1988;105:604–605.
9. Arciniegas E, Sutton AB, Allen TD, Schor AM. Transforming growth factor
␤1 promotes the differentiation of endothelial cells into smooth muscle-like
cells in vitro. J Cell Sci. 1992;103:521–529.
10. Kohler A, Jostarndt-Fogen A, Rottener K, Alliegro MC, Draeger A.
Intima-like smooth muscle cells: developmental link between endothelium
and media? Anat Embryol. 1999;200:313–323.
11. Frid MG, Moiseeva EP, Stenmark KR. Multiple phenotypically distinct
smooth muscle cell populations exist in the adult and developing bovine
pulmonary arterial media in vivo. Circ Res. 1994;75:669– 681.
12. Frid MG, Shekhonin BV, Koteliansky VE, Glukhova MA. Phenotypic
changes of human smooth muscle cells during development: late expression
of heavy caldesmon and calponin. Dev Biol. 1992;153:185–193.
13. Voyta JC, Via DP, Butterfield CE, Zetter BR. Identification and isolation of
endothelial cells based on their increased uptake of acetylated-low density
lipoprotein. J Cell Biol. 1984;99:2034–2040.
14. Hewett PW, Murray JC. Immunomagnetic purification of human microvascular endothelial cells using Dynabeads coated with monoclonal antibodies to PECAM-1. Eur J Cell Biol. 1993;62:451–454.
15. Condorelli G, Borello U, De Angelis L, Latronico M, Sirabella D, Coletta M,
Galli R, Balconi G, Follenzi A, Frati G, Cusella De Angelis MG, Gioglio L,
Amuchastegui S, Adorini L, Naldini L, Vescovi A, Dejana E, Cossu G.
Cardiomyocytes induce endothelial cells to trans-differentiate into cardiac
muscle: implications for myocardium regeneration. Proc Natl Acad Sci
U S A. 2001;98:10733–10738.
16. Altschul R. Histologic analysis of arteriosclerosis. Arch Pathol. 1944;38:
305–312.
17. Beranek JT, Cavarocchi NC. Undifferentiated vascular endothelial cells in
coronary allograft atherosclerosis. Internatl J Cardiol. 1990;28:127–128.
18. Beranek JT. Vascular endothelium-derived cells containing smooth muscle
actin are present in restenosis. Lab Invest. 1995;72:771.
19. Mironov YA, Hill CH, Starcher B, Shin JCH, Hoffman S, Markwald RR.
Endothelial-mesenchymal transformation in atherogenesis: a recapitulation of
embryonic heart cushion tissue morphogenesis. Ann Biomed Eng. 1995;
23(suppl 1):S29A. Abstract.
20. Lipton B, Bensch KG, Karasek MA. Histamine-modulated transdifferentiation of dermal microvascular endothelial cells. Exp Cell Res. 1992;199:
279–291.
21. Romero LI, Zhang DN, Herron GS, Karasek MA. Interleukin-1 induces
major phenotypic changes in human skin microvascular endothelial cells.
J Cell Phyiol. 1997;173:84–92.
22. Karasek MA. Origin of spindle-shaped cells in Kaposi sarcoma. Lymphology.
1994;27:41–44.
23. Fan JM, Huang XR, Ng YY, Nicolic-Paterson DJ, Mu W, Atkins RC, Lan
HY. Interleukin-1 induces tubular epithelial-myofibroblast transdifferentiation through a transforming growth factor-␤1–dependent mechanism in
vitro. Am J Kidney Dis. 2001;37:820–831.
24. Springhorn JP, Madri JA, Squinto SP. Human capillary endothelial cells from
abdominal wall adipose tissue: isolation using an anti-PECAM antibody. In
Vitro Cell Dev Biol Anim. 1995;31:473–481.
25. Bhowmick NA, Ghiassi M, Bakin A, Aakre M, Lundquist CA, Engel ME,
Arteaga CL, Moses HL. Transforming growth factor-␤1 mediates epithelial
to mesenchymal transdifferentiation through a RhoA-dependent mechanism.
Mol Biol Cell. 2001;12:27–36.
26. Brown CB, Boyer AS, Runyan RB, Barnett JV. Requirement of type III
TGF-␤ receptor for endocardial cell transformation in the heart. Science.
1999;283:2080–2082.
27. Paranya G, Vineberg S, Dvorin E, Kaushal S, Roth SJ, Rabkin E, Schoen FJ,
Bischoff J. Aortic valve endothelial cells undergo transforming growth
factor-␤–mediated and non-transforming growth factor-␤–mediated transdifferentiation in vitro. Am J Pathol. 2001;159:1335–1343.
28. Stahl PJ, Felsen D. Transforming growth factor-␤, basement membrane, and
epithelial-mesenchymal transdifferentiation. Am J Pathol. 2001;159:
1187–1192.
29. Hay ED. An overview of epithelio-mesenchymal transformation. Acta Anat.
1995;154:8–20.
30. Landerholm TE, Dong X-R, Lu J, Belaguli NS, Schwartz RJ, Majesky MW.
A role for serum response factor in coronary smooth muscle differentiation
from proepicardial cells. Development. 1999;126:2053–2062.
31. Miettinen PJ, Ebner R, Lopez AR, Derynck R. TGF-␤ induced transdifferentiation of mammary epithelial cells to mesenchymal cells: involvement of
type I receptors. J Cell Biol. 1994;127:2021–2036.
32. Markwald R, Eisenberg C, Eisenberg L, Trusk T, Sugi Y. Epithelialmesenchymal transformations in early avian heart development. Acta Anat.
1996;156:173–186.
33. Yang Y, Relan NK, Przywara DA, Schuger L. Embryonic mesenchymal cells
share the potential for smooth muscle differentiation: myogenesis is controlled by the cell’s shape. Develoment. 1999;126:3027–3033.
34. Hautmann MB, Adam PJ, Owens GK. Similarities and differences in smooth
muscle ␣-actin induction by TGF-␤ in smooth muscle versus non-smooth
muscle cells. Arterioscler Thromb Vasc Biol. 1999;19:2049–2058.
35. Crosby JR, Kaminski WE, Schatteman G, Martin PJ, Raines EW, Seifert RA,
Bowen-Pope DF. Endothelial cells of hematopoietic origin make a significant
contribution to adult blood vessel formation. Circ Res. 2000;87:728–730.
36. Gunsilius E, Duba H-C, Petzer AL, Kahler CM, Grunewald K, Stockhammer
G, Gabl C, Dirnhofer S, Clausen J, Gastl G. Evidence from a leukaemia
model for maintenance of vascular endothelium by bone-marrow-derived
endothelial cells. Lancet. 2000;355:1688–1691.
37. Shimizu K, Sugiyama S, Aikawa M, Fukumoto Y, Rabkin E, Libby P,
Mitchell RN. Host bone-marrow cells are a source of donor intimal smooth
muscle-like cells in the murine aortic transplant arteriopathy. Nat Med.
2001;7:738–741.
38. Campbell JH, Efendy JL, Han C-L, Campbell GR. Blood vessels from bone
marrow. Ann N Y Acad Sci. 2000;902:224–229.
39. Kaushal S, Amiel GE, Guleserian KJ, Shapira OM, Perry T, Sutherland FW, Rabkin E, Moran AM, Schoen FJ, Atala A, Soker S,
Bischoff J, Mayer JE Jr. Functional small-diameter neovessels created
using endothelial progenitor cells expanded ex vivo. Nat Med. 2001;
7:1035–1040.
Mature Vascular Endothelium Can Give Rise to Smooth Muscle Cells via
Endothelial-Mesenchymal Transdifferentiation: In Vitro Analysis
Maria G. Frid, Vishakha A. Kale and Kurt R. Stenmark
Downloaded from http://circres.ahajournals.org/ by guest on June 17, 2017
Circ Res. 2002;90:1189-1196; originally published online May 9, 2002;
doi: 10.1161/01.RES.0000021432.70309.28
Circulation Research is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 2002 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7330. Online ISSN: 1524-4571
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circres.ahajournals.org/content/90/11/1189
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Circulation Research can be obtained via RightsLink, a service of the Copyright Clearance Center, not the
Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further information
about this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation Research is online at:
http://circres.ahajournals.org//subscriptions/