Human breast microvascular endothelial cells retain phenotypic traits in long-term finite life span culture.
ABSTRACT Attempts to study endothelial-epithelial interactions in the human breast have been hampered by lack of protocols for long-term cultivation of breast endothelial cells (BRENCs). The aim of this study was to establish long-term cultures of BRENCs and to compare their phenotypic traits with the tissue of origin. Microvasculature was localized in situ by immunohistochemistry in breast samples. From this tissue, collagen-rich stroma and adipose tissue were dissected mechanically and further disaggregated to release microvessel organoids. BRENCs were cultured from these organoids in endothelial specific medium and characterized by staining for endothelial markers. Microvessels were a prominent feature of intralobular tissue as evidenced by immunostaining against endothelial specific markers such as CD31, VE-cadherin, and von Willebrand factor (VWF). Double staining against VE-cadherin and lymphatic vessel endothelial hyaluronan receptor-1 (LYVE-1) showed that blood and lymphatic vessels could be distinguished. An antibody against CD31 was used to refine protocols for isolation of microvasculature from reduction mammoplasties. BRENCs retained critical traits even at high passage, including uptake of low-density lipoprotein, and had E-selectin induced upon treatment with tumor necrosis factor-alpha. The first signs of senescence in passage 14 were accompanied by gain of trisomy 11. At passage 18 cells showed chromosomal aberrations and growth arrest as revealed by beta-galactosidase staining. We demonstrate here that breast microvasculature may serve as a large-scale source for expansion of BRENCs with molecular and functional traits preserved. These cells will form the basis for studies on the role of endothelial cells in breast morphogenesis.
- SourceAvailable from: René Villadsen[Show abstract] [Hide abstract]
ABSTRACT: Epithelial to mesenchymal transition (EMT) is a critical event in cancer progression and is closely linked to the breast epithelial cancer stem cell phenotype. Given the close interaction between the vascular endothelium and cancer cells, especially at the invasive front, we asked whether endothelial cells might play a role in EMT. Using a 3D culture model we demonstrate that endothelial cells are potent inducers of EMT in D492 an immortalized breast epithelial cell line with stem cell properties. Endothelial induced mesenchymal-like cells (D492M) derived from D492, show reduced expression of keratins, a switch from E-Cadherin (E-Cad) to N-Cadherin (N-Cad) and enhanced migration. Acquisition of cancer stem cell associated characteristics like increased CD44(high)/CD24(low) ratio, resistance to apoptosis and anchorage independent growth was also seen in D492M cells. Endothelial induced EMT in D492 was partially blocked by inhibition of HGF signaling. Basal-like breast cancer, a vascular rich cancer with stem cell properties and adverse prognosis has been linked with EMT. We immunostained several basal-like breast cancer samples for endothelial and EMT markers. Cancer cells close to the vascular rich areas show no or decreased expression of E-Cad and increased N-Cad expression suggesting EMT. Collectively, we have shown in a 3D culture model that endothelial cells are potent inducers of EMT in breast epithelial cells with stem cell properties. Furthermore, we demonstrate that basal-like breast cancer contains cells with an EMT phenotype, most prominently close to vascular rich areas of these tumors. We conclude that endothelial cells are potent inducers of EMT and may play a role in progression of basal-like breast cancer.PLoS ONE 01/2011; 6(9):e23833. · 3.53 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: BACKGROUND: Development of epithelial organs depends on interaction between the epithelium and the underlying mesenchyme including the vasculature. The aim of this study was to explore the morphogenic effect of endothelial cells on prostate epithelial cell lines in 3D culture and to establish an in vitro model for prostate branching morphogenesis. METHODS: A panel of eleven cell lines originating in normal or malignant prostate and primary prostate epithelial cells were cultured in reconstituted basement membrane (rBM) matrix with or without non-proliferating but metabolically active endothelial cells. Morphogenesis was evaluated by phase contrast microscopy and further characterized by immunocyto/histocemistry and confocal microscopy. RESULTS: Endothelial cells induced clonogenic potential of most prostate cell lines and formation of branching and mesenchymal-like colonies. One of the normal-derived cell lines in the panel (PZ-HPV-7) displayed unique properties in rBM culture by forming large and complex branching structures resembling the ductal architecture of the prostate. This ability was highly dependent on epithelial seeding density and soluble factors derived from the endothelial cells. High seeding density suppressed branching of PZ-HPV-7 but survival was compromised at low density in the absence of endothelium. CONCLUSIONS: We have generated an endothelial-based clonogenic assay to study prostate epithelial morphogenesis in three-dimensional context. This assay will be important tool to study prostate epithelial-endothelial interactions in 3D context and open up possibilities to study molecular regulation of prostate morphogenesis and cancer progression. Prostate © 2012 Wiley Periodicals, Inc.The Prostate 12/2012; · 3.84 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Branching morphogenesis is a mechanism used by many species for organogenesis and tissue maintenance. Receptor tyrosine kinases (RTKs), including epidermal growth factor receptor (EGFR) and the sprouty protein family are believed to be critical regulators of branching morphogenesis. The aim of this study was to analyze the expression of Sprouty-2 (SPRY2) in the mammary gland and study its role in branching morphogenesis. Human breast epithelial cells, breast tissue and mouse mammary glands were used for expression studies using immunoblotting, real rime PCR and immunohistochemistry. Knockdown of SPRY2 in the breast epithelial stem cell line D492 was done by lentiviral transduction of shRNA constructs targeting SPRY2. Three dimensional culture of D492 with or without endothelial cells was done in reconstituted basement membrane matrix. We show that in the human breast, SPRY2 is predominantly expressed in the luminal epithelial cells of both ducts and lobuli. In the mouse mammary gland, SPRY2 expression is low or absent in the virgin state, while in the pregnant mammary gland SPRY2 is expressed at branching epithelial buds with increased expression during lactation. This expression pattern is closely associated with the activation of the EGFR pathway. Using D492 which generates branching structures in three-dimensional (3D) culture, we show that SPRY2 expression is low during initiation of branching with subsequent increase throughout the branching process. Immunostaining locates expression of phosphorylated SPRY2 and EGFR at the tip of lobular-like, branching ends. SPRY2 knockdown (KD) resulted in increased migration, increased pERK and larger and more complex branching structures indicating a loss of negative feedback control during branching morphogenesis. In D492 co-cultures with endothelial cells, D492 SPRY2 KD generates spindle-like colonies that bear hallmarks of epithelial to mesenchymal transition. These data indicate that SPRY2 is an important regulator of branching morphogenesis and epithelial to mesenchymal transition in the mammary gland.PLoS ONE 01/2013; 8(4):e60798. · 3.53 Impact Factor
In Vitro Cell. Dev. Biol.—Animal 42:332–340, November/December 2006
? 2006 Society for In Vitro Biology
HUMAN BREAST MICROVASCULAR ENDOTHELIAL CELLS RETAIN PHENOTYPIC TRAITS
IN LONG-TERM FINITE LIFE SPAN CULTURE
VALGARDUR SIGURDSSON, AGLA J. R. FRIDRIKSDOTTIR, JENS KJARTANSSON, JON G. JONASSON,
MARGRET STEINARSDOTTIR, OLE WILLIAM PETERSEN, HELGA M. OGMUNDSDOTTIR, AND THORARINN GUDJONSSON1
Faculty of Medicine, University of Iceland (V. S., H. M. O., T. G.), Molecular and Cell Biology Research Laboratory, Icelandic Cancer
Society (V. S., A. J. R. F., H. M. O., T. G.), Reykjavik, Iceland, Structural Cell Biology Unit, Department of Medical Anatomy-A, The
Panum Institute, University of Copenhagen, Denmark (A. J. R. F., O. W. P.), St. Josefs Hospital, Iceland (J. K.), Department of
Pathology, University Hospital of Iceland (J. G. J.), Chromosome Laboratory, Department of Genetics and Molecular Medicine, University
Hospital of Iceland (M. S.)
(Received 25 April 2006; accepted 26 April 2006)
Attempts to study endothelial–epithelial interactions in the human breast have been hampered by lack of protocols
for long-term cultivation of breast endothelial cells (BRENCs). The aim of this study was to establish long-term cultures
of BRENCs and to compare their phenotypic traits with the tissue of origin. Microvasculature was localized in situ by
immunohistochemistry in breast samples. From this tissue, collagen-rich stroma and adipose tissue were dissected me-
chanically and further disaggregated to release microvessel organoids. BRENCs were cultured from these organoids in
endothelial specific medium and characterized by staining for endothelial markers. Microvessels were a prominent feature
of intralobular tissue as evidenced by immunostaining against endothelial specific markers such as CD31, VE-cadherin,
and von Willebrand factor (VWF). Double staining against VE-cadherin and lymphatic vessel endothelial hyaluronan
receptor-1 (LYVE-1) showed that blood and lymphatic vessels could be distinguished. An antibody against CD31 was
used to refine protocols for isolation of microvasculature from reduction mammoplasties. BRENCs retained critical traits
even at high passage, including uptake of low-density lipoprotein, and had E-selectin induced upon treatment with tumor
necrosis factor-?. The first signs of senescence in passage 14 were accompanied by gain of trisomy 11. At passage 18
cells showed chromosomal aberrations and growth arrest as revealed by ?-galactosidase staining. We demonstrate here
that breast microvasculature may serve as a large-scale source for expansion of BRENCs with molecular and functional
traits preserved. These cells will form the basis for studies on the role of endothelial cells in breast morphogenesis.
Key words: breast endothelial cells; isolation; cultivation; characterization
Blood vessels perfuse all tissues in the body and mediate meta-
bolic exchanges between tissues and blood. Furthermore, recent
data have demonstrated that blood vessels participate in embryo
development and tissue morphogenesis (Shekhar et al., 2000; Lam-
mert et al., 2001, 2003; Cleaver and Melton, 2003; Shen et al.,
2004). Although endothelial cells from different organs share many
morphological and functional features, subtleties at both levels have
been shown to be organ specific (McCarthy et al., 1991; Jackson
and Nguyen, 1997; Belloni and Nicolson, 1988; Bachetti and Mor-
bidelli, 2000; Bouis et al., 2001; Chi et al., 2003). In particular the
microenvironment seems to determine form and function as most
convincingly revealed in the blood–brain barrier (Wagner and Ri-
sau, 1994; Abbott, 2002).
The human breast gland is a good example of a dynamic organ
composed of a branching epithelium surrounded by vascularized
stroma. Indeed, the stroma accounts for more than 80% of the rest-
ing breast volume (Rønnov-Jessen et al., 1996) and collectively is
composed of cellular components such as fibroblasts, immune cells,
1To whom correspondence should be addressed.
and fat cells in addition to vascular-derived endothelial cells and
smooth muscle cells embedded in a collagen-rich extracellular ma-
trix. While blood vessels and particularly angiogenesis have re-
ceived considerable attention in relation to breast cancer (Gasparini
and Harris, 1995; Boudreau and Myers, 2003) the structure and
function of the vasculature in the normal human breast are poorly
understood. Recently, Naccarato et al. (2003) demonstrated that
morphological differences exist between intralobular and extralob-
ular vasculature. Whereas ducts were surrounded by many small
capillaries, lobules showed fewer but larger microvessels with a
sinusoidal phenotype. A similar pattern has been found by other
(Rønnov-Jessen et al., 1996).
Most of the information on human endothelial cells in culture is
derived from experiments with human umbilical vein endothelial
cells (HUVECs) since these are readily accessible (Bouis et al.,
2001). In general, HUVECs have been regarded as representative
also of microvascular endothelial cells (Manconi et al., 2000). While
this may be safely assumed for a number of investigations, when
studying cellular interactions at the level of individual organs the
tissue specificity must be considered. This was elegantly demon-
strated by the recent gene expression profiling of endothelial cells
BREAST ENDOTHELIAL CELLS
with (A) CD31, (B) VE-cadherin, (C) CD34, (D) LYVE-1. Note intralobular staining with CD31, VE-cadherin, and LYVE-1. CD34 stain
stromal cells broadly. Sections were counterstained with hematoxylin. Bar ? 50 ?m. Figure is published in color online at http://
Microvessels are a prominent feature of TDLU in the human breast. Cryostat sections of normal human breast tissue stained
of different tissue origins (Pasqualini et al., 2002; Pasqualini and
Arap, 2002; Chi et al., 2003).
There are only few published reports of in vitro studies of normal
BRENCs (Hewett et al., 1992; Hewett and Murray, 1993b; O’Hare
et al., 2001), mainly because of limited access to tissue material as
well as the lack of protocols for isolation and cultivation. Methods
for the isolation and culture of BRENCs have been regarded as
extremely laborious and time consuming. Hewett et al. (1993) es-
tablished a method for the isolation of BRENCs by sequential di-
gestion of the breast fat tissue with collagenase and trypsin followed
by specific selection of microvessel fragments with Ulex europaeus
agglutinin-1 (UEA)–coated magnetic beads. A modification of this
method included replacement of UEA with CD31 (PECAM) (Hewett
and Murray, 1993b). Rønnov-Jessen and Petersen (1993) estab-
lished a method for the isolation of different cellular compartments
of the human breast including microvessel organoids. The enriched
microvessel compartment is mostly derived from collagen-rich stro-
ma and provides an alternative source to adipose tissue for further
purification of BRENCs. With regard to increasing data from other
organs demonstrating that endothelial cells play an active role in
organogenesis and cell differentiation (Lammert et al., 2001, 2003;
Matsumoto et al., 2001; Shen et al., 2004), it is important to im-
prove conditions for both isolation and long-term culture of
BRENCs and to study them in the normal context of breast mor-
phogenesis. The aim of this study was to establish long-term cul-
tures of BRENCs and to compare the endothelial phenotype in cul-
ture with their counterpart in situ.
MATERIALS AND METHODS
Materials. Breast tissue specimens where obtained from reduction mam-
moplasties after informed consent. The study was approved by the Data Pro-
tection Commission and the National Bioethics Committee in Iceland (per-
mission number 99/111). Culture dishes were from Nunc (Roskilde, Den-
mark) and Chamber Slides from BD (Bedford, MA). Vitrogen, was from Co-
hesion Technologies (Palo Alto, CA). Growth factor–reduced matrigel was
obtained from Becton Dickinson (Bedford, MA). Collagenase IA was obtained
from Sigma, Brondby, Denmark (C-9891). Cell culture medium, Dulbecco
modified Eagle medium (DMEM)/F12, and fetal calf serum was obtained from
GIBCO (GIBCO BRL, Life Technologies, Grand Island, NY) and endothelial
growth medium, EGM-2 from Cambrex (Cambrex Bio Science, Walkersville,
MD). Anti-CD31 Dynabeads where obtained from Dynal Biotech (Dynal Bio-
tech ASA, Oslo, Norway). MACS columns and microbeads where purchased
from Miltenyi (Miltenyi Biotec GmbH, Gladbach, Germany).
Primary cell culture and isolation. Primary breast endothelial cells were
isolated from interstitial stroma or adipose tissue from breast biopsies as
described in more detail in the result section. Briefly, epithelial tissue and
collagen-rich interstitial stroma were separated from adipose tissue and
minced into small pieces and digested in DMEM/F12 medium supplemented
with collagenase-1A (900 U/ml) overnight at 37? C on a rotary shaker (60
SIGURDSSON ET AL.
staining with VE-cadherin (green) and Thy-1 (red). Blood vessels were clearly distinguishable from fibroblasts and epithelial cells by
expression of VE-cadherin. Fibroblasts and myoepithelial cells were recognized by their expression of Thy-1. (B) Double staining against
VE-cadherin (green) and CD31 (red) in endothelial cells shows colocalization. (C) Blood and lymphatic endothelial cells were identified
by expression of VE-cadherin (red) and LYVE-1 (green), respectively. (D) Extralobular expression of LYVE-1 (green). Nuclear staining
(blue). Bar ? 50 ?m. Figure is published in color online at http://inva.allenpress.com/invaonline/?request?index.html.
Blood and lymph endothelial cells stain differently in TDLUs of normal human breast tissue. (A) Immunofluorescence double
rpm). After enzymatic disaggregation the digest was differentially centrifu-
gated as described by others (Rønnov-Jessen and Petersen, 1993). This re-
sulted in four different components enriched in acini and ducts, large vessels,
microvessels, and fibroblasts. The microvessel pellet was washed twice with
5 ml incubation buffer (phosphate-buffered saline [PBS]/0.1% bovine serum
albumin [BSA]/2 mM ethylenediaminetetraacetic acid [EDTA]) and incubated
with anti-CD31 Dynabeads for 20 min at 4? C. Microvessel organoids bound
to the Dynabeads were isolated on a magnetic concentrator. Microvascular
organoids from adipose tissue were isolated as described by Hewett et al.
(1993). Briefly, adipose tissue was collagenase-treated for 2 h at 37? C. The
crude digest was centrifuged and microvessel organoids were isolated using
anti-CD31 Dynabeads. All breast endothelial organoids were seeded on vi-
trogen-coated T25 culture flasks and cultured in EGM-2 supplemented with
30% fetal bovine serum (FBS). Serum concentration could be reduced to 2–
5% after the first passage for short-term culture. When needed the MACS
cell sorting system was used for the selection of contaminating fibroblasts.
Anti–Thy-1 antibody (ASO2, Dianova, Hamburg, Germany) was used to de-
Immunocytochemistry and confocal microscopy. Breast biopsies were frozen
in n-hexan (Merck) and mounted in tissue freezing medium (Leica instru-
ments) for sectioning. Frozen biopsies were sectioned at a 5-?m setting in a
cryostat. The sections were dried for 15 min at room temperature and fixed
in methanol as described previously (Petersen and van Deurs, 1988). Primary
antibodies included anti-CD31 (JC70A, DakoCytomation, Denmark), anti-
VE-cadherin (BV9, abcam, Cambridge, UK), anti-CD34 (QBEnd/10, Novo-
castra, Newcastle, UK), anti-LYVE-1 (ab10278, abcam), anti-Thy1 (ASO2,
Dianova), anti-VWF (F8/86), anti-VEGFR2 (KDR/EIC, abcam), anti-CD105
(SN6h, NeoMarkers, Fremont, CA), anti-vimentin (V9, DakoCytomation), and
anti-keratin K19 (BA17, DakoCytomation). Rabbit anti-mouse immunoglob-
ulins (Z0259, DakoCytomation) were used as secondary antibodies and a
peroxidase conjugated anti-peroxidase mouse mAb was used as a tertiary
antibody (P850, DakoCytomation). The peroxidase reactions were performed
using 0.5 mg/ml 3,3-diaminobenzidine (DakoCytomation) and 0.02% H2O2
(Merck) for 10 min. The cultures were counterstained with hematoxylin. For
double-labelling experiments we used fluorescent isotype-specific secondary
antibodies (Molecular Probes, Invitrogen). Antibody incubations were carried
out for 30 min, and specimens were rinsed twice for 5 min each at room
temperature. Fluorescent nuclear counterstaining was performed with TO-
PRO-3 iodide (Molecular Probes, Invitrogen, Groco, Reykiacik, Iceland). Af-
ter staining specimens were mounted with coverslips using Fluoromount-G
(Southern Biotechnology). Immunofluorescence was visualized using a Zeiss
LSM 5 Pascal laser-scanning microscope (Carl Zeiss).
Functional assays. Acetylated low-density lipoprotein (AcLDL) uptake as-
say was performed on semiconfluent breast endothelial cells cultured on four-
BREAST ENDOTHELIAL CELLS
endothelial cells. Tissue from breast reduction mammoplasties was minced and separated into an adipose component and an epithelial/
stromal component. Each component was digested with collagenase for 4 and 24 h, respectively. Digestion of adipose tissue resulted in
relatively pure microvessel organoids. In contrast, digestion of epithelial/stromal component resulted in a mixture of cells and organoids.
This could be further purified into microvasculature by differential centrifugation. The microvasculature from either component was
incubated with anti-CD31 Dynabeads and isolated on a magnetic concentrator. This was seeded into collagen-coated culture flasks and
cultured in the EGM-2 plus serum. Maximum propagation was achieved when cells were cultured in 30% serum. For short-term culture
serum concentration could be lowered to 2–5 %, however, BRENCs did not grow without serum supplement. Figure is published in color
online at http://inva.allenpress.com/invaonline/?request?index.html.
Schematic presentation of the human breast including the TDLU and a protocol for the isolation of breast blood vessel
SIGURDSSON ET AL.
CD31 (green), (B) VWF (green), (C) VE-cadherin (green), (D) LYVE-1 (green). Breast endothelial cells in culture express classical
endothelial markers but not the lymphatic endothelial marker. Nuclear staining (red). Bar ? 50 ?m. Figure is published in color online
Characterization of breast endothelial cells in culture. Primary cultures of human breast endothelial cells stained with (A)
well chamber slides. Conjugated AcLDL-A488 was added at 10 ?g/ml to the
cells and the cells were then incubated for 4 h. The cultures were then
washed twice with PBS and fixed with 1% formalin solution. Chamber slides
were mounted with coverslips and AcLDL uptake visualized using a Zeiss
laser-scanning microscope. E-selectin expression was induced with tumor
necrosis growth factor-? (20 ng/ml) for 4 h on subconfluent breast endothelial
cells cultured on four-well chamber slides. Subsequently, the slides were
fixed with methanol and incubated with anti–E-selectin (ELAM) antibody
(Research Diagnostic clone 1.2B6, Concord, MA) for 30 min at room tem-
perature. Rabbit anti-mouse immunoglobulins were used as secondary anti-
bodies as described above. Breast endothelial tube formation assay was per-
formed on matrigel coated four-well chamber slides. Each well was coated
with 50 ?l matrigel and incubated at 37? C for 30 min. Endothelial cells
were trypsinized and seeded into the wells and formation of capillary-like
net was observed and photographed after 24 h of culture.
Senescence assay. Staining for senescence-associated ?-galactosidase (SA-
?-gal) was performed according to the manufacturers instructions(senescence
?-galactosidase staining kit, Cell Signalling Technology, Danvers, MA). Brief-
ly, cells were cultured on T25 cell culture flasks until semiconfluent. The
cells were then fixed and incubated with the staining solution, which con-
tained X-gal. Senescent cells showed ?-galactosidase activity at pH 6, which
was detected as a blue color. This procedure was used on primary breast
endothelial cells from three different individuals in passage 6, 13, and 18.
For comparison HUVECs from three different individuals were also stained
with this procedure.
Karyotype analysis. Karyotype analysis was performed at the Chromosome
Laboratory at the Landspitali University Hospital, using standard cytogenetic
procedures. Briefly, cells were incubated with MAS (metaphase arresting so-
lution, Genial Genetic Solutions, Ltd., Cheshire, UK) for 3 h, followed by
hypotonic treatment (0.0075 M KCl) for 20 min at 37? C and fixed with
methanol/acetic acid (1:3). Slides were aged for 2 d and G-banded with
trypsin solution and Leishmans stain. From each culture 30 cells were ana-
lyzed and karyotypes described following International system for human
cytogenetics (ISCN) recommendations (Mitelman, 1995).
Characterization of breast endothelial cells in situ. We first
mapped the spatial localization of the blood vasculature within the
terminal duct lobular unit (TDLU) using classical endothelial-spe-
cific markers. These include CD31, VE-cadherin, and VWF, which
are all expressed on different blood vessels. To discriminate be-
tween blood and lymphatic vessels we used an antibody against
LYVE-1 that is highly specific for lymphatic vessels (Fig. 1). When
stained with antibodies against CD31 and VE-cadherin, a promi-
nent intralobular expression was seen (Fig. 1A,B) and it was evident
that BRENCs were in close contact with the epithelial compartment.
BREAST ENDOTHELIAL CELLS
CHARACTERISTICS OF BREAST ENDOTHELIAL CELLS IN CULTURE
BRENCsb6BRENCs 13BRENCs 18
Tube formation on matrigel
Karyotype Normal NormalAbnormal
avWF, von Willebrand factor; SA-?-gal, senescence-associated ?-galacto-
sidase; LYVE-1, lymphatic vessel endothelial hyalvronan receptor; AcLDL,
acetylated low-density lipoprotein; VEGFR2, vascular endothelial growth fac-
tor receptor 2; Ck-18, cytokeratin-18; ND, not done.
bBRENCs, breast endothelial cells.
CD34 has been extensively used as a marker for endothelial cells
in situ. However, in our hands this marker was expressed on a wide
variety of other cell types such as fibroblasts and myoepithelial cells
(Fig. 1C, additional data not shown). Localization of lymphatic ves-
sels was studied by staining for LYVE-1, which is exclusively ex-
pressed on lymphatic vessels. The observed expression indicated
that intralobular lymphatic vessels were also a prominent feature of
the TDLU (Fig. 1D). The intralobular organization of blood and
lymphatic vessels was further demonstrated with immunofluores-
cence double labeling (Fig. 2). Double staining against VE-cadherin
and Thy-1 revealed a distinction between the vascular and the fi-
broblast/epithelial compartment (Fig. 2A). Double staining of VE-
cadherin with CD31 showed colocalization (Fig. 2B). To distinguish
between blood and lymphatic vascular cells we performed double
staining against VE-cadherin and LYVE-1 (Fig. 2C). Figure 2D
shows extralobular expression of LYVE-1 in a large lymphatic ves-
sel. In summary, the intimate relationship between the endothelial
and epithelial compartments inside the TDLU suggests the possi-
bility of a reciprocal cellular interaction across these compartments
in the normal breast gland.
Isolation and cultivation of breast endothelial cells. Breast micro-
vasculature was obtained by extraction from the collagen-rich in-
terstitial stroma or the surrounding adipose tissue with a modifi-
cation of methods published by Rønnov-Jessen et al. (1996),
Rønnov-Jessen and Petersen (1993), and Hewett et al. (1993), also
described in Material and Methods and outlined in Fig. 3. Modifi-
cations from previous methods included immunomagnetic purifica-
tion in a CD31 (PECAM) retaining column followed by a flow-
through in an anti–Thy-1 column in order to remove contaminating
fibroblasts. Enriched microvessel organoids were plated onto col-
lagen-coated tissue culture flasks and cultivated in the EGM-2 me-
dium (Fig. 3, schematic figure). This medium was further supple-
mented with 30% fetal calf serum which was indispensable for long-
term endothelial growth. Under these culture conditions primary
BRENCs were kept up to 18 passages with a split ratio of 1:4 at
each passage. For short-term culture purposes (4–6 passages) the
serum level could be reduced to 2% (Fig. 3). The success rate in
establishing long-term BRENCs cultures in terms of biopsies was
9 out of 35 (25.7%). This allowed for cryopreservation of several
vials from each biopsy for future purposes. A high viability after
cryopreservation was observed if cultures were provided with fresh
medium before handling and 55% fetal calf serum was added as a
Characterization of cultured BRENCs. BRENC cultures were
characterized in passage 6, 13, and 18. As seen in Fig. 4 and Table
1 BRENCs stained brightly with the endothelial markers CD31,
VWF, and VE-cadherin but were negative for the lymphatic marker
LYVE-1, the fibroblast marker Thy-1, and the epithelial cytokeratin
18. The endothelial lineage of the cultured BRENCs was further
evidenced by their ability to form capillary-like networks on top of
matrigel (Fig. 5A,B). Furthermore, BRENCs showed uptake of the
endothelial specific marker AcLDL when added to the culture me-
dium, and addition of tumor necrosis factor-? resulted in the in-
duction of another endothelial specific marker, E-selectin. Table 1
summarizes the marker expression and functional characteristics of
BRENCs remain karyotypically normal until the onset of senes-
cence. Karyotyping revealed that BRENCs were remarkably stable
even after long-term culture. Generally, at passage 14 a subpopu-
lation of BRENCs emerged exhibiting signs of senescence as evi-
denced by expression of SA-?-gal (data not shown). This was ac-
companied by chromosomal changes in terms of trisomy 11. For
comparison with BRENCs we used HUVECs and these started to
accumulate chromosomal changes already in the sixth passage with
the same split ratio. Interestingly, the same chromosomal abnor-
malities, i.e., trisomy 11, appeared in BRENCs and HUVECs. Fig-
ure 6 shows the karyotype of BRENCs and HUVECs at passage 13.
At passage 18 BRENCs expressed a widespread senescent pheno-
type as evidenced by the appearance of large cells with fragmented
nuclei and expression of SA-?-gal. Nevertheless, BRENCs retained
the ability to form tube-like structures even at the onset of senes-
cence after passage 14 (data not shown) which indicates that this
behavior is independent of cell growth.
We have characterized long-term cultures of human BRENCs and
compared their phenotypic traits to the tissue of origin. Further-
more, we optimized protocols for isolation of BRENCs from reduc-
tion mammoplasty. When compared to endothelial cells in situ, cul-
tured BRENCs clearly retained critical phenotypic traits until pas-
sage 14 when signs of senescence were seen concurrently with the
appearance of chromosomal instability. BRENCs were completely
senescent at passage 18.
The first successful isolation of human microvessel endothelial
cells was reported in 1979 when endothelial cells from adrenal
cortex, neonatal foreskin, and spleen were grown in vitro (Folkman
et al., 1979). Since then, human microvessel endothelial cells have
been isolated and cultured from many different organs such as the
brain (Dorovini-Zis et al., 1991; Lamszus et al., 1999), dermis (Dav-
ison et al., 1983; Jackson et al., 1990; Richard et al., 1998), lung
SIGURDSSON ET AL.
(A) Phase contrast micrograph of BRENCs on matrigel. Note the formation of capillary-like structures. (B) Immunofluorescence staining
of BRENCs in endothelial cells cultured on matrigel with CD31 (green). Nuclear staining (blue). (C) Monolayer culture of BRENCs
incubated with fluorescein isothiocyanate (FITC)-labelled AcLDL (green). Note the endothelial specific uptake of AcLDL (D) Monolayer
culture of BRENCs induced with tumor necrosis factor-? and stained with E-selectin immunoperoxidase. Bar ? 50 ?m. Figure is published
in color online at http://inva.allenpress.com/invaonline/?request?index.html.
Breast endothelial cells recapitulate a microvascular pattern on matrigel and retain functional properties in monolayer culture.
(Hewett and Murray, 1993a), bone marrow (Richard et al., 1998),
intestine (Haraldsen et al., 1995), and breast adipose tissue (Hewett
et al., 1993) However, most of these methods are time consuming
and laborious thus limiting their application.
Long-term BRENCs cultures were successfully established from
26% of the biopsies. One reason for this relatively low success rate
could be the phase of the menstrual cycle of each patient when
undergoing the reduction mammoplasty. It has been shown that the
vascular endothelial growth factor (VEGF) and insulin-like growth
factor-1 (IGF-1) levels increase in the normal breast tissue in the
luteal phase compared with the follicular phase of the menstrual
cycle which may provide a pro-angiogenic environment in the luteal
phase (Dabrosin, 2003a, 2003b).
The fact that we observe trisomy on chromosome 11 in passage
6 in HUVECs and passage 14 in BRENCs is in line with data from
other groups (Nichols et al., 1987; Johnson et al., 1992). Nichols
et al. (1987) serially subcultivated endothelial cells from adult hu-
man arteries and veins and compared them to endothelial cells from
umbilical cord; polyploidy, including trisomy 11, was only seen in
HUVECs. Johnson et al. (1992) demonstrated trisomy 11 in all but
one of 12 adult EC cultures. Interestingly, a number of oncogenes
are present on chromosome 11 including Ha-ras and the B cell
lymphoma 1 (bcl.1) (Johnson et al., 1992). Because the development
of polyploidy is a sign of senescence, it is possible that the early
appearance of polyploidy and trisomy 11 in HUVECs is related to
the fact that these cells are derived from a tissue that is at the end
of its in vivo life span when studied. The ability of BRENCs to
retain genotypic and phenotypic stability for up to 14 passages in
culture will allow the use of these cells for studies on their role in
normal development and morphogenesis of the human breast gland.
BREAST ENDOTHELIAL CELLS
left) and HUVECs (upper right) in passage 13. Trisomy 11 is seen in passage 13 in HUVECs but not in BRENCs. Phase contrast images
of BRENCs (lower left) and HUVECs (lower right) in passage 13. Bar ? 50 ?m. Figure is published in color online at http://
Breast endothelial cells show chromosomal stability even after long-term propagation in culture. Karyotypes of BRENCs (upper
The data presented here show that BRENCs retain characteristic
in situ markers in culture and that cultivation can be achieved for
up to 13 passages without any significant phenotypical or chromo-
somal changes. The ability to culture BRENCs provides the possi-
bility to create ‘‘designer’’ microenvironments in vitro, incorporating
different cell types for the study of mutual interactions during nor-
mal morphogenesis (Bissell et al., 2002). Such studies are currently
in progress in our laboratory.
Competing interests. The authors declare that they have no com-
Authors’ contributions. V. S. and A. F. performed cell isolation,
cell cultures and immunohistochemistry, J. K. and J. G. J. provided
access to tissue material and were consultants on this project in-
cluding proofreading of the article. M. S. carried out karyotype anal-
ysis, H. M. O. and O. W. P. contributed by critical reading of the
article and interpretation of data. T. G. coordinated the study and
was responsible for supervision of the laboratory work and writing
the article. All authors read and approved the final manuscript.
This work was supported by a Research Grant of Excellence from the
Icelandic Research Council, The Research Fund of the University of Iceland,
The Memorial Fund of Ingibjorg Gudjonsdottir Johnson and the Post-doctoral
Research Grants, from the University of Iceland and the Icelandic Research
Council (T. G.) as well as the European Commission Research Directorates
(O. W. P., contract HPRN-CT-2002-00246).
Abbott, N. J. Astrocyte-endothelial interactions and blood-brain barrier per-
meability. J. Anat. 200:629–638; 2002.
Bachetti, T.; Morbidelli, L. Endothelial cells in culture: a model for studying
vascular functions. Pharmacol. Res. 42:9–19; 2000.
Belloni, P. N.; Nicolson, G. L. Differential expression of cell surface glyco-
proteins on various organ-derived microvascular endothelia and en-
dothelial cell cultures. J. Cell. Physiol. 136:398–410; 1988.
Bissell, M. J.; Radisky, D. C.; Rizki, A.; Weaver, V. M.; Petersen, O. W. The
organizing principle: microenvironmental influences in the normal
and malignant breast. Differentiation 70:537–546; 2002.
Boudreau, N.; Myers, C. Breast cancer-induced angiogenesis: multiple mech-
anisms and the role of the microenvironment. Breast Cancer Res.
SIGURDSSON ET AL.
Bouis, D.; Hospers, G. A.; Meijer, C.; Molema, G.; Mulder, N. H. Endothelium
in vitro: a review of human vascular endothelial cell lines for blood
vessel-related research. Angiogenesis 4:91–102; 2001.
Chi, J. T.; Chang, H. Y.; Haraldsen, G., et al. Endothelial cell diversity
revealed by global expression profiling. Proc. Natl. Acad. Sci. USA
Cleaver, O.; Melton, D. A. Endothelial signaling during development. Nat.
Med. 9:661–668; 2003.
Dabrosin, C. Increase of free insulin-like growth factor-1 in normal human
breast in vivo late in the menstrual cycle. Breast Cancer Res. Treat.
Dabrosin, C. Variability of vascular endothelial growth factor in normal hu-
man breast tissue in vivo during the menstrual cycle. J. Clin. Endo-
crinol. Metab. 88:2695–2698; 2003b.
Davison, P. M.; Bensch, K.; Karasek, M. A. Isolation and long-term serial
cultivation of endothelial cells from the microvessels of the adult
human dermis. In Vitro 19:937–945; 1983.
Dorovini-Zis, K.; Prameya, R.; Bowman, P. D. Culture and characterization
of microvascular endothelial cells derived from human brain. Lab.
Invest. 64:425–436; 1991.
Folkman, J.; Haudenschild, C. C.; Zetter, B. R. Long-term culture of capillary
endothelial cells. Proc. Natl. Acad. Sci. USA 76:5217–5221; 1979.
Gasparini, G.; Harris, A. L. Clinical importance of the determination of tumor
angiogenesis in breast carcinoma: much more than a new prognostic
tool. J. Clin. Oncol. 13:765–782; 1995.
Haraldsen, G.; Rugtveit, J.; Kvale, D.; Scholz, T.; Muller, W. A.; Hovig, T.;
Brandtzaeg, P. Isolation and longterm culture of human intestinal mi-
crovascular endothelial cells. Gut 37:225–234; 1995.
Hewett, P. W.; Murray, J. C. Human lung microvessel endothelial cells: iso-
lation, culture, and characterization. Microvasc. Res. 46:89–102;
Hewett, P. W.; Murray, J. C. Human microvessel endothelial cells: isolation,
culture and characterization. In Vitro Cell. Dev. Biol. 29A:823–830;
Hewett, P.; Murray, J.; Price, E.; Watts, M.; Woodcock, M. Isolation and
characterization of microvessel endothelial cells from human mam-
mary adipose tissue. In Vitro Cell. Dev. Biol. 29A:325–331; 1992.
Hewett, P. W.; Murray, J. C.; Price, E. A.; Watts, M. E.; Woodcock, M. Iso-
lation and characterization of microvessel endothelial cells from hu-
man mammaryadipose tissue.
Jackson, C. J.; Garbett, P. K.; Nissen, B.; Schrieber, L. Binding of human
endothelium to Ulex europaeus I-coated Dynabeads: application to
96(Pt 2):257–262; 1990.
Jackson, C. J.; Nguyen, M. Human microvascular endothelial cells differ from
macrovascular endothelial cells in their expression of matrix metal-
loproteinases. Int. J. Biochem. Cell Biol. 29:1167–1177; 1997.
Johnson, T. E.; Umbenhauer, D. R.; Hill, R.; Bradt, C.; Mueller, S. N.; Levine,
E. M.; Nichols, W. W. Karyotypic and phenotypic changes during in
vitro aging of human endothelial cells. J. Cell Physiol. 150:17–27;
In VitroCell. Dev.Biol.
Lammert, E.; Cleaver, O.; Melton, D. Induction of pancreatic differentiation
by signals from blood vessels. Science 294:564–567; 2001.
Lammert, E.; Cleaver, O.; Melton, D. Role of endothelial cells in early pan-
creas and liver development. Mech. Dev. 120:59–64; 2003.
Lamszus, K.; Schmidt, N. O.; Ergun, S.; Westphal, M. Isolation and culture
of human neuromicrovascular endothelial cells for the study of an-
giogenesis in vitro. J. Neurosci. Res. 55:370–381; 1999.
Manconi, F.; Markham, R.; Fraser, I. S. Culturing endothelial cells of micro-
vascular origin. Methods Cell Sci. 22:89–99; 2000.
Matsumoto, K.; Yoshitomi, H.; Rossant, J.; Zaret, K. S. Liver organogenesis
promoted by endothelial cells prior to vascular function. Science
McCarthy, S. A.; Kuzu, I.; Gatter, K. C.; Bicknell, R. Heterogeneity of the
endothelial cell and its role in organ preference of tumour metastasis.
Trends Pharmacol. Sci. 12:462–467; 1991.
Mitelman, F. An international system for human cytogenetic nomenclature.
ISCN, S. Karger, Basel, 1995.
Naccarato, A. G.; Viacava, P.; Bocci, G.; Fanelli, G.; Aretini, P.; Lonobile,
A.; Montruccoli, G.; Bevilacqua, G. Definition of the microvascular
pattern of the normal human adult mammary gland. J. Anat.
Nichols, W. W.; Buynak, E. B.; Bradt, C.; Hill, R.; Aronson, M.; Jarrell, B.
E.; Mueller, S. N.; Levine, E. M. Cytogenetic evaluation of human
endothelial cell cultures. J. Cell. Physiol. 132:453–462; 1987.
O’Hare, M. J.; Bond, J.; Clarke, C., et al. Conditional immortalization of
freshly isolated human mammary fibroblasts and endothelial cells.
Proc. Natl. Acad. Sci. USA 98:646–651; 2001.
Pasqualini, R.; Arap, W. Profiling the molecular diversity of blood vessels.
Cold Spring Harbor Symp. Quant. Biol. 67:223–225; 2002.
Pasqualini, R.; Arap, W.; McDonald, D. M. Probing the structural and mo-
lecular diversity of tumor vasculature. Trends Mol. Med. 8:563–71;
Petersen, O. W.; van Deurs, B. Growth factor control of myoepithelial-cell
differentiation in cultures of human mammary gland. Differentiation
Richard, L.; Velasco, P.; Detmar, M. A simple immunomagnetic protocol for
the selective isolation and long-term culture of human dermal micro-
vascular endothelial cells. Exp. Cell Res. 240:1–6; 1998.
Rønnov-Jessen, L.; Petersen, O. W. Induction of alpha-smooth muscle actin
by transforming growth factor-beta 1 in quiescent human breast gland
fibroblasts. Implications for myofibroblast generation in breast neo-
plasia. Lab. Invest. 68:696–707; 1993.
Rønnov-Jessen, L.; Petersen, O. W.; Bissell, M. J. Cellular changes involved
in conversion of normal to malignant breast: importance of the stromal
reaction. Physiol. Rev. 76:69–125; 1996.
Shekhar, M. P.; Werdell, J.; Tait, L. Interaction with endothelial cells is a
prerequisite for branching ductal-alveolar morphogenesis and hyper-
plasia of preneoplastic human breast epithelial cells: regulation by
estrogen. Cancer Res. 60:439–449; 2000.
Shen, Q.; Goderie, S. K.; Jin, L.; Karanth, N.; Sun, Y.; Abramova, N.; Vincent,
P.; Pumiglia, K.; Temple, S. Endothelial cells stimulate self-renewal
Wagner, E. F.; Risau, W. Oncogenes in the study of endothelial cell growth
and differentiation. Semin. Cancer Biol. 5:137–145; 1994.
of neuralstem cells. Science