| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Department of Pathological Sciences, Medical School, University of Manchester, Manchester M13 9PT, U.K.;
St. Mary’s Hospital, Manchester M13 0JH, U.K.;
Paterson Institute for Cancer Research, Manchester M20 9BX, U.K.; and
§ CSIC, Velazquez 144, 28006, Madrid, Spain.
1Correspondence: Department of Pathological Sciences, Stopford Building, Medical School, University of Manchester, Manchester M13 9PT, U.K. E-mail: MDDPSCL2{at}FS1.SCG.MAN.AC.UK
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: angiogenesis • TGFß1 • HUVEC
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2
3)
. Although the regulatory mechanisms in angiogenesis
are not fully understood, evidence suggests that a cascade of events
occurs and that intervention at one of the steps is sufficient to
prevent or promote neovascularization.
The transforming growth factor ß (TGFß) family of growth factors
includes some of the most potent pleiotropic angiogenic factors known,
which exert biological effects through interaction with heteromeric
transmembrane serine/threonine kinase receptor complexes (types I and
II) (4
5
6)
. Ligand binding to the type II receptor
(TßRII) promotes heteromeric type I/II complex formation and allows
type II to phosphorylate the type I receptor. This activates Smad
proteins, which transduce signals to the nucleus (7
, 8)
.
Ligand-receptor access is finely controlled by several proteins.
For example, membrane betaglycan plays a role in presenting TGFß to
the signaling receptors whereas the soluble form of this receptor is an
antagonist of TGFß effects (9)
. Betaglycan is absent in
human umbilical vein endothelial cells whereas CD105, which shares a
71% sequence similarity in the transmembrane and intracellular regions
with betaglycan, is highly expressed by these cells (10)
.
A striking feature of CD105 is that it is constitutively phosphorylated
in vascular endothelial cells.
CD105 is a 180 kDa integral membrane glycoprotein that binds with high
affinity to TGFß1 and TGFß3 (kD =
50 pM) (10)
. It has been localized to human chromosome 9
(11)
and is the target gene for hereditary hemorrhagic
telangiectasia type I (12)
. Although CD105 is
overexpressed in endothelial cells of healing wounds, embryos,
infarcted tissues, psoriatic skin, synovial tissues of rheumatoid
arthritis, and a wide range of solid tumors (13
14
15
16
17
18
19)
, its
function in modulating endothelial responses to TGFß remains to be
established. Furthermore, recent work has shown that the reactivity of
CD105 in blood vessels of breast cancer tissues correlated with poor
prognosis (20)
, and elevated levels of soluble CD105 and
CD105/TGFß complexes in the plasma of breast cancer patients have
been linked to increased prevalence of metastatic disease
(21
; our unpublished data). These observations suggest
that CD105 is involved in the angiogenic process through interaction
with its ligand. Overexpression of CD105 in cell lines (e.g.,
myoblasts) had previously been reported to modulate their responses to
TGFß1 (22)
. More importantly, lack of CD105 in knockout
mice leads to defective vascularization and death by 11.5 p.c. day
(23)
.
To clarify the relationship between CD105 and TGFß in the control of neovascularization, we have used antisense oligodeoxynucleotides (AS ODN) to specifically block CD105 protein translation in human umbilical vein endothelial cells (HUVECs) and then examined the TGFß1-mediated responses of these cells.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Tissue culture plastic ware (Costar, Cambridge,
Calif.) used for growing HUVECs was precoated with 1% (w/v) gelatin
for 1 h at 37°C, then washed twice with phosphate-buffered
saline (PBS). HUVECs were cultured in medium supplemented with 20%
(v/v) fetal calf serum (FCS) (Life Technologies, Inc., Grand Island,
N.Y.) and confluent cells were subcultured using 0.05% (w/v) trypsin
and 0.02% (w/v) EDTA in PBS. Cells used in all the experiments did not
exceed passage 3 and were identified using antibodies to the specific
endothelial cell markers von Willebrand factor (DAKO, Glostrup,
Denmark) and CD31 (PECAM-1) (DAKO). A collection of cells from five
different umbilical cords was used to prevent the possible
interindividual variation.
Phosphorothioate-modified oligodeoxynucleotides
The sequence of the 16-mer AS ODN was 5'-ATGCTGTCCACGTGGG-3',
which was designed to target the initiation codon AUG on the CD105
mRNA. A scrambled control (SC ODN) having the same base composition
(5'-ACTCGTGCTACGGTGG-3') was generated by randomizing the bases of the
antisense sequence. GenEMBL homology searches (FASTA and BLAST) were
carried out and revealed no relation to any other known human genes.
Phosphorothioate-modified oligodeoxynucleotides were synthesized on an
automated DNA synthesizer (Applied Biosystems, Foster City, Calif.) and
purified by high-performance liquid chromatography.
HUVECs were treated with ODNs as follows: 5.6 µg lipofectACE (Life Technologies) was mixed with 90 µl serum-free DMEM (Life Technologies). ODN was added to the mixture at designated concentrations and incubated at room temperature for 15 min before transferring into subconfluent HUVEC cultures in medium containing 5% FCS. The cells were incubated at 37°C with daily replenishment of both the media and reagents. At the end of 72 h treatment, the cultures were gently washed with prewarmed PBS and used for further study.
Indirect immunofluorescence and flow cytometric analysis
Cell surface expression of CD105 protein was quantified by flow
cytometry as described previously (25)
. Briefly,
105 cells per tube were incubated with 50 µl
(10 µg/ml in PBS) of monoclonal antibody to CD105 (Mab E9)
(24)
or preimmunized mouse serum as negative control
antibody (10 µg/ml in PBS) on ice for 1 h and washed twice with
cold PBS. After incubation with FITC-labeled rabbit anti-mouse
F(ab)2 (1/40; DAKO) for 30 min on ice, the cells
were washed and resuspended in 0.3 ml of 2% buffered formalin and
analyzed on a Becton Dickinson FACScan flow cytometer.
The TßRII level in the cells was quantified by flow cytometry analysis using the same procedure as for CD105. The antibody against TßRII was produced in goats immunized with purified recombinant, human soluble TßRII; specific immunoglobulin G (IgG) was purified by TßRII affinity chromatography according to the supplier’s manual (R&D Systems, Abingdon, U.K.). This antibody and the FITC-conjugated anti-goat antibody (Santa Cruz Biotechnology, Santa Cruz, Calif.) were diluted 1:10 (25 µg/ml) and 1:20 (20 µg/ml), respectively, for cell staining. A preimmunized goat IgG (Sigma) at the same concentration as the anti-TßRII antibody was used as a negative control antibody.
Cell cycle analysis
The effects of 0.25 µM ODN on cell cycle distribution were
evaluated by flow cytometry analysis as described previously
(25)
. 1 x 106 cells were fixed
with 2 ml cold 70% (v/v) ethanol in PBS, immediately mixed, and slowly
agitated for 15 min at room temperature. After two washes with PBS, 0.7
ml of 0.2 mg/ml pepsin (Sigma) in 2 M HCl was added to the pelleted
cells for simultaneous proteolysis and DNA denaturation at 37°C for
30 min. The hydrolysis was terminated by addition of 2 ml of 1 M Tris
(pH 10). Cells were washed twice and incubated in 0.3 ml PBS containing
10 µg/ml propidium iodide (Sigma) for 15 min on ice. To determine the
proportion of cells in various phases of the cell cycle, propidium
iodide staining of the DNA in 2 x 104
nuclei was quantified by flow cytometry.
HUVEC DNA synthesis as determined by [3H]-thymidine
incorporation
To further evaluate the effects of 0.25 µM ODN on DNA
synthesis, [3H]-thymidine incorporation assays
were carried out as described previously (25)
. Briefly,
1 x 104 cells that had been treated with
0.25 µM ODN for 72 h per well were seeded onto a 48-well plate
in 0.5 ml of medium containing 5% FCS and incubated for 24 h.
Cells were labeled with 1 µCi per well of
[3H]-thymidine (Amersham Pharmacia, Amersham,
U.K.) for 3 h, followed by two washes with prewarmed PBS, and
fixed in situ with cold methanol for 1 h. The cells
were then washed with ice-cold 5% (v/v) tri-chloroacetic acid (TCA)
(Sigma), with five subsequent changes over the next 24 h to remove
any thymidine that had not been incorporated into DNA. They were then
washed in PBS, and digested by addition of 0.4 ml per well of 0.1 M
NaOH for 1 h, and the cell lysate was transferred to 4 ml of
scintillation mixture (Fisons Chemicals). The amount of
incorporated [3H]-thymidine was determined on a liquid
scintillation analyzer (Packard Tricard 4660, Canberra Packard).
Extraction of CD105 protein from HUVECs
CD105 protein was extracted from HUVECs by solubilizing 1 x 107 cells/ml of extraction buffer
[0.2% (v/v) Nonidet P-40 in 0.1 M Tris buffer (pH 7.3), 0.5 mM PMSF,
1 mM pepstatin, 0.1 mM leupeptin, 1 mM EDTA]. The cell lysate was
microfuged at 13,000 rpm (Micro Centaur) for 10 min at 4°C and the
supernatant was collected for immunoblotting analysis.
Immunoblotting analysis of CD105 protein
HUVEC lysate corresponding to 1 x
105 cells was added to an equal volume of sample
buffer [0.1 M Tris-HCl, 4% (w/v) sodium dodecyl sulfate, 0.001%
(w/v) bromphenol blue, 20% (v/v) glycerol] and proteins were resolved
on 4–7.5% (w/v) sodium dodecyl sulfate-polyacrylamide gel. The
fractionated proteins were electrophoretically transferred onto a PVDF
membrane (Hybond-C super, Amersham) using a Trans-Blot system (Bio-Rad
Laboratories, Hercules, Calif.). Filters were blocked with 2% (w/v)
bovine serum albumin in PBS, 0.1% (v/v) Tween for 2 h at room
temperature. To detect CD105 protein, biotinylated Mab E9 (1:2000; 0.5
µg/ml) in blocking solution was applied and filters were incubated
overnight at 4°C. Finally, the blots were incubated with horseradish
peroxidase-conjugated streptavidin (1:8000 in blocking solution) for
2 h at 4°C with shaking. The CD105 protein was visualized by the
use of the enhanced chemiluminescence (ECL) system (Amersham).
Blots were stripped by washing overnight in PBS-Tween containing 2% (v/v) mercaptoethanol, washed twice in PBS-Tween, and incubated overnight at 4°C with mouse monoclonal antibody to CD31 (1:2000 in blocking buffer; DAKO). CD31 protein was detected by the addition of rabbit anti-mouse antibody-conjugated horseradish peroxidase (1:2000), washed with PBS-Tween, and the blot was developed by ECL.
Northern blot analysis of HUVEC mRNA
Total RNA was extracted using the acid phenol method
(26)
and electrophoresed as described previously
(27)
. Five micrograms of RNA per track were
electrophoresed through a 1% (w/v) agarose (Seakem GTG, Flowgen Labs.,
Sittingbourne, U.K.) under denaturing conditions at 7 V/cm. This was
capillary transferred in 10 x SSC buffer to PAL Biodyne B
membrane (Flowgen), UV cross-linked, baked at 70°C for 30 min, and
hybridized overnight at 68°C in 7% (w/v) sodium dodecyl sulfate,
0.5% (w/v) bovine serum albumin, 0.5 M NaCl with a random primed
(Boehringer Mannheim, Lewes, U.K.) 
32P dCTP
(3000 Ci/mMol, Amersham Pharmacia) -labeled 2.3 kb CD105 cDNA fragment
excised from pcEXV-EndoL plasmid. The blots were washed down to
0.2 x SSC at 68°C and signals were visualized by
phosphorimaging (Molecular Dynamics). Blots were reprobed with an
-actin cDNA probe as an internal control for quantitative
comparison.
Radio-ligand assay for 125I-TGFß1 binding on HUVECs
1 x 106 HUVECs in 1 ml medium
containing 5% FCS were incubated for 30 min at 37°C with 0.1, 1.0,
and 10.0 ng/ml of 125I-TGFß1 (1000 Ci/mMol,
Amersham Pharmacia). Nonspecific binding was assessed in the presence
of a 40-fold excess of competing unlabeled TGFß1. Cells were washed
thrice in prewarmed Hank’s solution (Life Technologies) and lysed by
0.5 ml of 0.1 M NaOH. The cell lysate was transferred to a new tube for
measuring the cell-associated radioactivity in a gamma counter.
Cell proliferation as determined using direct cell counting
HUVECs were trypsinized and 1 x 104
cells in 0.5 ml medium supplemented with 5% FCS was seeded onto
48-well plates precoated with 1% (w/v) gelatin. Three groups of cells
(untreated, AS-, or SC-treated HUVECs) were included in the same
experiment. After a 2 h period in which the cells were allowed to
adhere, control media or media containing various concentrations of
TGFß1 plus 5% FCS were added. The media and TGFß1 were renewed
every 24 h. At 72 h, the cells were washed once with
prewarmed PBS and cell suspension was prepared by trypsinization. The
number of cells in each well was determined by cell count using a
Coulter counter (Coulter Electronics, Hialeah, Fla.).
Cell migration as determined by the multichannel wounding system
2 x 105 cells/well were seeded onto
1% gelatin-coated Thermanox (Nunc, Life Technology) coverslips and
incubated at 37°C until they reached confluence. The coverslips were
rinsed in PBS and, by using the mechanical wounding device, 11 parallel
lesions (400 µM wide) were produced across the cell monolayer
(28)
. The wounded cells were rinsed in PBS to remove
dislodged cells and cellular debris, and placed in a well containing
0.5 ml medium supplemented with 5% FCS and TGFß1 at designated
concentrations. After incubation for 18 h, the coverslips were
washed three times in PBS, fixed with 100% ethanol for 10 min, and
allowed to air dry prior to staining with 0.1% (w/v) methylene blue.
The movement of cells into the denuded area was quantified using a
semiautomated computerized image analysis system. For each coverslip
five fields of view (each field covered 5.3x105
µM (2)
were taken at random. The repaired area in each
field of view was measured; by using the lesion area for time zero, the
repaired area was then converted to give % recovery: % recovery = repaired area/(5.3x105) x 100%.
Capillary formation in the 3-dimensional collagen matrix
Native type I collagen was extracted from rat tail tendons. To
prepare the 3-dimensional collagen matrix, 1 ml/well of the 0.1% (w/v)
collagen solution was deposited on a 12-well plate (Costar, Cambridge,
Mass.) and polymerized for 2 h at 37°C. One milliliter of HUVEC
suspension (8x105 cells/ml) was then seeded and
cells were allowed to attach to the collagen. The medium was gently
removed and 1 ml of 0.1% collagen solution was laid on top of the cell
monolayers. After 2 h incubation at 37°C to allow subsequent
polymerization, 2 ml of HUVEC growth medium was poured into each well
and incubation was continued for 2 h; thereafter, TGFß1 in 2 ml
normal medium was added to the culture to give the final concentration
of 0.001 or 0.1 ng/ml. The medium and TGFß1 were renewed every day.
Photographs were taken at specified intervals under a phase-contrast
microscope, using three fields at random in each gel.
Isolation and viability identification of HUVECs from collagen
matrix
Collagen gel was washed with PBS and transferred to a universal
tube containing 2 ml of 0.1% (w/v) type IV collagenase (Sigma).
Digestion of the gel was carried out in 37°C water bath for 30 min.
Cells were washed and harvested by centrifugation. The proportion of
nonviable cells were determined by trypan blue exclusion.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, treatment of HUVECs with AS ODN induced a
concentration-dependent inhibition on CD105 expression, with a mean
62% reduction at 0.25 µM. Although slightly increased CD105 levels
were seen in SC-treated cells over untreated cells, the difference was
not statistically significant (P>0.05).
|
The ability of the cells to reexpress CD105 was assessed using HUVECs
treated with 0.25 µM AS ODN for 72 h. Thereafter the cells were
cultured in ODN-free medium for the designated time courses and CD105
levels were quantified by flow cytometry. At the time of AS ODN
removal, the cells were nearly confluent; therefore, the cell number
did not increase during the ensuing incubation period (data not shown).
Figure 1C
shows that once CD105 was suppressed by AS ODN, it
remained unchanged for 48 h. The slight increase, seen at 72 h and 96 h, was not significant compared to 0 h
(P>0.05). At 120 h after removal of AS ODN,
CD105 levels recovered from 51.7% to 67.4% (P>0.05
compared to 0 h), indicating that the antisense effect can persist
for more than 5 days in these cells.
The effect of the AS ODN on CD105 inhibition was also evaluated by
immunoblotting (Fig. 1D
). Treatment of the cells with 0.25
µM AS ODN gave rise to 52% reduction in CD105 expression relative to
CD31 levels, as quantified using a densitometer (Molecular Dynamics,
Sunnyvale, Calif.). In contrast, no marked inhibition on CD105 levels
was observed in control ODN-treated cells. That the same molecular mass
(180 kDa) of CD105 appeared in untreated or ODN-treated cells indicated
that ODN treatment did not alter the molecular mass of CD105 protein.
Antisense oligonucleotide induced CD105 mRNA degradation
To investigate whether the mechanisms by which AS ODN suppressed
CD105 protein expression involved CD105 mRNA degradation, CD105 mRNA
levels were examined by Northern blot. Figure 1E
indicates
that CD105 mRNA levels (2.3 kb) were markedly decreased by treatment
with 0.25 µM AS ODN. With reference to -actin, CD105 mRNA was
reduced by 67% as quantified on a PhosphorImager (Molecular Dynamics).
Sense and SC ODNs did not induce CD105 mRNA degradation.
CD105 suppression did not alter CD31 and TßRII expression
To investigate whether ODNs affected other cell surface protein
expression, CD31 levels on ODN-treated cells were quantified using flow
cytometry and immunoblotting. HUVECs used for CD31 quantification were
taken from the same batch of cells as for CD105 analysis shown in Fig. 1A, B
. CD31 levels were unchanged in AS-treated cells
compared with untreated or SC-treated cells (data not shown). This was
further proved when using immunoblotting analysis, in which CD105 but
not CD31 was suppressed by AS treatment (Fig. 1D
).
Quantification of TßRII levels on HUVECs by flow cytometry showed
that TßRII expression did not significantly differ in the three
groups of cells (Fig. 2A
).
|
Reduction of CD105 did not interfere with TGFß1 binding on HUVECs
Since TGFß1 binds CD105 on HUVECs with high affinity, it was
necessary to analyze whether ligand binding to the cell surface was
altered in CD105 reduced cells. TGFß1 binding to HUVECs was carried
out in medium supplemented with 5% FCS, as used for cell proliferation
and migration assays. Data pooled from four experiments indicate that
specific binding of 125I-TGFß1 to untreated,
SC-, or AS-treated cells increased in a concentration-dependent manner,
with no significant difference between the three groups of cells (Fig. 2B
). This suggests that suppression of CD105 expression did
not significantly affect ligand binding to the specific receptors.
Cell cycle distribution and DNA synthesis were not altered on
treatment ODN
As shown by flow cytometry, immunoblotting, and mRNA analysis,
0.25 µM AS ODN appeared to provoke significant inhibition on CD105
protein translation and mRNA degradation. The effect of ODNs on cell
cycle distribution and [3H]-thymidine
incorporation at this concentration was evaluated. Treatment of HUVECs
with 0.25 µM AS or SC ODN neither altered cell cycle distribution nor
affected the DNA synthesis rate of these cells (Table 1
), indicating that this concentration did not produce detectable
nonspecific effects.
|
Reduction of CD105 on HUVECs enhanced the inhibitory effects of
TGFß1 on cell proliferation
In cell proliferation assay, the untreated, AS-, or
SC-treated cells were included so as to make parallel comparison
between groups. The effect of TGFß1 on cell proliferation was
quantified by direct cell counting.
When tested on untreated cells, TGFß1 was inhibitory over a wide
range of concentrations. The maximal inhibition (24% decrease in cell
number) on cell growth was noted at 0.1 ng/ml. The inhibitory effect
was concentration dependent from 0.001 to 0.1 ng/ml and reached a
plateau from 1 to 10 ng/ml, implying saturation of the signal receptors
on the cell surface (Fig. 3A
).
|
A similar trend was observed in SC-treated cells wherein the lowest cell number was shown at 0.1 ng/ml with a 28% decrease. No significant difference was observed compared with untreated cells at corresponding concentration points, except a lower cell number was noted at 0.01 ng/ml (P<0.05).
In AS-treated cells, in which 60–80% CD105 had been eliminated,
TGFß1 exhibited augmented inhibitory effects on cell growth. A
significantly lower cell number was counted at TGFß1 concentrations
of 0.01 to 10 ng/ml over either untreated or SC-treated HUVECs at the
corresponding concentration points. The maximal inhibition occurred at
0.1 ng/ml TGFß1 with a 38% decrease in cell number (Fig. 3A
).
TGFß1 inhibited HUVEC migration
The rate of cell migration was examined by a multichannel wounding
system. Postconfluent HUVEC monolayers were wounded to produce 11
straight, parallel denuded areas and TGFß1 was added to the media
immediately after wounding. Eighteen hours after wounding, the cells
were seen to have migrated into the wounded area with no apparent
increase in cell proliferation (28)
. The extent of the
repaired area was quantified by computerized image analysis and the
recovery rate calculated by comparison with the initial denuded area
(T0).
In the absence of TGFß1, recovery rates of the untreated, SC-, or
AS-treated cells were 48%, 49% and 47%, respectively, which were not
statistically different (P>0.05). Addition of 0.001 ng/ml
TGFß1 did not affect the migration of untreated (46%) and SC-treated
(47%) cells (P>0.05 compared with the control media), but
significantly inhibited the migration of AS-treated HUVECs (41%)
(P<0.05 compared to the control media). In the presence of
0.1 ng/ml TGFß1, the recovery rate of both untreated and SC-treated
cells was reduced to 41%, which was significantly lower than the
control media (P<0.05). However, this concentration
exhibited maximal inhibition in AS-treated cells, with a significantly
reduced recovery rate of 32% (P<0.01 compared with either
control media or the other two groups of cells) (Fig. 3B
).
HUVEC tube formation in the 3-dimensional collagen matrix
The three groups of HUVECs started to elongate and connect
head-to-tail in 0.1% collagen gel within 24 h of culture.
Capillary-like networks formed from 48 h to 72 h in culture.
TGFß1 was applied to the culture 4 h after seeding cells into
the collagen matrix with daily replenishment. In the absence of
TGFß1, the length of the capillary-like structures in untreated
(5149±523 µM/field), SC-treated (4619±439 µM/field), or
AS-treated (5553±103 µM/field) cells was not statistically different
(P>0.05) which suggested that these cells displayed
an equal ability to form capillary-like networks. No significant
inhibition on untreated (4196±484 µM/field) and SC-treated
(3921±570 µM/field) cells occurred at 0.001 ng/ml TGFß1
(P>0.05 compared with control media), whereas the
same concentration shortened the length of the tubes in AS-treated
cells (2679±368 µM/field) (P<0.05 compared with
corresponding control media or the other two groups of cells at this
concentration) (Fig. 3C
). The presence of 0.1 ng/ml TGFß1
significantly reduced the length of capillary-like structures in both
untreated (3078±225 µM/field) and SC-treated (3112±416 µM/field)
cells (P<0.05 compared with control media) (Fig. 3C
). However, a striking feature observed in AS-treated
cells was that addition of 0.1 ng/ml TGFß1 not only markedly
inhibited the elongation of the tubes (1762±282 µM/field,
P<0.01 compared with either control media or the other two
groups of cells at the same concentration), but also stimulated
breakdown of the networks (Fig. 4
).
|
TGFß1 induced cell mortality in the collagen matrix
Cells were isolated from the collagen matrix for viability
identification after 72 h culture. Data collected from two
duplicate experiments indicated that the presence of 0.1 ng/ml TGFß1
increased the percentage of dead cells (Fig. 3D
).
Furthermore, the most remarkable difference between these cells was
that a significantly higher proportion of nonviable cells was noted in
AS-treated cells (58.3%) over untreated (22.5%) or SC-treated
(28.3%) cells (p for all <0.01), demonstrating that TGFß1 displayed
augmented cytotoxicity on CD105-deficient endothelial cells in the
3-dimensional collagen matrix.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
reported that in contrast to mice that lacked
TGFß1 wherein vascularization was unaffected, the absence of CD105 in
knockout mice leads to death by 11.5 day p.c. from defective vascular
development. We suggest that since we have undertaken our experimental
work on primary human endothelial cells indicate that the mouse model
is relevant to the human situation.
TGFß1 is an anti-proliferative factor over a wide range of cell
types. The present data confirm that TGFß1 behaves as a potent
inhibitor of endothelial cell growth, migration and capillary
formation, which is consistent with previous reports
(29
30
31)
. Although TGFß1 is anti-angiogenic in
vitro, it induces angiogenesis in in vivo assays such
as the chick chorioallantoic membrane assay (32)
. Ex
vivo studies in human showed that TGFß1 is present in tissues
undergoing neovascularization (16)
, which suggests a
positive stimulatory role for this growth factor. A possible
explanation for these conflicting observations may be the mixed
cellularity that is present in vivo whereby TGFß1 will
initially recruit inflammatory cells into the immediate vicinity of the
applied stimulus (32)
. Subsequent angiogenesis is then
mediated by positive regulators—for example, VEGF produced by
TGFß1-recruited inflammatory cells. Studies with transgenic mice
overexpressing TGFß1 support the notion that TGFß1 is an indirect
angiogenic factor in vivo, since these animals do not
develop inflammation or neovascularization within the affected organs
(3)
.
Using AS-treated HUVECs (CD105 reduced by 60–80%), we initially
evaluated the cell proliferation rate in a 2-dimensional system in
order to determine the association between the expression levels of
CD105 and the cellular responses to TGFß1. TGFß1 was found to be
inhibitory over a wide range of concentrations, with maximal inhibition
at 0.1 ng/ml. This concentration is compatible with the concentration
measured in the plasma (33)
. The presence of the plateau
from 1 to 10 ng/ml implies the saturation of the signal receptors,
although TGFß1 has been demonstrated to up-regulate the expression of
Smad 6 and Smad 7, both of which antagonize TGFß signaling
(34)
. This implies that higher ligand concentration may
not induce greater cellular responses due to the augmented expression
of these negative regulators. The cell growth rate was similar in
untreated and SC-treated HUVECs, except that a slightly lower cell
number was counted at 0.01 ng/ml in SC-treated HUVECs. In contrast,
addition of TGFß1 produced a significant growth inhibition in
AS-treated cells when compared to control and SC-treated cells exposed
to the same concentration of growth factor. No significant difference
in TßRII level was found between the three groups of cells, which
indicates that the inhibitory response to TGFß1 in AS-treated HUVEC
was not due to up-regulation of the type II receptor. This was
previously suggested as a possible mechanism for the anti-proliferative
effects of TGFß1 (35)
.
Endothelial cell migration is fundamental for the formation of new vessels from preexisting blood vessels. In the present study, the cell motility in response to TGFß1 was measured by a multichannel wounding system such that repair into the denuded area was due entirely to cell migration. TGFß1 clearly inhibited HUVEC migration, but the maximal inhibition was observed in AS-treated cells. Treatment of HUVECs with SC ODN did not alter the recovery rate compared with untreated cells. These results suggest that the presence of CD105 in HUVECs is capable of weakening the inhibitory effects of TGFß1 on cell migration.
It has been observed that HUVECs form extensive networks of
capillary-like structures with lumen after 24 h culture in a
3-dimensional collagen gel system (36)
. TGFß1 was shown
to be a direct-acting inhibitor of capillary tube formation (3
, 37)
. However, if the TGFß1 is applied when the new vessels
have been established, it may promote the phase of resolution by
inducing endothelial quiescence and vessel maturation. In the present
study, TGFß1 was added to the cultures 4 h after seeding the
cells, as no network had formed at this stage. The presence of 0.001
ng/ml TGFß1 showed no significant effect on the tube formation of
control and SC-treated cells, but an inhibitory action was noted in
AS-treated cells. The networks were markedly reduced by the addition of
TGFß1 (0.1 ng/ml) in all three groups of cells, but greater
inhibition in tube formation was observed in CD105-deficient cells.
This was initially thought to be due entirely to the augmented
inhibitory effects of TGFß1 on cell proliferation and migration.
However, the most striking feature of the AS-treated HUVECs was the
presence of massive cell mortality when compared to control cultures.
These data demonstrated that TGFß1 becomes cytotoxic when HUVEC CD105
levels are reduced below a certain threshold.
The observation that CD105 and TßRI and/or RII form a heteromeric
receptor complex in endothelial cells in the presence of ligand
provides circumstantial evidence that CD105 may be involved in the
modulation of TGFß signaling (38)
. It is noteworthy that
CD105 is constitutively phosphorylated in endothelial cells, with
eightfold higher phosphorylation in L-CD105 than that in S-CD105
(39)
. Together with the notion that L-CD105 is much more
efficient than S-CD105 in blocking the inhibitory effects of TGFß1 on
U937 cell growth (40)
, this raises the possibility that
the extent of phosphorylation of the cytoplasmic domain plays a
significant role in modulating TGFß signaling. The enhanced responses
of CD105-deficient cells to TGFß1, in which both extracellular and
cytoplasmic domains of CD105 were markedly depressed, provide direct
evidence that this receptor plays a remarkable negative role in TGFß1
signaling on vascular endothelial cells.
We analyzed the cell surface binding of TGFß1 to determine whether
interaction with specific receptors was altered.
Concentration-dependent binding was observed from 0.1 to 10 ng/ml
TGFß1, with no difference between the three groups of cells,
indicating that the reduced level of CD105 protein on cell surface was
unable to disturb the ligand binding to specific receptors. Though
CD105 is abundantly expressed in endothelial cells, only a small amount
(~10%) binds TGFß1 on cell surface (10)
; hence, it is
not surprising that down-regulation of this protein could not suppress
the cell surface binding of TGFß1.
We have previously demonstrated that CD105 can be shed into the media
(24)
forming CD105/TGFß1 complex in vitro
(33)
. Therefore, we postulate that the reduction of CD105
on HUVECs may permit more free TGFß1 to remain in the media, which
can then bind to its signal receptors and thus trigger enhanced
cellular responses to TGFß1.
The most important feature of an AS ODN is its ability to hybridize in an extremely sequence-specific manner to its corresponding mRNA without binding to related or unrelated mRNAs. The specificity of the AS ODN to CD105 mRNA was confirmed by Northern blot analysis of total RNA extracted from the same batch of cells used for CD105 protein analysis. That AS ODN treatment resulted in a significant decrease in CD105 mRNA levels, whereas both sense and SC ODN caused no effect indicated that the decrease in CD105 mRNA was a specific consequence of AS ODN treatment. The selective suppression of CD105 protein was further verified by quantifying two cell surface proteins, CD31 and TßRII, both constitutively expressed in HUVECs. Whereas CD105 was reduced by 60–80%, CD31 and TßRII remained unaltered in these cells, demonstrating the highly efficient and selective nature of this AS ODN.
How can our current observations be rationalized with the previously observed in vivo angiogenic response to TGFß1? It is clear that normal levels of CD105 insulate endothelial cells from the inhibitory effects of TGFß1 whereas reduced levels enhance TGFß1-mediated growth inhibition and inhibit angiogenesis in the collagen gel system. These data suggest that the development of an in vivo angiogenic response may depend on achieving a critical level of TGFß1 stimulation, which in turn is modulated by the level of CD105 expression. Clearly these findings have important clinical relevance for situations in so-called angiogenic diseases, but will require in vivo studies.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  N. A. Dallas, S. Samuel, L. Xia, F. Fan, M. J. Gray, S. J. Lim, and L. M. Ellis Endoglin (CD105): A Marker of Tumor Vasculature and Potential Target for Therapy Clin. Cancer Res., April 1, 2008; 14(7): 1931 - 1937. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. L. Moody, S. Singbrant, G. Karlsson, U. Blank, M. Aspling, J. Flygare, D. Bryder, and S. Karlsson Endoglin Is Not Critical for Hematopoietic Stem Cell Engraftment and Reconstitution but Regulates Adult Erythroid Development Stem Cells, November 1, 2007; 25(11): 2809 - 2819. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Masuyama, H. Nakatsukasa, N. Takamoto, and Y. Hiramatsu Correlation between Soluble Endoglin, Vascular Endothelial Growth Factor Receptor-1, and Adipocytokines in Preeclampsia J. Clin. Endocrinol. Metab., July 1, 2007; 92(7): 2672 - 2679. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. G. Docherty, J. M. Lopez-Novoa, M. Arevalo, A. Duwel, A. Rodriguez-Pena, F. Perez-Barriocanal, C. Bernabeu, and N. Eleno Endoglin regulates renal ischaemia-reperfusion injury Nephrol. Dial. Transplant., August 1, 2006; 21(8): 2106 - 2119. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Maio, M. Altomonte, and E. Fonsatti Is it the primetime for endoglin (CD105) in the clinical setting? Cardiovasc Res, March 1, 2006; 69(4): 781 - 783. [Full Text] [PDF]
|
|
|
|
|
|
M. Jerkic, A. Rodriguez-Barbero, M. Prieto, M. Toporsian, M. Pericacho, J. V. Rivas-Elena, J. Obreo, A. Wang, F. Perez-Barriocanal, M. Arevalo, et al. Reduced angiogenic responses in adult endoglin heterozygous mice Cardiovasc Res, March 1, 2006; 69(4): 845 - 854. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Fernandez-L, F. Sanz-Rodriguez, R. Zarrabeitia, A. Perez-Molino, R. P. Hebbel, J. Nguyen, C. Bernabeu, and L.-M. Botella Blood outgrowth endothelial cells from Hereditary Haemorrhagic Telangiectasia patients reveal abnormalities compatible with vascular lesions Cardiovasc Res, November 1, 2005; 68(2): 235 - 248. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Muenzner, M. Rohde, S. Kneitz, and C. R. Hauck CEACAM engagement by human pathogens enhances cell adhesion and counteracts bacteria-induced detachment of epithelial cells J. Cell Biol., August 29, 2005; 170(5): 825 - 836. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Pece-Barbara, S. Vera, K. Kathirkamathamby, S. Liebner, G. M. Di Guglielmo, E. Dejana, J. L. Wrana, and M. Letarte Endoglin Null Endothelial Cells Proliferate Faster and Are More Responsive to Transforming Growth Factor {beta}1 with Higher Affinity Receptors and an Activated Alk1 Pathway J. Biol. Chem., July 29, 2005; 280(30): 27800 - 27808. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. B. Schmidt-Weber, M. Letarte, S. Kunzmann, B. Ruckert, C. Bernabeu, and K. Blaser TGF-{beta} signaling of human T cells is modulated by the ancillary TGF-{beta} receptor endoglin Int. Immunol., July 1, 2005; 17(7): 921 - 930. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Lux, C. Beil, M. Majety, S. Barron, C. J. Gallione, H.-M. Kuhn, J. N. Berg, P. Kioschis, D. A. Marchuk, and M. Hafner Human Retroviral gag- and gag-pol-like Proteins Interact with the Transforming Growth Factor-{beta} Receptor Activin Receptor-like Kinase 1 J. Biol. Chem., March 4, 2005; 280(9): 8482 - 8493. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Tang, B. Low, S. A. Rutherford, and Q. Hao Thrombin induces endocytosis of endoglin and type-II TGF-{beta} receptor and down-regulation of TGF-{beta} signaling in endothelial cells Blood, March 1, 2005; 105(5): 1977 - 1985. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Lebrin, M. Deckers, P. Bertolino, and P. ten Dijke TGF-{beta} receptor function in the endothelium Cardiovasc Res, February 15, 2005; 65(3): 599 - 608. [Abstract] [Full Text] [PDF]
|
