crosstalk between vegf and bmp9 signalling in the context ... · ii crosstalk between vegf and bmp9...
TRANSCRIPT
Crosstalk between VEGF and BMP9 signalling in the context of preeclampsia
by
Valentin Sotov
A thesis submitted in conformity with the requirements for the degree of Master of Science
Department of Immunology University of Toronto
©Copyright by ValentinSotov2013
ii
Crosstalk between VEGF and BMP9 signalling in the context of preeclampsia
Valentin Sotov
Master of Science
Department of Immunology
University of Toronto
2013
Abstract
Preeclampsia is a pregnancy related disorder, characterized by proteinuria and hypertension. The
pathogenesis of preeclampsia is poorly understood; however, two proteins, called sFlt-1 and
sEng, were found to be highly elevated in the maternal plasma weeks prior to the onset of
clinical symptoms. sFlt-1 and sEng are thought to inhibit VEGF and TGF-β receptor signalling
respectively. In order to elucidate how these proteins may contribute to preeclampsia, we looked
at their ability to affect the secretion of endothelin-1 (ET-1), a powerful vasoconstrictor, shown
to be dysregulated in preeclampsia. We found that both TGF-β1 and BMP9, a recently described
ligand for sEng, induce ET-1 secretion through Smad1/5/8 and p38 pathways. Moreover, we
report that sEng and VEGF can efficiently block ET-1 secretion, induced by BMP9. We propose
that the balance between VEGF and BMP9 signalling is disturbed during preeclampsia, leading
to excessive release of ET-1, which in turn may cause hypertension.
iii
Dedication
I dedicate this thesis to my father, Alexander Sotov. He was the person who encouraged me to
pursue science in the first place, so none of this will be possible without him. His love, faith and
constant support helped me immensely during the completion of my Master’s degree.
iv
Acknowledgements
I would like to thank my supervisor, Dr. Michelle Letarte, who gave me the opportunity to work
on this fantastic project. Without her guidance and support none of this would ever be possible. I
owe a special thanks to Allison Gregory, who started this project and was incredibly helpful
throughout my time in the laboratory. I also would like to thank Dr. Mirjana Jerkic, who
welcomed me into the lab and had great patience in teaching me how to do science. My labmates
created an amazing atmosphere in the lab, and I should definitely thank Madonna and Daniela
for being there for me when I needed it the most. Thank you so much guys.
I also would like to thank my graduate committee and specifically Dr. Alberto Martin and Dr.
Isabella Caniggia, their help with this project cannot be overstated. I also extend my most sincere
thanks to Zhijie Li from the Rini Lab, for all his help with the surface Plasmon resonance
experiments.
v
Acknowledgements of Contributions
I wish to acknowledge the contribution of Allison Lindsay Gregory from our laboratory and
Zhijie Li from Dr. James Rini research team who participated and guided me in all surface
plasmon resonance experiments, including those described in Figure 9. These experiments were
performed in the Department of Biochemistry, at the University of Toronto. I am also grateful to
Dr Rini for allowing these studies and for the helpful discussions.
vi
Table of Contents
Table of Contents
Table of Contents ........................................................................................................................... vi
List of Figures .............................................................................................................................. viii
List of abbreviations ...................................................................................................................... ix
Chapter 1 Introduction .....................................................................................................................1
1 Introduction .................................................................................................................................1
1.1 Overview ..............................................................................................................................1
1.2 Preeclampsia ........................................................................................................................1
1.2.1 Overview of the disease ...........................................................................................1
1.2.2 Pathogenesis of preeclampsia ..................................................................................3
1.3 Overview of TGF-β Receptors and Signal Transduction ..................................................15
1.3.1 TGF-β superfamily ................................................................................................15
1.3.2 TGF-β receptor signalling ......................................................................................15
1.3.3 Effects of TGF-β superfamily signalling ...............................................................22
1.4 Endothelin-1 .......................................................................................................................25
1.4.1 Structure and Function of ET-1 .............................................................................25
1.4.2 Role of ET-1 in Preeclampsia ................................................................................27
1.5 Thesis Objectives ...............................................................................................................29
1.5.1 Rationale ................................................................................................................29
1.5.2 Hypothesis..............................................................................................................29
1.5.3 Specific Objectives ................................................................................................29
2 Materials and Methods ..............................................................................................................30
2.1 Cell culture .........................................................................................................................30
vii
2.2 Cell based stimulation and inhibition assays .....................................................................30
2.3 Western Blotting ................................................................................................................30
2.4 Enzyme-Linked Immunosorbent Assay (ELISA) for ET-1 ...............................................31
2.5 Biacore Biosensor Analysis ...............................................................................................32
2.6 Statistical Analysis .............................................................................................................32
3 Results .......................................................................................................................................33
3.1 BMP9 and TGF-β1 induce ET-1 secretion in mouse embryonic endothelial cells ...........33
3.2 sEng (∆586) binds with high affinity to BMP9, but not to TGF-β1 ..................................35
3.3 sALK1-Fc and sEng block BMP9 – but not TGF-β1 –induced ET-1 secretion ................39
3.4 VEGF interferes with BMP9 but not TGF-β1 signalling to block ET-1 secretion ...........42
3.5 Additive effects of BMP9 and TGF-β1 on ET-1 secretion ................................................45
3.6 Analysis of the ability of inhibitors to block their respective phosphorylation
pathways ............................................................................................................................47
3.7 SMAD1 and p38 phosphorylation are involved in BMP9 – and TGF-β1 –induced
ET-1 secretion ....................................................................................................................51
3.8 VEGF may lower BMP9 –induced p38β phosphorylation ................................................55
4 Discussion .................................................................................................................................57
4.1 BMP9 and TGF-β1 induce ET-1 secretion from mouse embryonic endothelial cells .......57
4.2 Potential role of sEng in PE ...............................................................................................58
4.3 VEGF signalling and hypertension ....................................................................................61
4.4 Pathways involved in ET-1 production..............................................................................62
4.5 Conclusion .........................................................................................................................64
Bibliography ..................................................................................................................................66
viii
List of Figures
Figure 1 ……………………………………………………………………………. 4
Figure 2 ……………………………………………………………………………. 6
Figure 3 ……………………………………………………………………………. 8
Figure 4 ……………………………………………………………………………. 12
Figure 5…………………………………………………………………………….. 14
Figure 6 ……………………………………………………………………………. 17
Figure 7 ……………………………………………………………………………. 20
Figure 8 ……………………………………………………………………………. 34
Figure 9 ……………………………………………………………………………. 37
Figure 10 ……………………………………………………………………………40
Figure 11 ……………………………………………………………………………53
Figure 12 ……………………………………………………………………………46
Figure 13 ……………………………………………………………………………49
Figure 14 ……………………………………………………………………………53
Figure 15 ……………………………………………………………………………56
Figure 16 ……………………………………………………………………………60
ix
List of abbreviations
ActRII : Activin receptor type II
ALK : Activin receptor-like kinase
AT1 : Angiotensin II type I receptor
AT2 : Angiotensin II type II receptor
AT1-AA : Angiotensin II type I receptor agonistic autoantibodies
BMP : bone morphogenic protein
BMPRII : bone morphogenic protein receptor type II
Eng : endoglin protein
eNOS : endothelial nitric oxide synthase
ET-1 : endothelin-1
GDF : growth differentiation factor
HHT : Hereditary Hemorrhagic Telangiectasia
HHT1 : Hereditary Hemorrhagic Telangiectasia type 1
HHT2 : Hereditary Hemorrhagic Telangiectasia type 2
HELLP : hemolysis, elevated liver enzymes, low platelets
HUVEC : Human umbilical vein endothelial cells
I-Smads : Inhibitory Smads
Id-1 : Inhibitors of differentiation-1
x
KD : affinity constant
mEng : membrane endoglin
MMP : matrix metalloproteinase
PAI-1 : Plasminogen activator inhibitor-1
PE : Preeclampsia
PlGF: Placental growth factor
R-Smads : Receptor activated Smads
sEng : soluble endoglin
sFlt1 : soluble fms-like kinase (soluble VEGF receptor type 1)
SPR : surface plasmon resonance
TβRII : Transforming growth factor β receptor II
TGF-β : Transforming growth factor β
TNF-α : Tumor necrosis factor α
VEGF : vascular endothelial growth factor
1
Chapter 1 Introduction
1 Introduction
1.1 Overview
Preeclampsia (PE) affects 5-8% of all pregnant women worldwide and is one of the leading
causes of maternal death in childbirth. Hallmark symptoms of PE include hypertension and
proteinuria after 20 weeks of gestation (1). To date, there is no definite treatment of PE and the
only available therapy is to deliver the baby. Pathogenesis of PE is not clearly understood, but it
is evident that the placenta plays a crucial role in the disease (2). Specifically, during PE, the
placenta produces a number of anti-angiogenic factors, such as sFlt-1 (soluble VEGF receptor 1)
and soluble endoglin (sEng), which have been shown to cause systemic endothelial dysfunction
(3, 4). It has been postulated that an increase in sFlt-1 and sEng, which block VEGF and TGF-
β signalling pathways respectively, is a major cause of hypertension and proteinuria, associated
with PE. However, recent studies also implicate increased ET-1 secretion as one of the key
events in PE associated hypertension (5). ET-1 is a very potent vasoconstrictor, elevated in
preeclamptic patients, though the exact mechanism responsible for the increase in ET-1 secretion
during preeclampsia remains unknown. Therefore, we examined the effects of VEGF and TGF-β
superfamily ligands on ET-1 secretion. This thesis will focus on signalling pathways employed
by VEGF and TGF-β superfamily ligands with respect to ET-1 secretion. To better understand
the context of the experiments performed in this thesis, the introduction will give a quick
overview of the pathogenesis of preeclampsia, followed by a discussion of TGF-β superfamily
receptor signalling and the role of ET-1 in the context of PE.
1.2 Preeclampsia
1.2.1 Overview of the disease
Preeclampsia continues to be the most common cause of maternal and neonatal morbidity and
mortality in the world (6). Hallmark symptoms of PE are maternal systemic hypertension and
proteinuria after 20 weeks of pregnancy. Two subtypes are recognized: early and late onset PE.
Early onset PE occurs before 34 weeks of gestation and is thought to be a fetal disorder, often
associated with fetal growth restriction and incomplete trophoblast invasion. In contrast, late
2
onset PE occurs after 34 weeks of pregnancy and is considered a maternal disorder usually
associated with normal uterine resistance and low rate of fetal involvement with no signs of
growth restriction. Though both types of PE share a number of features, this thesis will focus on
early-onset PE (7, 8).
1.2.1.1 Clinical Presentation and Diagnosis
The hallmark symptoms of PE are proteinuria and hypertension. Classic PE is diagnosed by a
systolic blood pressure ≥ 140 mm Hg or a diastolic pressure ≥ 90 mm Hg on two or more
occasions and at least 4 hours apart, after week 20 of gestation. However, PE is a highly
heterogeneous disease; the extent of hypertension varies between patients and around 20% of
cases show minimal to no signs of proteinuria, making the disease much harder to diagnose (9).
PE is categorized into mild and severe forms. About 5% of PE cases will progress to severe
disease, associated with headaches, liver dysfunction, pulmonary edema, renal failure, eclampsia
or thrombocytopenia (1, 10). By definition, preeclampsia should be a precursor for eclamptic
seizures; however, 20% of women progress to eclampsia without proteinuria and in some cases
without hypertension (11, 12). There are no reliable tests for eclamptic seizures. Magnesium
sulfate, which is often given during preeclampsia to prevent seizures, has a number of known
side effects, including areflexia and respiratory distress (13).
Another complication of severe PE is hemolysis, elevated liver enzymes, low platelet (HELLP)
syndrome, which occurs in 2% of women with PE. HELLP syndrome can be life threatening for
both the mother and fetus, though most deaths result from complications of premature birth
before 28 weeks of pregnancy (14, 15). The mild form of PE is rarely dangerous for the fetus or
the mother, while the severe form can be fatal and accounts for 10%–15% of maternal deaths
worldwide (16).
There is no test to predict preeclampsia and in many cases diagnosis is made too late to keep the
symptoms in check. Recent findings, however, indicate that an increase in sFlt-1 and sEng
concentrations in the maternal circulation precedes the clinical symptoms and could be used to
predict preeclampsia several weeks in advance. In fact, an increase in sEng levels occurs as early
as 2-3 months before the onset of maternal symptoms and has a strong association with disease
severity (17). Several studies measuring the levels of sFlt-1, VEGF and PlGF (placental growth
factor), in the maternal serum, confirmed that sFlt-1 rises shortly after sEng (18, 19). Moreover,
3
the rise in sFlt-1 is often accompanied by a decrease in VEGF and PlGF levels. Though neither
sFlt-1 nor sEng alone can accurately predict preeclampsia, studies have indicated that a rise in
the sFlt-1 to PlGF ratio accompanied by an increase in sEng, provides clinicians with an
opportunity to accurately predict preeclampsia as early as one month before the onset of clinical
symptoms (20). These biomarkers can help in the early detection of PE.
1.2.1.2 Treatments, prognosis
Once diagnosis of PE is established, patients start undergoing frequent blood pressure
measurements and blood tests to check kidney and liver function. The fetus is monitored through
heart rate measurements and ultrasound. Preeclampsia rarely progresses to the severe stage, and
the mild stage can usually be managed expectantly until 37 weeks of pregnancy (21). Women
who progress to severe preeclampsia are monitored and are expected to deliver the baby on time
unless: a) hypertension can no longer be managed through medication; b) the patient develops
either eclampsia or HELLP syndrome; or c) the fetus shows signs of distress. By the 34th
week
of gestation, risks associated with the birth of the baby are minimal, and delivery is suggested
(1). Patients with preeclampsia are usually given a solution of magnesium sulfate, shown to
significantly decrease incidents of eclampsia (22).
Interestingly, while delivery of the placenta cures all symptoms associated with preeclampsia,
recent studies have suggested that there are long-term effects associated with disease.
Specifically, 20% of women with PE develop hypertension within 7 years of the pregnancy,
compared with 2% among women with normal pregnancies. Additionally, preeclamptic women
were 3 times more likely to develop cardiovascular disorders later in life (23). These numbers are
shared with other non-pregnancy related cardiovascular and especially hypertensive disorders,
suggesting a common etiology (24).
1.2.2 Pathogenesis of preeclampsia
A number of factors have been shown to contribute to the pathogenesis of preeclampsia. Figure 1
illustrates some of the known molecules and pathways, thought to play an important role in
preeclampsia.
4
Figure 1. A summary of key factors involved in PE pathogenesis.
A number of factors, including excessive production of AT1-AA (Angiotensin 1 auto-
antibodies), genetic predisposition, oxidative stress and decreased heme oxygenase activity have
been linked to placental dysfunction, which in turn results in the production of anti-angiogenic
factors, such as sFlt-1 and sEng, to induce endothelial dysfunction and subsequent hypertension
(HTN), proteinuria and other complications.
Figure reproduced with permission from The American Physiology Society.
Wang A, Rana S, Karumanchi SA. 2009. Preeclampsia: the role of angiogenic factors
in its pathogenesis. Physiology (Bethesda) 24: 147-58.
5
1.2.2.1 Regulation of Trophoblast Invasion
Preeclampsia has been described as a disease of theories, highlighting the fact that, despite 100
years of active research, its pathogenesis is yet to be understood. However, most researchers
agree that failure of the cytotrophoblast to invade the spiral arteries occurs in most PE cases.
During the first half of gestation, cytotrophoblast start migrating from the fetal compartment into
the uterine wall and associated vasculature, thus attaching the fetus to the mother. By doing so,
trophoblast cells invade and modify spiral arteries into low-resistance high blood-flow vessels,
capable of supplying the growing fetus with enough nutrients (25, 26). To accomplish this
remarkable transformation, trophoblast cells undergo a process called molecular mimicry, where
they lose molecular markers associated with epithelial cells, but upregulate endothelial markers,
which allow them to displace native endothelial cells and modify spiral arteries. However, during
preeclampsia, this process is incomplete and spiral arteries remain high-resistance low blood-
flow vessels, that may lead to hypoxia (Figure 2). The exact mechanism of trophoblast invasion
has not been elucidated; however, a number of molecules have been shown to affect this process.
To invade the spiral artery, the trophoblast must digest the extracellular matrix (ECM). To do
this, the trophoblast upregulates metalloproteinases, like MMP9, which are very efficient in
degrading ECM (27, 28). Therefore, failure of ECM digestion may ultimately result in a poor
trophoblast invasion. Interestingly, Lin et al, showed that trophoblast cells from preeclamptic
patients failed to upregulate expression of MMP9 at the mRNA level, and showed reduced
invasion (29). Other studies have shown that TGF-β3 and TGF-β1 signalling inhibit trophoblast
proliferation and invasion, probably by stimulating tissue inhibitors of metalloproteinases (30).
Interestingly, TGF-β3 expression is induced by hypoxia inducing factor 1 (HIF-1α), which is
itself upregulated during hypoxia (31). Therefore, hypoxia could be responsible for poor
trophoblast invasion. Not surprisingly, the plasma concentration of TGF-β3 was found to be
elevated in preeclamptic women (32).
Additionally, endoglin has been shown to play a role in trophoblast invasion. Specifically,
antibodies raised against endoglin were shown to enhance trophoblast differentiation and
invasion, presumably by interfering with TGF-β1 and TGF-β3 signalling (33). Later reports
6
Figure 2. Abnormal Placentation in Preeclamptic Women.
In normal pregnancy, the invasive trophoblast invades the spiral artery and transforms it into a
low resistance high volume artery. This allows the fetus to receive all required nutrients for
successful growth. During preeclampsia, however, the trophoblast fails to invade, resulting in a
small caliber high resistance spiral artery.
Figure reproduced with permission from The American Physiology Society.
Wang A, Rana S, Karumanchi SA. 2009. Preeclampsia: the role of angiogenic factors
in its pathogenesis. Physiology (Bethesda) 24: 147-58.
7
confirmed these findings, showing that the loss of endoglin promotes the invasion of extravillous
trophoblast (34). Recent studies have indicated that the plasma of preeclamptic women can
inhibit trophoblast invasion, suggesting that certain circulating soluble factors can actively block
trophoblast invasion. As of now these factors have yet to be identified (35). Overall, it is likely
that both endoglin and TGF-β play a crucial role in remodeling of spiral arteries.
To complete the invasion process, trophoblast cells need to change their adhesion molecule
profile from epithelial to endothelial, thus insuring a proper placentation process. However,
during preeclampsia trophoblast cells continue to express α6β4, αVβ5 and E-cadherin molecules,
which are usually present on epithelial cells and are associated with a non-invasive phenotype. In
fact, antibodies against E-cadherin were shown to significantly enhance the invasion process,
implying that E-cadherin normally blocks invasion. Alternatively, blockade of VE-cadherin
function, using antibodies, resulted in much reduced invasion, suggesting that the role of VE-
cadherin is to enhance trophoblast invasion (36). Experiments confirmed that preeclamptic
trophoblast cells do not show expression of αVβ3, α1β1integrins as well as VE-cadherin, which
were found to be essential for successful invasion (37). Overall, it is clear that upregulation of
specific integrin molecules greatly potentiates the invasive capabilities of the trophoblast;
however, in PE, trophoblast cells fail to acquire endothelial cells markers, leading to shallow
invasion and reduced blood flow (37, 38). It is thought that hypoxia, induced by the failure of
trophoblast to remodel spiral arteries, acts on placenta to induce several anti-angiogenic
molecules, such as sFlt-1. sFlt-1 will in turn block both VEGF and PlGF signalling in an attempt
to increase blood circulation and bring more nutrients to the fetus.
1.2.2.2 Dysregulation of VEGF/PlGF Signalling Pathway
VEGF and PlGF proteins are structurally and functionally similar. The difference lies in the
ability of VEGF to bind to two distinct receptors: VEGFR-1 and VEGFR-2, while PlGF can only
bind to VEGFR-I. It is thought that PlGF enhances VEGF function by competing for VEGFR-I
and forcing VEGF to bind to the more active receptor, VEGFR-1 (Figure 3) (39). In response to
ischemia, the placenta starts producing a variety of anti-angiogenic proteins (20). One of the
most widely recognized proteins is sFlt-1 (or sVEGFR-1), which is primarily secreted from the
syncytiotrophoblast into the maternal circulation. sFlt-1 is a splice variant of VEGFR-1 (Flt-1),
which lacks the intracellular and membrane regions. VEGFR-1 can bind VEGF and
8
Figure 3. VEGF signalling in normal and preeclamptic pregnancies.
During nomal pregnancy, VEGF can bind to either VEGFR-1 or VEGFR-2 to phosphorylate and
activate eNOS to produce the very potent vasodilator NO. In PE, however, high levels of sFlt-1
bind circulating free VEGF and PlGF, blocking their signalling and subsequent NO production,
leading to hypertension.
Figure adapted from Luizon MR, Sandrim VC, Palei AC, Lacchini R, Cavalli RC, Duarte G,
Tanus-Santos JE. Epistasis among eNOS, MMP-9 and VEGF maternal genotypes in
hypertensive disorders of pregnancy. Hypertension Research (2012) 35, 917–921.
9
PlGF, therefore sFlt-1 is extremely efficient at scavenging both molecules. It should be noted,
that by binding to VEGF, sFlt-1 also blocks signalling of VEGFR-2, as VEGF can signal through
both receptors (40). Studies have shown that preeclamptic women have a 6-fold increase in
serum concentration of sFlt-1 compared to normal pregnancies. Not surprisingly, elevated sFlt-1
leads to a low concentration of free VEGF and PlGF in the serum (20). Interestingly, inhibition
of VEGF alone was not sufficient to induce PE like symptoms in pregnant rats. Both PlGF and
VEGF pathways needed to be blocked to induce PE, suggesting that PlGF plays a key role in
pathogenesis of PE (3).
The important role of sFlt-1 in PE is underlined by experiments showing that mice injected with
sFlt-1, developed several of the symptoms occurring in PE patients, including proteinuria,
hypertension and renal failure (4). Moreover, recent studies have indicated the development of
hypertension and proteinuria in patients undergoing anti-VEGF therapy as part of their cancer
treatment (41). Therefore, it appears that sFlt-1 by itself can mimic the majority of PE
symptoms.
VEGF is a known angiogenic molecule that stimulates production of NO and PGI2, both of
which are extremely potent vasodilators (Figure 3). VEGF can induce Akt phosphorylation,
leading to phosphorylation and activation of eNOS, which is a major NO-producing enzyme
(42). VEGF can also upregulate eNOS expression through ERK phosphorylation (43). Therefore
VEGF can induce vasodilatation by upregulating or activating eNOS. Overall, a hallmark
symptom of PE, systemic hypertension, could be partially attributed to the lack of free VEGF
and PlGF molecules in circulation (44). Additionally, both VEGF and VEGF receptors are
highly expressed by the glomerular endothelium and VEGF was shown to have a stabilizing
effect on glomerular endothelial cells. A podocyte specific VEGF-knockout mice experienced
proteinuria and glomerular endotheliosis, highlighting the importance of locally produced VEGF
for glomerular endothelium function (45). It is likely that the lack of VEGF signalling in the
glomerular endothelium is a primary cause of proteinuria in PE patients.
1.2.2.3 Heme-oxygenase 1 and PI3K/Akt Pathways
Heme-oxygenase 1 (HO-1) is an inducible enzyme that catalyzes the degradation of heme into
biliverdin and carbon monoxide. Therefore, not only can HO-1 protect cells against oxidative
damage induced by heme, but it also releases CO, which acts as a blood vessel relaxant.
10
Therefore, low levels of HO-1could lead to endothelial dysfunction and hypertension, observed
in PE. Indeed, chorionic villous samples taken at 11 weeks of gestation showed a much lower
expression of HO-1 (46). Multiple reports have shown that women who smoke during pregnancy
are less likely to develop preeclampsia suggesting that inhaling exogenous CO can be protective
during PE (47). In fact, HO-1 is one of the few molecules capable of blocking release of sEng
and sFlt-1. Zhou et al. showed that HO-1 acts downstream of TNF-α to regulate sEng and sFlt-1
production (Figure 4) (48). It appears that HO-1 exerts its protective function by activating the
PI3K/Akt pathway (49, 50).
Recently, the PI3K/Akt pathway has emerged as a key signalling cascade in preeclampsia.
Firstly, activation of PI3K/Akt stimulates the expression of both VEGF and VEGF receptors
through HIF-1α transcription factor. Therefore, downregulation of PI3K/Akt activity can
potentially exacerbate decreased VEGF signalling, whereas lack of VEGF leads to reduction in
VEGF and VEGFR-1 expression, which in turn results in even more complete blockade of
VEGF signalling (51, 52). Activation of the PI3K/Akt pathway is required for endothelial cell
homeostasis and survival after vascular injury. In fact, several molecules including VEGF
promote endothelial cell survival through the PI3K/Akt pathway(53). Moreover, inhibition of the
PI3K-Akt pathway was shown to directly suppress sFlt-1 secretion from human placenta hypoxia
models (54).
1.2.2.4 The Angiotensin II type I Agonistic Autoantibodies (AT1-AA)
The circulating renin-angiotensin system (RAS) is a signalling cascade, which regulates blood
pressure and electrolyte balance. It starts with secretion of renin, which cleaves angiotensinogen
into angiotensin I (ANG I). While ANG I is not biologically active, it can be cleaved into ANG-
II by angiotensin-converting enzyme (ACE). ANG-II then can bind one of two angiotensin
receptors: AT1 and AT2, expressed on a number of different cell types, including vascular
smooth muscle cells and placental vasculature. Once bound, ANG-II initiates a signalling
cascade to induce vasoconstriction by inducing adolsterone production. Adolsterone increases
reabsorption of water and ions in the kidney, leading to an increase in blood pressure (55). As
ANG-II can readily increase blood pressure during preeclampsia, several studies compared its
concentration between normal and preeclamptic pregnancies. Paradoxically, ANG-II was present
in much lower quantities in preeclamptic women compared to healthy controls (56).
11
However, in a pioneering study done by Wallukat et al., researchers found increased levels of
AT1-AA in individuals with PE compared to healthy pregnancies. These autoantibodies were
shown to bind and activate the AT1 receptor (57). Subsequent studies were done to elucidate the
function of these antagonistic autoantibodies in preeclampsia. AT1-AA are present between 18-
22 weeks of pregnancy, before the onset of the clinical symptoms. The mechanism of action of
these antibodies is complex. While the direct mechanism of activating AT1 receptor certainly
plays a major role in inducing hypertension, recent studies have shown that AT1-AA also induce
expression of pre-pro-endothelin, a very potent vasoconstrictor (58). Moreover, the serum of rats
injected with AT1-AA, contained higher levels of both sFlt-1 and sEng, suggesting that AT1-AA
themselves could potentially initiate the release of these factors into the circulation (Figure 4)
(59). Additionally, these antibodies were shown to interfere with the invasive ability of human
trophoblast in vitro, suggesting that AT1-AA could be responsible for the initial ischemia as well.
In fact, administration of AT1-AA, isolated from preeclamptic patients, was shown to induce
preeclampsia in pregnant rats (60). The important role of AT1-AA was further demonstrated by
interfering with auto-antibody production in several animal models of preeclampsia, resulting in
a significant reduction in blood pressure (60).
Though AT1-AA are not specific for preeclampsia and occur in many other hypertensive
disorders, the mechanism involved in the production of these antibodies remains unclear. Rats
exposed to the chronic reduced uterine perfusion pressure (RUPP) model of preeclampsia tested
positive for AT1-AA. Therefore, ischemia itself appears to promote the generation of these
antibodies (61).
1.2.2.5 Generation and Function of Soluble Endoglin
Soluble endoglin is also increased in the serum of preeclamptic women. In fact, during PE there
is a 5-fold increase in sEng levels. sEng starts to rise around 8 to 10 weeks before the onset of
preeclampsia and correlates better with disease severity than sFlt-1 (17). However, unlike sFlt-1,
for which alternate splice variants have been described, none have thus far been reported for
sEng. Based on work from tumor cells, it was proposed that sEng is cleaved from plasma
membrane endoglin by the metalloprotease MMP14 (or MT1-MMP) and released into the serum
(62). Multiple mechanisms have been proposed to explain sEng generation from the placenta.
AT1-AA are thought to induce production of sEng by upregulating TNF-α expression.
12
Figure 4. Model of sEng and sFlt-1 generation by AT1-AA.
AT1-AA bind to the AT1 receptor and induce expression of TNF-α. TNF-α can then trigger
secretion of sEng and sFlt-1, which cause endothelial dysfunction and impaired angiogenesis.
Interestingly, HO-1 can block TNF-α-induced sEng and sFlt-1 generation.
Figure adapted from Zhou CC, Irani RA, Zhang Y, Blackwell SC, Mi T, Wen J, Shelat H,
Geng YJ, Ramin SM, Kellems RE, Xia Y. 2010. Angiotensin receptor agonistic autoantibody-
mediated tumor necrosis factor-alpha induction contributes to increased soluble endoglin
production in preeclampsia. Circulation 121: 436-44.
13
Indeed, several studies have showed that TNF-α upregulates a number of metalloproteases,
which in turn could be responsible for cleaving sEng from the cell surface (48). Specifically,
MMP14 has been shown to be expressed in preeclamptic placentae and capable of generating
sEng by cleaving proteins from the membrane (63). Interestingly, Hawinkels et al. used site-
directed mutagenesis to determine that endoglin is cleaved at the G-L bond at positions 586-587,
releasing the complete extracellular domain into the circulation (62). However, soluble endoglin
that has been isolated from the serum of preeclamptic women is resolved by SDS-PAGE as a
65kDa band, as opposed to the 80kDa mass of the recombinant ∆586 protein (4) (Figure 5). It is
therefore unlikely that the circulating soluble form of endoglin (sEng) observed in PE, is
produced by MMP14 cleavage at position 586. Mass spectrometry analysis of purified endoglin
from sera of PE patients, could only identify peptides up to Arginine 406 (4), missing the C-
terminal portion (Figure 5). The exact structure of soluble endoglin associated with PE, remains
to be established and current findings with recombinant (∆586) will have to be reexamined when
the true nature of sEng produced by the placenta during PE is identified.
The functional characteristics of sEng were described in a study by Walshe et al.where mice
were infected with an adenovirus, engineered to induce the expression of recombinant sEng
(∆586). An increase in sEng was shown to contribute to endothelial dysfunction by impairing
perfusion, vasodilatation and barrier function (64). Additionally, sEng was shown to inhibit
TGF-β1 –induced phosphorylation of eNOS and to enhance lung and liver microvascular
permeability (4). In that same study, injection of sEng (∆586) alone resulted in a small increase
in arterial pressure, however, combination of sFlt-1 and sEng exacerbated hypertension and
proteinuria induced by sFlt-1 alone. The authors concluded that sEng interferes with TGF-β1 and
TGF-β3 binding to the receptor, thus blocking eNOS phosphorylation and subsequent
vasodilatation (4).
Though some papers have reported the ability of sEng to interfere with TGF-β1 signalling (4),
recent studies have shown that recombinant sEng (∆586) has little to no affinity for TGF-β
molecules, but rather has high affinity for another TGF-β superfamily ligand, namely bone
morphogenic protein BMP9 (and related BMP10) (65). Overall, recombinant soluble endoglin
appears to have a destabilizing effect on the vascular endothelium, which probably contributes to
the pathogenesis of preeclampsia.
Figure 5. Structure of different forms of endoglin
Membrane endoglin (A) is a homodimeric glycoprotein
consists of 658 amino acids and
90kDa. It is composed of an orphan domain (with unknown structural features) followed by a
zona pellucida (ZP) region (generally
transmembrane region and a short
many investigators corresponds to the 586 amino acids of the extracellular domain
transmembrane and cytoplasmic regions.
form of sEng (C), identified in the sera of preeclamptic patients was shown to contain at least
406 amino but is likely a bit larger to correspond to
(A)
(B)
(C)
Figure 5. Structure of different forms of endoglin.
is a homodimeric glycoprotein of 180kDa. Each polypeptide chain
and the glycosylated chain has a molecular mass of approximately
is composed of an orphan domain (with unknown structural features) followed by a
(generally involved in protein interactions), a juxtamembrane and
transmembrane region and a short cytoplasmic region. A recombinant soluble
corresponds to the 586 amino acids of the extracellular domain
cytoplasmic regions. Its molecular mass is around 80kDa.
f sEng (C), identified in the sera of preeclamptic patients was shown to contain at least
but is likely a bit larger to correspond to its estimated molecular mass of 6
14
. Each polypeptide chain
molecular mass of approximately
is composed of an orphan domain (with unknown structural features) followed by a
, a juxtamembrane and
soluble form (B) used by
corresponds to the 586 amino acids of the extracellular domain, and lacks the
is around 80kDa. The physiological
f sEng (C), identified in the sera of preeclamptic patients was shown to contain at least
molecular mass of 65kDa.
15
1.3 Overview of TGF-β Receptors and Signal Transduction
1.3.1 TGF-β superfamily
The TGF-β superfamily, including TGF-βs, BMPs, activins and other related proteins, plays a
key role in cell proliferation, migration, differentiation and survival. More than 30 proteins have
been identified, all of which share the same cysteine knot structure (66). All ligands are
synthesized as pre-proproteins and require the pro-domain to form dimers. Therefore,
dimerization occurs intracellularly, before pro-domains are cleaved; proteins are then secreted
from the cell (67). All ligands signal through a heteromeric complex of Type I and Type II
serine/thereonine kinase receptors. Five type II receptors and seven type I receptors (also known
as Activin receptor like kinases or ALK), have been identified in human. Though most of the
ligands were shown to have affinity for either type I (BMP-2, 4, 7, 9) or type II (TGF-β1, TGF-
β3) receptors, some can only bind to an already formed type I – type II receptor complex,
suggesting that some receptors can associate with each other on the cell surface in the absence of
ligand.
Moreover, TGF-β receptors are often found in a complex with type III receptors also referred to
as co-receptors (betaglycan and endoglin). Betaglycan and endoglin are two known co-receptors
with some structural and functional homology (68). These co-receptors have been shown to
modulate cell signalling depending on the ligand and cell type in question (69). For example,
betaglycan facilitates interaction of TGF-β2 to its receptor TβRII (70). Endoglin on the other
hand, was shown to generally enhance signalling of TGF-β1/β3 and BMP9 (71).
1.3.2 TGF-β receptor signalling
1.3.2.1 Initial binding to the receptor
TGF-β1 has a high affinity for its type II receptor, TβRII. All type II receptors are constitutively
active and can undergo autophosphorylation. Phosphorylation of Ser213 and Ser409 has been
identified as essential for TβRII kinase activity, while phosphorylation of Ser416 may have an
inhibitory effect (72). For a long time, it was though that in the absence of ligand, each receptor
existed on the cell surface as a homodimer. However, recent studies on TβRI and TβRII, using
single molecule imaging and total internal reflection fluorescence microscopy technology,
clearly showed that receptors are monomers, and dimerize upon ligand binding (73, 74). Once
16
ligand is bound, TβRII associates with and phosphorylates type I receptor in its GS domain
(TTSGSGSG). Wieser et al, have shown that mutations of this sequence strongly inhibit smad-
dependent signalling initiated by TGF-β, therefore this phosphorylation is central to TGF-β
receptor activity (75). Interestingly, a number of phosphatases have been shown to modulate
TGF-β receptor signalling. Protein Phosphatase 1 (PP1) was found to inhibit TβRI activity and
be upregulated by TGF-β signalling, thus serving as a negative feedback mechanism. PP2A was
found to have dual effects on TBRI activity, depending on which regulatory subunit of the
phosphatase was activated (76). As of yet, there are no reports of phosphatases that affect TβRII
activity (76). Additionally, ALK5 expression was also shown to be regulated through the
ubiquitination pathway (77).
Though TGF-β1 binds primarily to TβRII and recruits the ALK5 receptor, it was also shown to
associate with another type I receptor – ALK1. While ALK5 is present almost universally on
most cell types, ALK1 is mainly present in endothelial cells; therefore co-activation of ALK1
and ALK5 receptors by TGF-β occurs primarily in the endothelium. Mutation studies have
shown that TGF-β itself does not bind to ALK1, but rather that ALK5 activation, by TGF-β
binding, allows association between two type I receptors. The kinase activity of ALK5 was
required for efficient ALK1 activation (78). Therefore, in endothelial cells, TGF-β1 can activate
both ALK5 and ALK1 type I receptors; however, ALK5 activation is necessary for ALK1
function (Figure 6).
Other members of the TGF-β superfamily, capable of activating ALK1 receptors are BMP9 and
the structurally related BMP10. Unlike TGF-β, BMP9 and BMP10 have a strong affinity towards
the type I receptor ALK1 (79). ALK1 associates with a type II receptor, BMPRII or Activin type
II receptor (ActRII) to mediate BMP9 and BMP10 effects (79). Phosphorylation of the type I
receptor leads to a conformational change and recruitment and phosphorylation of Smad1/5/8
proteins.
Type III receptors, or co-receptors, have been shown to mediate type I receptor signalling;
specifically, endoglin was shown to modulate TGF-β1 and TGF-β3 responses; however, the
exact mechanism of this process is uncertain. Both, intracellular and extracellular domains of
endoglin physically associate with TβRII and ALK5 receptors (80). In fact, studies have shown
that endoglin can bind TGF-β1 and TGF-β3 with high affinity in the presence of TβRII (81).
17
Figure 6. Function of endoglin with respect to TGF-β1 signalling in endothelial cells.
Endothelial cells contain both ALK1 and ALK5 receptors, meaning that TGF-β1 can induce
both Smad2/3 and Smad1/5/8 phosphorylation. The Type III receptor, endoglin, was shown
to modulate TGF-β1 signalling. Specifically, endoglin can inhibit ALK5 activation and
subsequent Smad2/3 phosphorylation. ALK5 signalling is thought to have quiescent effects
in ECs. ALK1 signalling, on the other hand, appears to be enhanced by the presence of endoglin
on the cell surface. Therefore, endoglin potentiates ALK1 and Smad1/5/ 8 activation.
Figure adapted from Fonsatti E, Nicolay HJ, Altomonte M, Covre A, Maio M. 2010.
Targeting cancer vasculature via endoglin/CD105: a novel antibody-based diagnostic and
therapeutic strategy in solid tumours. Cardiovasc Res 86: 12-19.
18
However, it appears that endoglin function is very complex, as it was shown to enhance TGF-
β1–induced Smad1/5/8 phosphorylation, but to interfere with TGF-β1–induced Smad2/3
phosphorylation (82). Moreover, endoglin is extremely important for BMP9 signalling, as
endoglin deficient endothelial cells were unable to induce Smad1/5/8 phosphorylation in
response to BMP9 (83). Since both BMP9 and TGF-β1 induce Smad1/5/8 phosphorylation
through ALK1 receptor, it appears that endoglin facilitates ALK1 and blocks ALK5 signalling
(Figure 5). However, other studies have reported that endoglin has no effect on Smad2/3
phosphorylation (84), while Santibanez et al. demonstrated that endoglin enhances ALK5
signalling (85). Therefore, the function of endoglin is extremely cell type and context dependent.
1.3.2.2 Smad Pathways
Smads are intracellular proteins, which relay signalling from TGF-β ligands to the nucleus.
There are three types of Smads: R-Smad (receptor regulated), Co-Smad (common mediator) and
I-Smad (Inhibitory). Smads1, 2, 3, 5, 8 are R-Smads and can be phosphorylated by type I
receptors. Binding of Smad proteins to the type I receptors is aided by the presence of Smad
anchor for receptor activation (SARA) protein. Once phosphorylated, Smads dissociate from the
receptors and form heterotrimeric or dimeric complexes with Smad4. All R-Smads have two
conserved domains, termed mad homology (MH) domains 1 and 2 (86). In all R-Smads, with the
exception of Smad2, MH1 domain is responsible for DNA binding, while MH2 domain mediates
protein-protein interactions. Specifically, the MH2 domain is required for interaction with TβRI
receptor. Interestingly, the same region of Smad has been implicated in other protein-protein
interactions, suggesting that TGF-β signalling is achieved through a sequential process of
competitive interactions (87).
Upon phosphorylation, Smad complexes translocate to the nucleus, where they are able to
interact with transcription factors and be recruited to specific promoter elements. Usually, each
type I receptor is associated with only a subset of Smads. For example, ALK4, ALK5, ALK7 all
phosphorylate Smad2 and Smad3, while ALK1, ALK2, ALK3 and ALK6 induce Smad1, 5, 8
phosphorylation (88). All heteromeric complexes of Smads contain Smad4, implying that Smad4
is central to TGF-β signalling. Yet, Smad4 null mice and a number of Smad4 lacking tumors
show TGF-β-induced transcription. It is not known whether this is mediated by other Smads, or
represents Smad-independent signalling (89). While R-Smads and Co-Smads are present
19
universally in almost all if not all cell types, I (inhibitory)-Smads expression is highly regulated.
In fact, TGF-β receptor signalling is a known activator of Smad6 and Smad7 expression,
suggesting that I-Smads act as a negative feedback loop to control TGF-β signalling (90-92).
TGF-β1 activates ALK5 signalling cascade, meaning that its primary pathway goes through
phosphorylation and dimerization of Smad2/3. The dimer then associates with Smad4 and
translocates to the nucleus to initiate gene activation. In endothelial cells, however, TGF-β1 was
shown to also signal through ALK1, activating the Smad1/5/8 signalling cascade (78, 93).
Extensive research showed that though ALK1 is the primary receptor, which initiates Smad1/5/8
phosphorylation, the presence of active ALK5 receptor is an absolute requirement for this
process. It appears that binding of TGF-β1 to ALK5 receptor initiates the association of ALK5 to
ALK1, which in turn induces Smad1/5/8 signal transduction (78). BMP9, on the other hand, was
shown to bind mostly the type I receptor ALK1 and occasionally ALK2. ALK1 activation leads
to Smad1/5/8 phosphorylation; therefore, both TGF-β1 and BMP9 are capable of activating the
ALK1 receptor and induce Smad1/5/8 phosphorylation (83).
1.3.2.3 Non-canonical pathways
P38 pathway
The TGF-β receptor family not only induces signalling through Smad proteins, but is also
capable of activating other signalling molecules such as MAPKs, ERK, p38 and others (Figure
7). Perhaps the most recognized non-canonical pathways initiated by TGF-β are p38 MAPK and
JNK pathways. Both p38 and JNK exist at the 3rd
level of MEK phosphorylation, meaning that
there are least two sequential phosphorylation events before p38 or JNK can become
phosphorylated. It is thought that JNK is activated through MKK4 and p38 MAPK through
MKK3/6 (94). Further upstream, both MKK4 and MKK3/6 are activated through a TGF-β
activated kinase (TAK1). Multiple studies have shown that TAK1 deficient cells are unable to
signal through p38 or JNK pathways (95, 96). Though Watkins et al showed that TAK1 can
directly interact with TβRII, most studies agree that TRAF6 is needed for efficient recruitment of
TAK1 to the receptor (97). Interestingly, JNK and p38 pathways are completely independent
from Smad signalling, as Smad2/3 or Smad4 deficient cells are perfectly capable of activating
these pathways (98). Recent studies revealed that TGF-β receptors cannot only be
phosphorylated on their serine/threonine resides, but can also be activated through
20
Figure 7. Canonical and Non-Canonical signalling pathways of TGF-β.
Upon binding to its receptor, TGF-β1 induces phosphorylation of TβRII and TβRI, which leads
to recruitment, phosphorylation and dimerization of Smad2/3 molecules. This dimer then
interacts with Smad4 and the whole complex translocates to the nucleus to activate gene
transcription. This pathway is referred to as the canonical TGF-β1 pathway. Additionally, TβRI
can also activate PI3K/Akt pathways as well as TAK, which would in turn phosphorylate JNK
and p38 proteins. Moreover, activation of TβRII can also lead to RAS/RAF activation and
subsequent ERK phosphorylation. These non-Smad pathways are usually called non-canonical
pathways.
Figure adapted from Connolly EC, Freimuth J, Akhurst RJ. 2012. Complexities of TGF-beta
targeted cancer therapy. Int J Biol Sci 8: 964-78.
21
phosphorylation of tyrosine residues. TβRII cytoplasmic domain contains three tyrosine residues,
which upon phosphorylation can recruit scaffold proteins, which in turn initiate non-canonical
signalling pathways (99).
ERK pathway
Several studies have shown that TGF-β can induce ERK activation. Interestingly, in some cells
ERK phosphorylation occurs rapidly within minutes of TGF-β binding, suggesting a direct
method of recruitment (100). In contrast, other cells show a delayed response, happening hours
after stimulation, implying that protein synthesis is required for signalling to occur (101). As in
the case of p38 signalling, phosphorylation of tyrosine residues on TβRII plays a crucial role in
ERK signalling. Moreover, tyrosine residues on TβRI can also be phosphorylated by TβRII,
which then serve as docking sites for Grb2 and Shc proteins. Grb2 protein is usually in a
complex with Sos protein and therefore brings Sos to the TβRI – TβRII complex. The
ShcA/Grb2/Sos complex is then capable of activating RAS at the plasma membrane, resulting in
activation of c-RAF, MEK and ERK (102). Mutations in ShcA have been shown to interfere with
ShcA /Grb2 interactions, causing decreases in ERK phosphorylation. Therefore, it appears that
ShcA is essential for efficient ERK signalling and actively holds the ShcA/Grb2/Sos complex
together (102). Moreover, as in the case with p38, Smad involvement is not needed to induce
ERK phosphorylation (103).
Akt pathway
Several findings suggest that TGF-β can also activate PI3K/Akt pathway in a Smad independent
manner (104). The exact mechanism of this activation is not yet clear; however, recent studies
have shown that TβRI can indirectly associate with one of the p85 subunit of PI3K (105). It is
thought that PI3K/Akt pathway signals downstream to activate mammalian target of rapamycin
(mTOR), which is a key regulator of protein synthesis. Inhibition of either PI3K or Akt
phosphorylation showed a significant decrease in mTOR activation level, suggesting a direct link
between PI3K/Akt and mTOR pathways (106). Interestingly, TGF-β-induced apoptosis and
growth inhibition, which are Smad-dependent processes, can be attenuated by PI3K/Akt
activation, implying that two pathways are in direct conflict with one another. Later studies have
22
shown that phosphorylated Akt can directly bind to Smad3 and prevent its translocation to the
nucleus, thus inhibiting Smad3 signalling (107).
BMP9 signalling
BMP9 can also activate non-Smad pathways; however, no extensive research has been done to
elucidate the exact mechanisms of Smad-independent pathways. It is likely that BMP9 initiates
non-canonical signalling similarly to TGF-β, by phosphorylating tyrosine residues allowing for
protein docking (108). Several papers have shown that BMPs, including BMP9, can induce
MAPK pathways, through TAK1, leading to p38 MAPK or JNK activation. Additionally, BMPs
were also shown to induce ERK phosphorylation (109). Interestingly, BMP9 was also found to
have an inhibiting effect on PI3K/Akt pathway, which is activated by TGF-β signalling,
implying that ALK1 and ALK5 pathways could function in opposition to each other (110).
1.3.3 Effects of TGF-β superfamily signalling
1.3.3.1 Functional role of TGF-β1
Members of TGF-β superfamily have multiple and diverse functions, including but not limited to
proliferation, differentiation, migration, wound repair and inflammation (86). There are three
isoforms of TGF-β: -β1, -β2 and -β3. They have very similar properties in vitro, but appear to
have different functions in vivo (111). TGF-β1 is secreted as a proprotein, the prodomain being
required for dimerization. Upon release, the prodomain is usually cleaved from the protein, but
remains non-covalently attached to TGF-β1, inhibiting the ability of TGF-β1 to bind to its
receptor. Dimeric TGF-β attached to the prodomain, is called latency-associated protein (LAP).
Early in the assembly of TGF-β1, LAP is in complex with latent TGF-β binding protein (LTBP)
through disulfide bonds. LTBP was shown to have covalent interactions with the ECM, tethering
TGF-β1 in the extracellular matrix and preventing it from entering the bloodstream. It is
therefore thought that the major pool of TGF-β1 exists in the extracellular matrix, while only a
fraction of TGF-β1 is in circulation (112). Indeed, the concentration of TGF-β1 in plasma is only
around 1ng/ml (113).
One of the most acknowledged functions of TGF-β proteins is in wound repair. Upon injury,
TGF-β1 is rapidly secreted by macrophages, platelets and keratinocytes (114). TGF-β1 is
thought to be important for cell migration during wound repair, as well as production of ECM
23
proteins, which provide structural integrity for the wound (115). Many animal models with
impaired wound repair show reduced expression of TGF-β, and usually benefit greatly from
exogenous administration of TGF-β (116). Paradoxically, recent studies found that Smad3 null
mice experience accelerated wound healing (117). Subsequently, multiple studies have reported
that blocking TGF-β signalling leads to a significantly faster wound repair (118-120). Therefore,
the intricacies of how wound repair is regulated by TGF-β signalling remain to be determined.
Additionally, TGF-β plays an important role in immunity. TGF-β was shown to inhibit IL-2
production in T cells, thus significantly inhibiting T cell proliferation (121). Moreover, TGF-β is
a known inducer of FoxP3 during naïve T cell activation, forcing T cells into regulatory T cell
(Treg) lineage (122). Overall, the majority of the studies agree that TGF-β has strong
immunosuppressive properties (123).
The importance of TGF-β in angiogenesis is highlighted by the fact that mice deficient in TβRII,
ALK1, ALK5 or endoglin, experience a very similar phenotype, showing lethal abnormalities in
cardiovascular development (124). ALK5 activation usually blocks angiogenesis by up-
regulating fibronectin and plasminogen activator inhibitor type 1 (PAI-1), which has been shown
to inhibit EC migration (125). Surprisingly, TGF-β-induced ALK1 activation stimulates EC
migration, proliferation and tube formation (78). In fact, multiple studies have reported that
ALK1 and ALK5 pathways act in opposition to each other in terms of ECM production (126)
and endothelial cell proliferation and migration (78). The mysterious nature of TGF-β is further
outlined in cancer pathogenesis, where TGF-β was shown to be highly beneficial at the early
stages by inducing cell cycle arrest and apoptosis. However, during the late metastatic stages,
cells become resistant to quiescent effect of TGF-β1 and start benefiting from its angiogenic
properties (127).
1.3.3.2 Functional role of BMP9
BMP9 is also known as GDF2 (Growth and differentiation factor 2) and has been identified as a
ligand for the ALK1 receptor (83). Like all other TGF-β1 ligands, BMP9 is synthesized as a
large pre-pro-protein. Once dimerization occurred, the pro-domain is cleaved from the active
BMP9; however, it still continues to be attached to BMP9 through non-covalent interactions
(128). Bidart et al. looked at the expression of BMP9 in different organs and found that
hepatocytes are by far the best producers of BMP9. Moreover, they found that similarly to TGF-
24
β, BMP9 circulates in active and non-active forms. However, unlike TGF-β, the pro-domain of
BMP9 does not seem to bind to ECM, meaning that BMP9 is not inhibited to enter the
circulation, resulting in a much higher concentration of BMP9 in the serum. The estimated level
of BMP9 in the blood is around 5-10ng/ml, which is much higher than the EC50 for ALK1
activation. Indeed, aortic endothelial cells showed endogenously phosphorylated Smad1/5/8
proteins, presumably due to sustained activation by circulating BMP9 molecules (129).
The functional properties of BMP9 are poorly understood. In fact, BMP9 knockout mice did not
show defects in angiogenesis (130), even though multiple studies identified BMP9 as an
important factor for vasculature maintenance (79). Further studies revealed that another protein,
BMP10, is present at high levels in the circulation and can also bind to the ALK1 receptor and
can function as a substitute for BMP9 (130). Knockout of the BMP9/10 receptor, ALK1, is
embryonically lethal, while haploinsufficiency in ALK1 leads to a genetic disorder called
hereditary hemorrhagic telangiectasia type 2 (HHT2), which is characterized by abnormal vessel
formation (131). Knockout of endoglin, which also interacts directly with BMP9/BMP10, also
leads to a very similar disease, HHT1.
Early studies described BMP9 as a vascular quiescence factor (132). Subsequent papers used
constitutively active ALK1 to confirm that BMP9 inhibits VEGF-induced angiogenesis and
endothelial cell proliferation (79). Recent studies, however, reported the ability of BMP9 to
induce proliferation of multiple types of endothelial cells both in vitro and in vivo, most likely by
up-regulating VEGF receptors and Angiopoetin-1/Tie2 expression (133). These findings are not
surprising, as ALK1 signalling up-regulates ID-1 and ID-3 protein expression (134), which are
strong promoters of angiogenesis (135). Interestingly, the soluble form of ALK1 was recently
reported to inhibit tumor growth in mouse models (136). As with other TGF-β superfamily
ligands, the functional role of BMP9 remains controversial and is likely extremely cell and
context dependent.
Apart from angiogenesis, BMP9 was also found to have many other effects in vitro and in vivo.
As most other BMPs, BMP9 can function as an osteogenic and chondrogenic factor (137).
Additionally, BMP9 can also regulate metabolism, by inhibiting glucose production and
upregulating key enzymes of lipid homeostasis (138). Moreover, two recent papers reported that
BMP9 can induce secretion of Endothelin-1 (ET-1), a very potent vasoconstrictor (139, 140).
25
1.4 Endothelin-1
1.4.1 Structure and Function of ET-1
1.4.1.1 Biosynthesis of ET-1
Endothelin was first described in 1985 as an endothelium-derived constrictor factor (141). Since
then, 3 different isotypes of ET-1 have been found, termed ET-1, ET-2 and ET-3. ET-1 is by far
the most prominent form, and responsible for the majority of functions exerted by ETs (142).
Endothelin isoforms are highly homologous; in fact ET-2 and ET-3 differ from ET-1 only by 3
and 6 amino acids respectively. All three are synthesized as pre-pro-proteins, with a 17 amino
acid leader sequence that targets pre-pro-ET-1 to the endoplasmic reticulum(143). In the cell, the
pre-domain (165 amino acids) is cleaved from pro-ET-1 by furin-like proteases, leaving 38
amino acids, termed Big-ET1. Big-ET1 is then further cleaved into active ET-1 (21 amino acids)
by endothelin-converting enzymes (ECE) (144). It is thought that activation of ET-1 by ECE is a
rate-limiting step in ET-1 production (145). ET-1 is mainly produced by endothelial cells in
relatively small quantities, but also by other cell types, such as cardiomyocytes and vascular
smooth muscle cells. The reported concentration of ET-1 in the serum was much below the EC50
of the ET-1 receptors, suggesting that ET-1 acts as a local agent (146).
1.4.1.2 The receptors of ET-1
There are four known receptors for endothelins called ETA, ETB1, ETB2 and ETC, all of which are
G-protein coupled receptors. ETA binds ET-1 and ET-2 with greater affinity than ET-3, while
ETB appears to bind all isoforms with the same strength. These receptors are expressed in
different levels in various cell types (147). Studies revealed that ETA receptors are mostly present
on vascular smooth muscle cells (VSMC) and upon activation initiate calcium ion influx and
subsequent vasoconstriction. ETB receptors, on the other hand, are mostly present in endothelial
cells, and induce release of NO and prostacyclin, causing vasodilatation (148). Typically upon
ET-1 injection, arterial pressure briefly goes down due to the extremely fast secretion of NO,
induced via ETB receptors, but then quickly rises due to the constriction effect via ETA receptors
(149). Interestingly, an isotype of ETB receptors has been recently found to be present in VSMC.
It was named ETB2 receptor, as opposed to ETB1 found on endothelial cells; upon activation,
ETB2 initiates contraction of the smooth muscle cells (150). Constriction by ETA receptor can
usually last for up to one hour. Studies have shown that vasoconstriction is initiated by prolonged
26
influx of Ca2+
from the extracellular space. Removal of ions from the extracellular space resulted
in the absence of sustained vasoconstriction (149). Interestingly, NO donors were found to
interfere with channels responsible for ET-1-induced calcium ion influx, implying that ETA and
ETB receptors might work in opposite ways (151). Moreover, several studies have reported that
NO decreases ET-1 production from endothelial cells, once again suggesting a delicate balance
between ETA and ETB receptors (152). Though it would seem logical to inhibit ETA receptor,
rather than inhibiting both types of receptors, most studies agree that inhibiting both ETA/B
receptors produces a much more effective vasodilatation (146). It is likely, that exclusive
inhibition of the ETA receptor, allows the ETB receptor on VSMC to mediate ET-1-induced
vasoconstriction.
1.4.1.3 Physiological role of ET-1
Endothelin-1 is one of the most potent vasoconstrictors known, second only to urotensin II (153).
In humans, ET-1 increases mean arterial blood pressure, reduces heart rate and causes sustained
vasoconstriction in the pulmonary vasculature (154). Since its discovery in 1988, multiple
studies have shown a link between ET-1 and various cardiovascular disorders, such as heart
failure, hypertension, renal failure and many others. Blockade of ETA receptor was shown to
result in an increase in forearm blood flow in healthy humans (155), while inhibition of ETB
signalling yielded local vasoconstriction. Therefore, it appears that endothelial cells are
constantly stimulated to induce production of ET-1, which plays an important role in
maintenance of vascular tone. The importance of ET-1 is highlighted by the fact that the ET-1
homozygous knockout phenotype is lethal with morphological abnormalities of craniofacial and
thoracic blood vessels (156). Mice deficient in receptors for endothelin-1 show a similar
phenotype (157). ET-1 heterozygous mice, paradoxically suffer from increased systemic blood
pressure, even though serum ET-1 concentration was found to be much lower than in wild type
mice. The most probable explanation is the fact that the ETB receptor is no longer activated,
leading to decreased in NO production and subsequent hypertension (156, 158). This is further
strengthened by the fact that mice deficient in ETB receptor also experience hypertension (158).
Apart from its cardiovascular effect, ET-1 is also associated with inflammation processes. In
fact, ET-1 was shown to activate macrophages, leading to release of multiple pro-inflammatory
cytokines, including IL-1, IL-6 and TNF-α. In turn, both TNF-α and IL-6 were able to induce
27
ET-1 secretion from endothelial cells (159). It appears that ET-1 works in concert with TNF-α to
up-regulate adhesion molecules on ECs. Blockade of ET-1 receptors has been shown to
significantly decrease the accumulation of neutrophils and myeloperoxidase activity in the
ischemic myocardium (160).
1.4.2 Role of ET-1 in Preeclampsia
1.4.2.1 ET-1 levels in PE
Several studies have looked into ET-1 as being a potential contributor to preeclampsia-associated
hypertension. Most studies have found a 2 to 3 fold increase in ET-1 concentration in the serum
of preeclamptic women and at least one study found a correlation between ET-1 levels and
severity of the disease (161, 162). Additionally, Nishikawa et al. looked at serum nitrate levels in
preeclamptic patients and showed a negative correlation between ET-1 levels and those of
vasodilators NO and cGMP (162). Another group compared the expression levels of ECE and
found it was up-regulated in preeclamptic women compared to healthy controls, implying that
increased activity of ECE might be contributing to the increase in ET-1 production (163).
Moreover, ET-1 was shown to induce oxidative stress and inhibit proliferation of JEG-3 cells (a
trophoblast line), suggesting that increased levels of ET-1 may contribute to extravillous
trophoblast differentiation in vivo (164).
1.4.2.2 ET-1 in preeclampsia models
There are several animal models, which were developed to study preeclampsia. One of the most
widely used models is RUPP (Reduced uterine perfusion pressure), to partially block blood flow
to the uterus at the later stages of gestation (165). As a result, the placenta experiences acute
hypoxia and ischemia, an approximation of what happens when the trophoblast fails to invade
the spiral artery. This model was applied to a wide variety of animals, including rodents, dogs
and non-human primates. Most studies reported that animals were experiencing all hallmark
symptoms of preeclampsia, such as elevated sFlt-1, renal injury and endothelial dysfunction
(165-167). Other studies looked at the expression of ET-1 in RUPP animals and concluded that
pre-pro-endothelin mRNA was significantly elevated compared to healthy pregnant controls
(168, 169). Moreover, endothelial cells incubated with serum from RUPP rats secreted
significantly more ET-1 when compared to cells incubated with healthy control serum (170). The
28
most interesting finding, however, was that hypertension induced by this model can be
completely abolished using an ETA specific antagonist (171).
Another method to induce PE-like disease in animals is to introduce exogenous sFlt-1 molecules.
Several groups reported proteinuria, hypertension and renal failure in rats infused with
recombinant sFlt-1 or using an adenovirus for delivery of the gene (3). One study looked at ET-1
production in these animals and reported significantly higher levels of ET-1 mRNA in renal
cortex. They also looked at the ability of ETA antagonist to relieve sFlt-1-induced hypertension
and found that by blocking ET-1 signalling could completely abolish the hypertensive response
(18).
In yet another model, AT1-AA are delivered into pregnant mice, leading to preeclampsia-like
symptoms. In the same study, the authors reported an increase in ET-1 mRNA levels, suggesting
that induced hypertension was in part due to ET-1 signalling (58). Zhou et al. looked into the
mechanism of ET-1 secretion and found that AT1-AA induced TNF-α expression, which leads to
IL-6 expression, which is necessary for ET-1 production. In fact, ETA antagonists were able to
completely abolish the induced hypertension, implying a key role for ET-1 in this model of
preeclampsia. Additionally, inhibition of either TNF-α or IL-6 expression resulted in normal
pregnancy with no hypertension and no ET-1 up-regulation (172).
1.4.2.3 Anti-ET-1 therapeutics
It is clear that ET-1 is a key mediator of hypertension in preeclampsia. Multiple studies have
shown that administration of ET-1 blocking agents can attenuate, if not completely abolish, PE-
associated hypertension (171-173). However, genetic studies with ET-1 or ETA receptor deficient
mice showed a number of serious birth defects. In fact, administration of endothelin receptor
antagonists is contraindicated during pregnancy (174). Other groups looked into the possibility
of administrating ET antagonists during specific times in pregnancy; however, more studies are
needed to evaluate the safety of this method (175). In summary, it is evident that ET-1 plays a
crucial role in preeclampsia. However, direct inhibition of ET-1 signalling might not be the best
way to manage rising blood pressure. Indirect methods of inhibiting ET-1 secretion might prove
safer and more beneficial in fighting preeclampsia-associated hypertension.
29
1.5 Thesis Objectives
1.5.1 Rationale
Preeclampsia is responsible for more than 15% of all premature births in North America, yet
pathogenesis of the disease is poorly understood and currently there is no cure but to deliver the
baby prematurely. The crucial role of ET-1 in preeclampsia-associated hypertension has been
documented by several groups; however, the exact mechanism underlying increased ET-1
generation is not established. Recent reports that some members of TGF-β superfamily can
induce ET-1 secretion raise the possibility that they could be responsible for increased levels of
ET-1 during preeclamptic pregnancies. Moreover, recent findings have shown that sEng binds
BMP9 with high affinity, creating an intriguing possibility that sEng production could be
beneficial for preeclampsia patients by blocking BMP9 signalling and subsequent ET-1
generation. Therefore, we investigated the ability of endothelial cells to produce ET-1 in
response to TGF-β1 and BMP9. Additionally, we examined signalling pathways responsible for
ET-1 secretion and looked at the effects of sEng and VEGF, two known factors involved in PE,
on BMP9 – and TGF-β1 –induced ET-1 secretion.
1.5.2 Hypothesis
sEng counteracts sFlt-1-induced hypertension by blocking BMP9 signalling and subsequent
secretion of ET-1.
1.5.3 Specific Objectives
(I) To test the effects of sEng and VEGF on BMP9 – and TGF-β1 –induced secretion of
ET-1 from mouse endothelial cells.
(II) To examine and compare different signalling pathways responsible for ET-1 secretion
from endothelial cells stimulated with either BMP9 or TGF-β1.
30
2 Materials and Methods
2.1 Cell culture
Mouse embryonic endothelial cells (EC) were a generous gift from Dr E. Dejana (Milan, Italy).
They were derived from mouse embryos at E9.0 and immortalized using polyoma middle T
antigen (176). Cells were cultured in MCDB131 medium plus 1% glutamine, 1%
penicillin/streptomycin, 15% heat-inactivated fetal bovine serum (All GIBCO) and 25 µg/ml
endothelial mitogen (Biomedical Technology Inc.), as previously described (177).
2.2 Cell based stimulation and inhibition assays
Endothelial cells were seeded at 5∙105 cells per well on 6-well plates and incubated in complete
medium at 37oC and 5% CO2. After 48 hours, cells were washed with phosphate-buffered saline
(PBS, GIBCO) and incubated in serum-free media for 3 hours. After another wash, cells were
incubated in serum-free MCDB131 media with various inhibitors: increasing concentrations of
dorsomorphin (10, 20, 40 µM; Bioscience), SB525334 (1, 2, 4 µM; Selleckchem), or BIRB796
(0.2, 0.5, 1 µM; Selleckchem) for 30 minutes. Cells were then treated with 40 pM of either TGF-
β1 or BMP9 for another 30 minutes.
2.3 Western Blotting
Cell lysates were obtained by scraping cells from the 6-well plates and solubilizing in 500 µL of
cell lysis buffer (0.05 M Tris-HCl pH 7.4, 0.01 M NaCl, 1% Triton X-100, 0.01 M sodium
pyrophosphate, 0.25 M NaF, 0.001 M Na3VO4; all from Sigma), supplemented with 1%
ProteoBlockTM protease inhibitor cocktail (Fermentas). Samples were then rotated for 30
minutes at 4oC and centrifuged at 10,000g for 30 minutes at 4
oC, and the supernatants collected.
Protein concentration measurements were done in duplicate according to the protocol from
BioRad Protein Assay II kit. Once concentration was established, a 20 µg protein aliquot was
mixed with reducing buffer (Invitrogen), supplemented with 62 µg/mL DL-Dithiothreitol, in 3:1
ratio. Samples were then heated at 95oC for 5 minutes, centrifuged for 1 minute at 10,000g and
fractionated on 10 or 12 well 4-12% Bis-Tris Gels (Invitrogen). Samples were fractionated at
150 volts for approximately 1 hour in SDS Running Buffer (NuPAGE MES, Novex, Invitrogen)
31
and transferred to a PVDF membrane in Transfer Buffer (0.025 M Tris pH 7.4, 0.192 M glycine,
20% methanol; all Sigma) at 4oC for 90 minutes at 200 amps.
Membranes were blocked in 5% milk (Bio-Rad) for 1 hour and incubated with one of the
following antibodies: Phospho-Smad1/5/8 (1:1000), Phospho-Smad2/3 (1:1000), Phospho-p38α
(1:1000), phospho-p38β (1:1000), purchased from Cell Signalling; total Smad2/3 (1:1000), total
Smad1/5/8 (1:1000), phospho-p38 (1:1000), purchased from Santa-Cruz, or β-actin (1:1000),
purchased from Sigma. Membranes were incubated overnight at 4oC, washed 3 times in TBS-T
buffer (0.02 M Tris base, 0.137 M NaCl, 0.1% Tween 20; Sigma) for 10 minutes and incubated
in secondary antibody (anti-mouse or anti-rabbit, purchased from GE Healthcare, 1:10,000
dilution in 5% milk). After 3 washes, membranes were incubated in Western Lighting TM
Chemi luminescent Reagent Plus and developed using Konica Minolta SRX-101A developer.
Radiographs were scanned and analyzed using FluorChem software. Stripping was done using 1
M glycine solution, pH=2.5 (Sigma).
2.4 Enzyme-Linked Immunosorbent Assay (ELISA) for ET-1
Cells were seeded in 6-well plates at 5·105 cells per well and incubated at 37
oC for 48 hours.
Cells were washed with PBS and starved for 12 hours in MCDB131 media without any
additives. Cells were then incubated for 30 minutes with inhibitors (dorsomorphin at 40 µM;
Bioscience, BIRB796 at 1 µM; Sellechchem, SB525334 at 8 µM; Sellechchem) and stimulated
with various ligands, including TGF-β1 (1 ng/ml, R&D), BMP9 (1 ng/ml, R&D), VEGF (50
ng/ml; R&D), soluble ALK1-Fc (125 ng/ml, R&D) and recombinant sEng [100 nM; this protein
was generated by Allison Gregory and contains only the extracellular region, 1-586 amino acids
(∆586)]. After 10 hours of stimulation, media was collected, centrifuged for 20 minutes at
10,000g, transferred to a new tube and stored at -80oC. Cell lysates were collected and tested for
protein concentration to ensure similar cell numbers in each sample. Protein concentrations
estimated in the various samples were within 10% of each other.
The ET-1 ELISA kit was used according to the protocol supplied by EnzoBiosciences. Briefly,
100 µL of the stored media samples were added to the 96-well plate coated with ET-1 antibodies
and incubated for 2 hours at 23oC. The plate was then washed 5 times in supplied washing buffer
and incubated with antibodies to ET-1 for 1 hour. Once again, the plate was washed 5 times and
32
visualizing solution was added for 30 minutes. Stop solution was then added to stop the reaction
and the plate was read at a wavelength of 450 nm. A standard curve was constructed using the
supplied ET-1 protein.
2.5 Biacore Biosensor Analysis
All experiments were done using BIAcore X (BIAcore, Uppsala, Sweden). First, research grade
CM5 sensor chips were activated using the standard NHS (N-hydroxysuccinimide)/EDC (N-
ethyl-N’-3-diethylaminopropyl) procedure. For coupling, proteins were diluted to 20 nM in 10
nM sodium acetate buffer, pH 6. Soluble receptors were then passed over the activated surface
until the desired amount of molecules was covalently attached to the chip. 1 M ethanolamine was
then used to quench any remaining active sites. The data were collected at 10 Hz using at least 5
injections of each ligand. Flow rate was maintained at 2-20 µL/min. Regeneration of the chip
with attached sALK1-Fc molecules was carried out using 3 M MgCl2. Regeneration of the chip
with sTβRII receptor was done using 0.4 M urea, 0.9 M MgCl2, 0.025 M EDTA, 0.9 M
guanidine-HCl, 0.01 M glycine, pH 9.5.
BIAevaluation 4.1 software was used to measure Kon and Koff of each interaction, and
subsequently the KD was calculated using these numbers. A hyperbolic curve was then fitted to
the graph. Concentration at which RU (relative units) difference was half-maximum represents
the affinity of the interaction (KD). The highest observed RU was taken as the maximum.
2.6 Statistical Analysis
Data are reported as mean ± SEM. Comparisons were performed by ANOVA, and significant
differences were evaluated post hoc using the Newman-Keuls method. Results are expressed as
the mean ± SEM with P< 0.05 representing significance.
33
3 Results
3.1 BMP9 and TGF-β1 induce ET-1 secretion in mouse embryonic endothelial cells
Recent studies have reported that BMP9 and TGF-β1 are potent inducers of ET-1 production in
human endothelial cells (139, 140). Since ET-1 is a major vasoconstrictor protein (141)
implicated in PE pathogenesis (178), I examined the ability of BMP9 and TGF-β1 to stimulate
ET-1 secretion in endothelial cells.
To better understand the kinetics of ET-1 release from EC, I exposed cells to increasing
concentrations of either BMP9 or TGF-β1 (Figure 8). Unstimulated EC secrete close to 15 pg/ml
of ET-1 in 10 hours. This value did not vary significantly throughout the experiments. When
treated with low concentrations of BMP9 (0.1, 0.5 ng/ml), EC showed only a slight increase in
ET-1 secretion; however, when stimulated with 1 ng/ml of BMP9, EC demonstrated a significant
2-fold increase in ET-1 secretion. EC secreted more ET-1 when treated with higher
concentrations of BMP9 (2, 4 ng/ml); however, doubling BMP9 concentration from 2 to 4 ng/ml
produced only a modest increase in ET-1 secretion (from 42 to 50 pg/ml), suggesting that near
maximum stimulation had been achieved. A software-constructed hyperbolic curved showed a
high degree of correlation with the acquired data, resulting in R2 = 0.99.
Treatment of EC with TGF-β1 also induced secretion of ET-1. However, unlike BMP9, low
concentrations of TGF-β1 had a pronounced effect on the ability of endothelial cells to secrete
ET-1. In fact, 0.4 ng/ml of TGF-β1 induced a significant increase in ET-1 levels. Cells reached
their maximum ET-1 secretion of 47 pg/ml when treated with 0.8 ng/ml of TGF-β1. Subsequent
increase in TGF-β1 concentration failed to induce a further rise in ET-1 secretion, suggesting
that saturation was achieved. A fitted hyperbolic curve showed good correlation with R2 = 0.86.
This is most likely due to the large standard errors in the linear portion of the TGF-β1 titration
curve, resulting from small changes in ligand concentration causing big changes in ET-1 release,
making numbers more variable. Overall, treatment with either BMP9 – or TGF-β1 –induced a
strong increase in ET-1 secretion in mouse embryonic endothelial cells.
34
Figure 8. BMP9 and TGF-β1 induce ET-1 secretion from mouse endothelial cells.
EC were starved for 12 hours in serum-free media and incubated with increasing concentrations
of BMP9 (A) or TGF-β1 (B) for 10 hours. Media were collected and tested for the presence of
ET-1 by ELISA and results are expressed as pg/ml media. A hyperbolic curve was fitted by the
software for each ligand, based on acquired results. Data are the mean ± SEM of 3 experiments,
all done in duplicate. * and # denote P< 0.05 versus control value.
0.0 0.5 1.0 1.5 2.00
10
20
30
40
50
60
#
#
##
ET-1 [pg/m
l]
[ng/ml]
R2 = 0.86
TGF-ββββ1
#
(A)
0 1 2 3 40
10
20
30
40
50
60 *
*
ET-1 [pg/m
l]
BMP9 [ng/ml]
R2= 0.99
*
(B)
35
3.2 sEng (∆586) binds with high affinity to BMP9, but not to TGF-β1
In contrast to sFlt-1, the exact sequence of the physiological form of sEng, how it is released
from the placenta and its role in PE remain relatively unknown. As a membrane protein,
endoglin was shown to physically associate with a number of TGF-β superfamily receptors and
to modulate responses to several TGF-β ligands (82, 177, 179). The function of endoglin as a
soluble protein, however, is poorly understood and a recombinant form of sEng (∆586; 80kDa
monomer) is used in most studies, which represents the complete extracellular domain of the
membrane protein, which is of larger size than the physiological form of sEng purified from PE
sera (65kDa monomer) (see Figure 5) (4). Thus, the complete amino acid sequence of the
physiological form of sEng remains to be determined. Several papers have proposed that sEng
interferes with TGF-β1 signalling (4, 180); however, recent studies indicate that recombinant
sEng (∆586) has a much higher affinity for BMP9 (65, 181).
To examine the functional properties of sEng, we used BIAcore X to evaluate binding affinities
of sEng to TGF-β1 and BMP9 ligands (Figure 9). As a proof of principle, the KD value between
two known high affinity receptor/ligand pairs was also measured. Binding of increasing
concentration of TGF-β1 (0.1 nM – 200 nM) to immobilized soluble TβRII dimer, yielded a KD
= 0.1nM. That is an extremely tight interaction; in fact, we were unable to strip TGF-β1 from
TβRII, which made our calculations less reliable. It is likely that the KD value was
underestimated (Figure 9A). Next, we examined the strength of binding between BMP9 and
recombinant sALK1-Fc, which are known to interact with each other from both functional and
structural studies (83, 181). My experiments confirmed that BMP9 binds to soluble ALK1 with
an extremely high affinity with a KD= 1.8nM, which though weaker than TGF-β1/TβRII
interaction, is still of high affinity.
We then looked into the binding affinities of sEng for TGF-β1 and BMP9. Purified dimeric
recombinant sEng (∆586) was fractionated through size exclusion chromatography and coupled
to the CM5 chip. Various concentrations of TGF-β1 and BMP9 were passed over sEng to look at
the strength of the interactions. TGF-β1 showed no significant binding to the immobilized sEng,
with an estimated KD value of 1230 nM. In none of the trials were we able to fully saturate the
bound sEng, indicating extremely low affinity of interaction. It is evident that TGF-β1 does not
36
bind to recombinant sEng at lower and more physiological concentrations. Interestingly, sEng
was found to have a high affinity towards BMP9. Our experiments showed that sEng/BMP9
interaction has a KD = 5 nM, which is in the same range as the ALK1 – BMP9 interaction.
Together these data demonstrate that, despite multiple reports of recombinant sEng being able to
block TGF-β1 signalling, it cannot do so by scavenging TGF-β1 directly, but rather that it has a
strong affinity towards BMP9 and could scavenge that ligand.
37
(A)
(B)
(C)
(D)
38
Figure 9. sEng binds with high affinity to BMP9, but not to TGF-β1.
Commercially available recombinant TβRII (A) or ALK1-Fc (B) were amine-coupled to a CM5
chip and increasing concentrations of TGF-β1 or BMP9 were passed over their respected
receptors to determine the on and off constants and calculate the affinity constants. sEng (∆586)
was coupled to a different CM5 chip and increasing concentrations of TGF-β1 (C) or BMP9 (D)
were injected over the captured protein. Colored lines represent the actual data acquired in the
course of the experiment and dotted lines show the expected graph assuming 1:1 binding mode
using the calculated KD values.
These experiments were done with Allison Lindsay Gregory from our laboratory and Zhijie Li in
Dr. James Rini’ s laboratory, Department of Biochemistry, University of Toronto.
39
3.3 sALK1-Fc and sEng block BMP9 – but not TGF-β1 –induced ET-1 secretion
During preeclampsia, the placenta produces a number of anti-angiogenic factors, including
circulating sEng and sFlt-1, which show a strong correlation with PE severity (182, 183). Our
own experiments (Figure 8) and recent papers concluded that BMP9 is the most relevant ligand
for recombinant sEng. Knowing that BMP9 is a strong inducer of ET-1 secretion, we
hypothesized that one of the functions of sEng in the context of PE, could be to scavenge BMP9
and interfere with its ability to block ET-1 secretion.
To examine if sEng is capable of blocking ET-1 secretion induced by BMP9 or TGF-β1 ligands,
we treated cells with either BMP9 or TGF-β1 in the presence or absence of 25 nM ALK1-Fc or
sEng. Treatment with 25 nM ALK1-Fc alone (in the absence of ligand) induced a small but
significant reduction in ET-1 secretion from 16 to 8 ng/ml, implying that endothelial cells
themselves secrete a small amount of BMP9, which can be blocked with addition of its soluble
receptor. The addition of sEng (∆586) alone decreased ET-1 secretion, but to a lower extent than
sALK1-Fc did (Figure 10A).
sALK1-Fc was used as positive control, since this protein is known to have high affinity for
BMP9 but not TGF-β1. EC treated with 1 ng/ml of BMP9 showed a significant (2.5-fold)
increase in ET-1 secretion. As expected, addition of 25 nM sALK1-Fc abolished BMP9 –
induced ET-1 secretion. EC treated with a combination of BMP9 and sEng (∆586), showed
significant inhibition in ET-1 secretion compared to BMP9 treated cells (Figure 9B).
Treatment with 1 ng/ml of TGF-β1 produced a significant (2.5-fold) increase in ET-1 secretion.
However, neither sALK1-Fc nor sEng could block TGF-β1–induced ET-1 secretion (Figure 9C).
Taken together these data indicate that sALK1-Fc and sEng (∆586), can interfere with ET-1
release through direct binding to BMP9.
40
Control BMP9 BMP9 + sALK1 BMP9 + sEng0
20
40
60
80
100
120
#
#
*
% of BMP9-induced ET-1
0
20
40
60
80
100
120
Control TGF-ββββ1+ sEngTGF-ββββ1 + sALK1
% of TGF- ββ ββ1-induced ET-1
TGF-ββββ1
*
Control sALK1 sEng
0
4
8
12
16
20
ET-1 [pg/m
l]
*
(A)
(B)
(C)
41
Figure 10. sALK1-Fc and sEng block BMP9 – but not TGF-β1 –induced ET-1 secretion.
EC were starved for 12 hours in serum-free media and incubated with 25 nM of sALK1-Fc or
sEng (∆586), alone (A) or in combination with 1 ng/ml of BMP9 (B) or TGF-β1 (C) for 10
hours. Media were collected in duplicate and analyzed by ET-1 ELISA. Each column represents
data mean ± SEM of 5 experiments. *and # denote significance versus control and stimulated
levels respectively, P< 0.05.
42
3.4 VEGF interferes with BMP9 but not TGF-β1 signalling to block ET-1 secretion
Several studies have implicated altered PlGF/VEGF signalling pathways in PE pathogenesis (4,
182). sFlt-1, shown to be elevated in the serum of preeclamptic women, directly binds to both
PlGF and VEGF, blocking their downstream signalling. Several studies have reported lower
levels of free PlGF and VEGF proteins in the serum of women with PE (20, 183, 184). Lack of
VEGF signalling is thought to be directly responsible for the observed hypertension and
proteinuria. In fact, external administration of sFlt-1 has been shown to lead to PE-like
symptoms (4). Since both PlGF and VEGF have been implicated in PE, we examined if they
could affect TGF-β superfamily receptor signalling. Specifically we looked at the ability of either
PlGF or VEGF to alter ET-1 release, which is known to be dysregulated in PE (178).
To test PlGF/VEGF involvement in ET-1 secretion, EC were incubated with 1 ng/ml of BMP9 or
TGF-β1 in the presence or absence of PlGF or VEGF (Figure 11). First, we looked at the effect
of VEGF and PlGF on basal ET-1 secretion. Neither molecule significantly affected the basal
level of ET-1 secreted in the media (Figure 10A). Cells treated with BMP9 showed a robust 2-
fold increase in ET-1 secretion. Surprisingly, when cells were treated with a combination of
BMP9 and VEGF, we saw a significant decrease in ET-1 secretion, compared to cells treated
with BMP9 alone. Treatment with PlGF did not have any effect on ET-1 secretion (Figure 10B).
Similar to BMP9, TGF-β1–induced a strong increase in ET-1 secretion. Cells co-incubated with
both TGF-β1 and VEGF did not show any difference in ET-1 secretion compared to cells treated
with TGF-β1 alone. Similarly, PlGF was not able to influence TGF-β1 stimulated ET-1 secretion
(Figure 10C). In summary, PlGF did not have any effect on ET-1 secretion, induced by either
BMP9 or TGF-β1. VEGF, on the other hand, strongly inhibited ET-1 secretion induced by
BMP9, however failed to do so with cells treated with TGF-β1.
43
0
20
40
60
80
100
120
140
*
TGF-ββββ1 + PlGFTGF-ββββ1 + VEGF
% of TGF- ββ ββ1-induced ET-1
Control TGF-ββββ1
Control BMP9 BMP9 + VEGF BMP9 +PlGF0
20
40
60
80
100
120
#
*
% of BMP9-induced ET-1
Control VEGF PlGF
0
4
8
12
16
20
ET-1 [pg/m
l]
(C)
(B)
(A)
44
Figure 11. VEGF inhibits BMP9 – but not TGF-β1 –induced ET-1 secretion.
EC were seeded in 6-well plates and incubated at 37oC for 48 hours. Cells were then starved for
12 hours in serum-free media and incubated with 50 ng/ml of either VEGF or PlGF (A) in the
presence or absence of 1 ng/ml of either BMP9 (B) or TGF-β1 (C). After 10 hours, media were
collected for determination of ET-1 by ELISA. Each sample was run in duplicate. Data are
shown as mean ± SEM of 5 experiments. * and # denote significance (P< 0.05) compared to
control and stimulated levels respectively.
45
3.5 Additive effects of BMP9 and TGF-β1 on ET-1 secretion
The fact that VEGF can lower ET-1 levels induced by BMP9 but not by TGF-β1 suggests that
there could be differences in signalling cascade between these two proteins. If two ligands use
different pathways to induce ET-1 secretion, we should see an additive effect when both ligands
are present. In fact, a recent study reported that co-incubation of endothelial cells with both TGF-
β1 and BMP9 proteins lead to an additional 29% increase in ET-1 secretion (140). To see if
mouse embryonic endothelial cells respond similarly, I incubated cells with 1 ng/ml of both
TGF-β1 and BMP9 (Figure 12).
Unstimulated cells produced 15.5 pg/ml of ET-1. Treatment with either BMP9 or TGF-β1
produced a similar 3-fold increase in ET-1 secretion. However, their combination showed a 5-
fold increase in ET-1 levels, compared to cells treated with either protein individually.
These results suggest that TGF-β1 and BMP9 use different pathways to induce ET-1 secretion.
Cells treated with 1 ng/ml of TGF-β1 were fully saturated and an additional increase in TGF-β1
concentration did not result in an increase of ET-1 secretion (Figure 1B). However, treatment
with TGF-β1 and BMP9 clearly induced a stronger response than each protein individually and
gave an additive effect, suggestive of distinct pathways of ET-1 secretion.
46
0
20
40
60
80
&#
*
BMP9 + TGF-ββββ1 TGF-ββββ1BMP9
ET-1 (pg/m
l)
Control
*
Figure 12. Combination of BMP9 and TGF-β1 has an additive effect on ET-1 secretion.
EC were starved for 12 hours in serum-free media and incubated with BMP9, TGF-β1 or a
combination of both ligands for 10 hours. Media were collected and analyzed for ET-1 by
ELISA. *P<0.05 compared to control levels, # and & P<0.05 compared to BMP9 and TGF-β1
treated cells respectively. Bars represent mean ± SEM of 5 experiments.
47
3.6 Analysis of the ability of inhibitors to block their respective phosphorylation pathways
BMP9 and TGF-β1 are known ligands for the TGF-β receptor superfamily. BMP9 was recently
shown to bind with high affinity to ALK1 to induce strong Smad1 phosphorylation, while TGF-
β1 is a strong inducer of Smad2/3 phosphorylation (79, 83, 185). Furthermore, in endothelial
cells, TGF-β1 was also shown to signal through ALK1 receptor to initiate Smad1/5/8
phosphorylation (84). Besides the canonical Smad pathways, both ligands are known to activate
other signalling cascades including ERK and p38 (186, 187). Recent studies have implicated
Smad1 and p38 in BMP9-induced ET-1 release from human endothelial cells (139, 140). Given
that VEGF could interfere with BMP9– but not TGF-β1–induced ET-1 secretion, we
hypothesized that these proteins use distinct pathways to stimulate ET-1. In order to examine the
pathways responsible for ET-1 release, we treated cells with inhibitors of Smad1/5/8
(dorsomorphin), Smad2/3 (SB525334), and p38 (BIRB796) phosphorylation. Before using
inhibitors in ELISA experiments, we looked at their specificity in blocking their respective
pathways.
Cells were serum starved for 3 hours and incubated for 30 minutes with either TGF-β1 or BMP9
in the presence or absence of various inhibitors. Preliminary experiments were performed to
determine the optimal concentration of the inhibitors. Dorsomorphin was able to significantly
reduced Smad1/5/8 phosphorylation at 40 µM; further increase in concentration of the inhibitor
did not affect the phosphorylation. SB525334 completely blocked phosphorylation of Smad2/3 at
8 µM; therefore this concentration was chosen for our experiments. BIRB796 at 1 µM resulted in
a mild decrease of p38 phosphorylation. Increase in BIRB796 concentration did not show further
down-regulation of p38 phosphorylation (data not shown).
TGF-β1 was the only molecule capable of inducing Smad2/3 phosphorylation. BMP9 showed
only a small increase in Smad2/3 phosphorylation, which did not reach significance. Since
Smad2/3 phosphorylation is mediated by ALK5 receptor, neither BIRB796 nor dorsomorphin
could interfere with ALK5 signalling. On the other hand, SB525334, an ALK5 inhibitor
completely abolished Smad2/3 phosphorylation induced by TGF-β1 treatment (Figure 13A).
Both BMP9 and to a lesser extent TGF-β1 were able to induce Smad1/5/8 phosphorylation.
TGF-β1–induced phosphorylation was blocked by both dorsomorphin and SB525334, implying
48
that both receptors are required for TGF-β1 activation of the Smad1/5/8 pathway. These results
are in agreement with the reports showing that TGF-β1 binds to ALK5 to promote association
with ALK1 receptor and can activate both Smad1/5/8 and Smad2/3 pathway simultaneously
(78). As expected, BIRB796 could not affect Smad phosphorylation induced by TGF-β1 and
BMP9.
BMP9 induces Smad1/5/8 phosphorylation by directly binding to and activating the ALK1
receptor (79, 83). Therefore, the ALK1 inhibitor, dorsomorphin, was the only molecule tested
capable of inhibiting Smad1/5/8 phosphorylation induced by BMP9. Neither SB525334 nor
BIRB796 produced any noticeable effect (Figure 13B).
Treatment with BMP9 and TGF-β1 resulted in a slight increase in p38 phosphorylation, which,
however, did not reach significance. Phosphorylation of p38 is activated by a number of
signalling pathways and by stress signals; it is therefore possible that p38 phosphorylation was
already induced in the control state. Among the 3 inhibitors, only dorsomorphin and BIRB796
showed a trend towards lowering control levels of p38 phosphorylation. These results are in
agreement with several papers reporting the ability of dorsomorphin to block p38
phosphorylation (188, 189). The ALK5 inhibitor, SB525334, was unable to affect p38
phosphorylation levels (Figure 13C).
In summary, these results indicate that SB525334 is a very potent inhibitor of TGF-β1 effects. It
blocks both Smad1/5/8 and Smad2/3 phosphorylation induced by TGF-β1. Dorsomorphin
blocked Smad1/5/8 phosphorylation initiated by BMP9 and TGF-β1, but also blocked p38
phosphorylation as an off-target effect. BIRB796 is a specific inhibitor of p38 and did not
influence phosphorylation of Smad1/5/8 or Smad2/3 molecules.
49
(B)
(C)
(A)
Control Stimulation Dorsomorphin BIRB796 SB525334
0
4
8
12
16
20
Phospho-Smad2/3 / Total Smad2/3
BMP9
TGF-ββββ1
*
#
Control Stimulation Dorsomorphin BIRB796 SB525334
0
5
10
15
20
25
30
35
#
#
*
Phospho-Smad1/5/8 / Total-Smad1/5/8 *
P=0.11
Control Stimulation Dorsomorphin BIRB796 SB525334
0.0
0.4
0.8
1.2
1.6
P=0.07
Phospho-p38 / ββ ββ-actin
P=0.07
P=0.15
50
Figure 13. Effects of inhibitors on Smad1/5/8, Smad2/3 and p38 phosphorylation.
Cells were seeded on 6-well plates and cultured for 48 hours at 37oC. After 3 hour starvation in
serum-free media, cells were incubated for 30 minutes with the following inhibitors:
dorsomorphin (D; 40 μM), SB525334 (S; 8 μM) or BIRB796 (B;1 μM). Afterwards, cells were
treated with 1 ng/ml of either BMP9 or TGF-β1. Cell lysates were resolved by SDS-PAGE and
analyzed by Western blot with antibodies to phosphorylated Smad1/5/8 (PSmad1/5/8), Smad2/3
(PSmad2/3) and p38 (P-p38) as well as total Smad1/5/8 (Smad1/5/8), Smad2/3 (Smad2/3) and β-
actin. Densitometry was performed to quantify each band. The extent of Smad phosphorylation
was quantified relative to total Smad levels, while p38 phosphorylation was quantified relative to
β-actin levels. The graphs represent the mean ± SEM of 3 experiments, done in duplicate.
51
3.7 SMAD1 and p38 phosphorylation are involved in BMP9 – and TGF-β1 –induced ET-1 secretion
Recent studies have looked into pathways involved in ET-1 release. Specifically various
inhibitors were used to show that p38 and Smad1 pathways are extremely important for BMP9-
induced ET-1 secretion (139, 140). Both studies also concluded that ERK phosphorylation was
not involved in ET-1 secretion. However, TGF-β family receptor signalling can be extremely cell
dependent; BMP9, for example, was able to promote cell growth of multiple endothelial cell
lines (133); however BMP9 inhibited proliferation and invasiveness of breast cancer cells (190).
It is, therefore highly important to examine signalling pathways involved in ET-1 release in
mouse embryonic endothelial cells.
EC cells were serum-starved for 12 hours and incubated with inhibitors of Smad1, Smad2 and
p38 pathways for 30 minutes. Cells were then treated with 1ng/ml of either BMP9 or TGF-β1 for
10 hours. Media were collected for ET-1 ELISA measurements (Figure 13). Treatment of EC
with dorsomorphin but not SB525334 inhibitor resulted in a decrease of ET-1 secretion,
implying that cells produce small amounts of BMP9, which contribute to the background level of
ET-1. These results are in agreement with our previous experiments, showing that sALK1-Fc is
capable of lowering background levels of ET-1, presumably by binding and neutralizing free
BMP9 molecules in the media. Treatment with BIRB796 did not affect baseline ET-1 secretion
(Figure 13A).
Dorsomorphin is an inhibitor of ALK1 (the receptor for BMP9) and was extremely efficient at
blocking BMP9-induced ET-1 release, completely abolishing it. SB525334, the inhibitor of
ALK5, a receptor not used by BMP9, showed no effect on BMP9-induced ET-1 secretion.
Treatment with BIRB796 showed a significant reduction in ET-1 levels, though this inhibitor
was half as efficient as dorsomorphin, implying both p38 dependent and independent pathways
(Figure 13B).
A similar picture was observed with cells treated with TGF-β1. Dorsomorphin, which blocks
both Smad1/5/8 and p38 pathways, completely inhibited ET-1 secretion back to baseline levels,
implying that the majority of ET-1 is produced via these two pathways. ALK5 catalytic activity
is needed for ALK1 activation as well as for p38 phosphorylation. Not surprisingly, SB525334,
an ALK5 inhibitor, completely abolished ET-1 secretion in TGF-β1-treated cells. Incubation
52
with p38 inhibitor showed a significant decrease in ET-1 secretion, suggesting that indeed the
p38 pathway is involved in TGF-β1 stimulated ET-1 release (Figure 13C).
Overall, both BMP9 and TGF-β1 use similar pathways to induce ET-1 production, specifically
phosphorylation of Smad1 and p38. Most of the ET-1 is getting produced through activation of
the Smad1/5/8 signalling cascade, while about half of ET-1 may be produced through the p38
pathway.
53
Control Dorsomorphin SB525334 BIRB796
0
4
8
12
16
ET-1 [pg/m
l]
Control BMP9 BMP9 + Dorsomorphin BMP9 + SB525334 BMP9 + BIRB 796
0
20
40
60
80
100
120
#
#
*
% of BMP9 -induced ET-1
0
20
40
60
80
100
120
TGF-ββββ1 + BIRB796TGF-ββββ1 + SB525334TGF-ββββ1 + DorsomorphinTGF-ββββ1
#
#
#
*
% of TGF- β
1 −
β1
−β
1 −
β1
−induced ET-1
Control
(B)
(C)
(A)
54
Figure 14. Smad1 and p38 pathways contribute to both BMP9 – and TGF-β1 –induced
ET-1 secretion.
Endothelial cells were serum starved for 12 hours and then incubated with various inhibitors for
30 minutes (A). Subsequently, either 1 ng/ml of BMP9 (B) or TGF-β1 (C) was added to the
cells. After a 10 hour incubation, supernatants were collected and analyzed by ET-1 ELISA.
*P<0.05 with respect to controls; # P<0.05 versus ligand stimulation. Bars represent mean ±
SEM for at least 3 experiments done in duplicate.
55
3.8 VEGF may lower BMP9 –induced p38β phosphorylation
Even though the literature indicates that both BMP9 and TGF-β1 should induce a significant
increase in p38 phosphorylation (186), mouse embryonic endothelial cells showed only a small
increase in phospho-p38 levels. Yet, treatment with the p38 inhibitor resulted in a significant
decrease in ET-1 levels, implying that p38 is involved in ET-1 secretion. To explore this further
we examined two isotypes of p38, which are present in endothelial cells: p38 alpha (p38α) and
p38 beta (p38β). In this experiment we looked into the phosphorylation status of these two p38
isotypes in endothelial cells at two different time points.
Cells were serum-starved for 3 hours and then treated with either BMP9 or TGF-β1 in the
presence of absence of VEGF. We looked at the phosphorylation state of p38α and p38β after 1
and 4 hours of treatment (Figure 14). Similar to what we saw previously (Figure 12), levels of
phospho-p38 regardless of isotypes remained unchanged after 1 hour of incubation with either
BMP9 or TGF-β1. After 4 hours, however, we saw an increase in p38β phosphorylation levels in
cells treated with either ligand. Surprisingly, VEGF treatment resulted in somewhat lower levels
of phospho-p38β. Additionally, cells that were co-incubated with either BMP9 or TGF-β1 along
with VEGF, showed a small decrease of p38β phosphorylation compared to the cells treated with
just BMP9 or TGF-β1 alone. The phosphorylation state of p38α, on the other hand, remained
unchanged throughout the experiment.
In summary, it appears that p38β is being phosphorylated in response to BMP9 or TGF-β1
treatments. However when cells are incubated with BMP9 or TGF-β1 in the presence of VEGF,
phospho-p38β levels were reduced, compared to untreated cells, implying that VEGF may block
BMP9– and TGF-β1–induced p38β phosphorylation. The increase in phosphorylation only after
4 hours of incubation suggests that protein synthesis is needed to initiate phosphorylation of
p38β. It is likely to be an indirect way of p38 activation, potentially through genes activated by
the canonical Smad pathway.
0.0
0.4
0.8
1.2
1.6
BMP9Control
Phopho-p38 / ββ ββ-actin
Figure 15. VEGF lowers BMP9
Cells were serum-starved for 3 hours and then treated with 1 ng/ml of BMP9 or TGF
presence of absence of 50 ng/ml of VEGF. Cells were then incubated for 1 or 4 hours, lysed and
resolved on SDS-PAGE gel and analyzed by Western Blot using antibodie
and β, and for β-actin. The graph shows Phospho
β-actin after 4 hours of incubation. Bars represent mean ±
VEGFTGF-ββββ1BMP9 BMP9 + VEGF
p38 α
p38 β
TGF-ββββ1 + VEGF
VEGF lowers BMP9 –induced p38β phosphorylation.
starved for 3 hours and then treated with 1 ng/ml of BMP9 or TGF
presence of absence of 50 ng/ml of VEGF. Cells were then incubated for 1 or 4 hours, lysed and
PAGE gel and analyzed by Western Blot using antibodies for Phospho
actin. The graph shows Phospho-p38α and Phospho-p38β expression relative to
actin after 4 hours of incubation. Bars represent mean ± SEM for 4 experiments.
56
starved for 3 hours and then treated with 1 ng/ml of BMP9 or TGF-β1 in the
presence of absence of 50 ng/ml of VEGF. Cells were then incubated for 1 or 4 hours, lysed and
s for Phospho-p38α
p38β expression relative to
for 4 experiments.
57
4 Discussion
Our studies confirmed previous findings, indicating that sEng (∆586) binds BMP9 and not TGF-
β1 with high affinity. Additionally, I found that BMP9 and TGF-β1 can induce ET-1 secretion
from mouse endothelial cells and hypothesized that the function of sEng in preeclampsia could
be to block BMP-9 signalling and subsequent ET-1 secretion. Therefore, I examined the ability
of sEng to block ET-1 secretion from BMP9 treated ECs, and conclude that sEng can indeed
block BMP9 signalling. Furthermore, I discovered that VEGF, but not PlGF, could block BMP9-
induced ET-1 secretion, suggesting a new way for VEGF to inhibit vasoconstriction. Finally, I
used inhibitors of ALK1, ALK5 and p38 pathways to dissect the molecular mechanism behind
BMP9 and TGF-β1 induction of ET-1 secretion. My findings indicate that ALK1 and to a lesser
extent p38 pathways are important for ET-1 secretion. Though my data did not show a
significant down-regulation by VEGF of p38β phosphorylation induced by BMP9, there was a
trend. It is, therefore, possible that VEGF blocks BMP-9 –induced ET-1 secretion by down-
regulating p38β pathway; however, more experiments are needed to support this claim.
Additionally, the fact that VEGF cannot block TGF-β1 –induced ET-1 secretion suggests that
other signalling pathways are involved in this process.
4.1 BMP9 and TGF-β1 induce ET-1 secretion from mouse embryonic endothelial cells
Recently two papers were published reporting the ability of BMP9 and TGF-β1 to induce ET-1
secretion from human umbilical or bovine aortic endothelial cells (139, 140). Both studies raised
the possibility that this mechanism might be important for hypertensive disorders, including
preeclampsia(139, 140). We performed similar experiments with mouse embryonic ECs and
confirmed that incubation with either BMP9– or TGF-β1–induced release of endothelin-1. Both
papers reported a two-fold increase in ET-1 production in treated cells, which is quite
comparable with our own findings (Figure 1). Thus, our studies extend the observations to a
distinct type of endothelial cells and in different species suggesting that stimulation of ET-1
secretion by BMP9 and TGF-β1 is likely a general mechanism with relevance to systemic
hypertension and a condition such as preeclampsia.
58
4.2 Potential role of sEng in PE
The role of sFlt-1 in preeclampsia is still not fully understood; however, it is clear that it exerts
its effects by interfering with VEGF and PlGF signalling. The functional role of sEng, however,
has yet to be defined. Several studies looked at the effects of recombinant sEng (∆586) and
concluded that it contributes to endothelial dysfunction (64) and a slight increase in systemic
blood pressure (4). It was postulated that sEng is able to scavenge TGF-β1 and interfere with
ALK5 signalling. However recent studies (65, 181) and our own experiments (Figure 2) clearly
showed that the recombinant sEng (∆586) only binds with high affinity to BMP9 and not to
TGF-β1. Both studies (65, 181) determined that recombinant sEng binds TGF-β1 in the µM
range or weaker, which is consistent with our estimated KD value of 1.2 µM. It is clear that with
that kind of binding affinity, recombinant sEng would not be able to scavenge circulating TGF-
β1 under physiological concentrations. However, binding of recombinant sEng to BMP9 was
strong; our experiments showed a KD = 4.8nM for sEng/BMP9 interaction. Similar studies were
performed by other groups, including Castonguay et al. who determined that recombinant sEng-
Fc binds to BMP9 with an affinity of 40 pM (181). This is somewhat surprising, considering that
the same group published a previous paper, claiming that the KD of sALK1-Fc for BMP9 was 2
nM (191). It is unlikely that sEng can bind BMP9 with an affinity 100 times greater than for its
natural receptor ALK1, especially considering that the amount of endoglin on the cell surface is
at least 100 times more than the amount of ALK1. This would cause BMP9 to preferentially bind
to endoglin, ignoring ALK1 receptors. It should be mentioned that the group used proteins
expressed as chimeras with the immunoglobulin Fc domain. It is possible that the addition of the
Fc domain could alter the protein structure, making a more stable protein with higher affinity for
BMP9. Recently, a new study was published showing that KD value of sEng/BMP9 interaction is
around 4 nM, which is similiar to the value obtained in our experiments (65). In summary, I
conclude that BMP9 and not TGF-β1 is a binding partner of recombinant sEng (∆586), and we
suggest that the true physiological sEng in circulation would have a similar relative affinity for
these ligands. However, this should be demonstrated experimentally.
Next I examined the function of circulating BMP9. Research on BMP9 effects on endothelial
cells reported somewhat controversial results as BMP9 was found to be both a quiescence (132),
and an angiogenic factor (133) for endothelial cells. Moreover, recent studies reported that
BMP9 can induce ET-1 production. Since ET-1 has been implicated in PE, I hypothesized that
59
the role of sEng in preeclampsia could be to sequester BMP9 molecules and inhibit ET-1
secretion. I first looked at the ability of soluble ALK1-Fc to inhibit BMP9-induced ET-1
secretion. As expected, ALK1-Fc was able to completely abolish ET-1 stimulation. From my
BIAcore studies, I knew that sEng was not as effective at binding BMP9 as ALK1; therefore, it
was not surprising that sEng was only half as effective at blocking ET-1 secretion. Neither
recombinant sALK1-Fc nor sEng (∆586) could block TGF-β1–induced ET-1 secretion, which
correlates well with our BIAcore studies.
Interestingly, cells incubated with sALK1-Fc alone showed a significant decrease in basal levels
of ET-1 secretion over a period of 10 hours, indicating that unstimulated endothelial cells can
produce a fair amount of BMP9. This hypothesis was further strengthened by the fact that cells
treated with an inhibitor of ALK1 signalling, dorsomorphin, also showed a decrease of ET-1
release compared to unstimulated cells. It is likely that BMP9 molecules, present in the cultures,
are responsible for the background accumulation of ET-1. Levels of BMP9 have yet to be
measured in patients with preeclampsia.
Multiple studies have reported increased levels of ET-1 in preeclampsia; in fact, blockade of ET-
1 receptors was able to completely abolish PE-associated hypertension in mouse and rat models
(171). It is therefore possible that one of the functions of sEng is to scavenge BMP9 molecules to
inhibit ET-1 secretion and subsequent hypertension. Release of sEng could represent a protective
mechanism, initiated in attempt to keep hypertension under control (Figure 16). During
preeclampsia an increase in circulating sFlt-1 molecules would diminish VEGF signalling,
leading to an increase in ET-1 secretion, since BMP9 signalling would remain unopposed. ET-1
would in turn, lead to an increase in blood pressure. We hypothesize that release of sEng might
be a response to VEGF blockade aimed at reducing hypertension by inhibiting ET-1 secretion.
Interestingly, elevated sEng is usually detected 1-2 weeks in advance of a detectable increase in
sFlt-1. Therefore, it is unlikely that sEng is being released in response to sFlt-1. However, AT1-
AA have been shown to be present in PE patients at much earlier time points and are thought to
be at least partially responsible for the secretion of sFlt-1 and sEng. Therefore, it is possible that
sEng is released in response to AT1-AA, which would eventually initiate sFlt-1-induced
hypertension.
60
Figure 16. Model of Preeclampsia.
In normal pregnancy, BMP9 signalling leading to ET-1 secretion is counteracted by VEGF
signalling, which prevents excessive ET-1 secretion and keeps blood pressure at the normal
level. However, during preeclampsia, high levels of sFlt-1 block VEGF signalling. ALK1
signalling becomes unopposed leading to increase in ET-1 secretion, which in turn results in
hypertension. We hypothesize that sEng gets released in response to hypertension and represents
a physiological attempt to control blood pressure. sEng would then bind BMP9 and block ALK1
signalling, therefore down-regulating ET-1 secretion and subsequent hypertension.
61
4.3 VEGF signalling and hypertension
By far the most interesting finding of my work is the fact that VEGF can block BMP-9-induced
ET-1 secretion (Figure 4B). As far as I know, we are the first group to report a negative effect of
VEGF on ET-1 secretion in endothelial cells. The role of VEGF in preeclampsia has never been
disputed as multiple groups agree that the lack of VEGF signalling due to the increased
production of sFlt-1 by the placenta can mimic the majority of preeclampsia symptoms,
including hypertension and proteinuria (192). Hypertension, induced by the injection of sFlt-1
was usually attributed to the lack of eNOS activation (193). NO is one of the most common
vasodilators in physiology, however quantifying NO production is not an easy task. NO levels in
preeclampsia were examined by several groups; somewhat controversial findings were reported.
Various studies found NO levels to be either reduced (194), normal (195) or increased (196) in
preeclampsia. In theory, reduced NO could explain PE-associated hypertension, while increased
NO could represent a physiological response to already induced hypertension.
In this study, I propose a new mechanism by which sFlt-1 can induce hypertension. I report that
ET-1 secretion induced by BMP9 can be efficiently blocked by VEGF. It is interesting to note
that BMP9 concentration in the serum is 5 to 10 ng/ml, meaning that BMP9 is constantly
activating ALK1 signalling in endothelial cells. This was clearly demonstrated by Bidrart et al;
they showed phosphorylated Smad1/5/8 molecules in the nucleus of aortic endothelial cells,
attributed to BMP9 in circulation (129). Therefore, there should be a constant secretion of ET-1
molecules into the serum. At the same time, other studies report that the concentration of ET-1 in
circulation is quite low (197), suggesting additional factors that keep ET-1 production in check.
We propose that in healthy individuals VEGF signalling opposes BMP-9-induced secretion of
ET-1, thus keeping a low concentration in circulation and preventing hypertension. Upon release
of sFlt-1, however, VEGF signalling is blocked. BMP9-induced secretion of ET-1 is no longer
inhibited, leading to an increase in ET-1 circulating levels and subsequent systemic hypertension
(Figure 14).
Interestingly, PlGF was not able to replicate the effect of VEGF on ET-1 secretion. One possible
explanation is the difference in signalling. PlGF signals through VEGFR-1 (Flt-1), while VEGF
signals through both VEGFR-1 and VEGFR-2 (198). A number of studies reported that certain
functions can be mediated strictly by one receptor. For example, adipose tissue expansion was
62
induced by VEGF exclusively through VEGFR-2, as an antibody to VEGFR-1 was unable to
block the expansion (199). Moreover, Koolwijk et al. examined the VEGF effect on tube
formation by human microvascular endothelial cells and found that VEGFR-2 is the only
receptor involved in this process (200). Therefore, our data support that VEGF mediates its
blocking effect on ET-1 secretion via VEGFR-2.
4.4 Pathways involved in ET-1 production
VEGF was able to inhibit BMP9-induced ET-1 secretion, however failed to block ET-1 secretion
induced by TGF-β1. One possible explanation might be that BMP9 and TGF-β1 use different
signalling pathways to induce ET-1 release. I therefore incubated cells with both ligands together
to see if they have an additive effect. Indeed, cells co-incubated with BMP9 and TGF-β1
secreted significantly more ET-1 than in response to either ligand alone. It should be noted that
we used a concentration of 1 ng/ml for both ligands. Our titration experiments (Figure 1) showed
that cells treated with TGF-β1 become saturated at a concentration of 0.8 ng/ml and addition of
extra TGF-β1 molecules did not result in an increase in ET-1 secretion. Therefore, it is likely that
additional ET-1 released by cells co-incubated with the two ligands was initiated by a separate
pathway, which is not activated by TGF-β1 signalling alone. To examine the differences between
TGF-β1 and BMP9 signalling leading to ET-1 secretion we tested inhibitors of ALK5 activation
and of Smad1/5/8 and p38 pathways.
Our experiments showed that BMP9 signals exclusively through ALK1 receptor to induce ET-1
secretion, which is not surprising as ALK1 activity is needed to activate all BMP9-related
signalling. These results are in agreement with previous studies showing that BMP9-induced ET-
1 release is dependent on Smad1 and p38 phosphorylation (139). Similarly, in our studies
inhibition of either pathway led to a significant decrease in ET-1 secretion (Figure 12B).
Interestingly, Park et al, showed that though HUVEC cells need Smad1 protein to initiate ET-1
secretion, the presence of Smad4 protein is not required (139).
I also performed similar experiments to elucidate the pathways responsible for TGF-β1–induced
ET-1 secretion. I observe that activation of Smad1/5/8 and p38 signalling pathways are also
necessary for TGF-β1 to induce ET-1 secretion. In fact, inhibition of ALK1 pathways reduced
ET-1 production to background levels. The p38 inhibitor was not as effective as dorsomorphin,
but was still capable of significantly decreasing ET-1 secretion. It appears that these two
63
pathways together are responsible for most of the ET-1 release, suggesting that Smad2/3
phosphorylation is less important in this process. Still, the ALK5 inhibitor, SB525334,
completely blocked ET-1 secretion, implying that the kinase activity of ALK5 is necessary.
These results are in agreement with our own data (Figure 11) and multiple other studies (82,
(201, 202), showing that activation of ALK5 is required for ALK1 signalling.
Interestingly, Castañares et al. demonstrated that TGF-β induces ET-1 secretion through
ALK5/Smad3 and not through ALK1 pathways. In fact, knockdown of ALK1 and Smad1
proteins by siRNA had absolutely no effect on ET-1 secretion, suggesting that the ALK1
pathway does not play a role in this process (203). My results, on the other hand, clearly
demonstrate that ALK1 pathway is needed for TGF-β-induced ET-1 secretion. This discrepancy
could possibly be explained by difference in cell types used in two studies. Castañares et al. used
bovine aortic endothelial cells (BAEC), while I used mouse embryonic endothelial cells for our
experiments. Additionally, Star et al, clearly showed that the p38 pathway is also involved in
TGF-β-induced ET-1 secretion, which was also found in my studies (140).
Though TGF-β1 and BMP9-induced only a slight increase in p38 phosphorylation, BIRB796
could significantly decrease ET-1 secretion, implying that the p38 pathway is involved in this
process. One explanation for the phenomenon is that I might not have looked at the p38
phosphorylation at the right time point. Usually, p38 gets activated almost instantly by TβRII
associated kinases (97); however several papers raised the possibility that p38 could become
phosphorylated hours after the initial stimulation (204, 205). Recent studies have also shown that
though there are 4 isoforms of p38 molecules, TGF-β1 does not activate all of them (206).
Therefore, it is possible that I did not detect p38 phosphorylation, because only one isoform
became phosphorylated. Moreover, a recent study by Ferrari et al, looked at crosstalk between
VEGF and TGF-β1 signalling, reporting that co-incubation of cells with both ligands could shift
the phosphorylation pattern from p38α to p38β molecules (207). I therefore thought that co-
incubation of endothelial cells with both VEGF and BMP9 may affect p38 phosphorylation in a
similar manner. Since endothelial cells primarily express p38α and p38β proteins, we tested
p38α and p38β phosphorylation following stimulation with either TGF-β1 or BMP9 in the
presence or absence of VEGF.
64
My results showed that neither BMP9 – nor TGF-β –induced p38α phosphorylation, at either 1
or 4 hours after treatment. That was surprising, since it was reported that TGF-β1 specifically up-
regulated p38α phosphorylation in multiple cell types (208). P38α is thought to be activated by
TGF-β to induce apoptosis, while phosphorylation of p38β is usually associated with survival
and is induced by VEGF. However, in my study, both BMP9 – and TGF-β1 –induced
phosphorylation of p38β after 4 hours of treatment. Interestingly, although our findings were not
statistically significant, co-incubation of VEGF with either ligand suggested that VEGF may
interfere with p38β phosphorylation induced by BMP9 and TGF-β1 ligands and could explain
how VEGF is able to block BMP9-induced ET-1 secretion. This mechanism, however, would not
explain why VEGF is unable to block TGF-β1 –induced ET-1 release from endothelial cells. It is
possible that p38 phosphorylation is not as important for TGF-β1 signalling, or alternatively
TGF-β1 uses a different isoform of p38 to induce ET-1 secretion. There is also the possibility
that TGF-β1 uses a completely different pathway to induce ET-1 secretion, though the fact that
p38 inhibitor BIRB76 had some blocking effect suggests that the p38 pathway does play some
role in TGF-β1 signalling. Further studies are needed to elucidate the exact signalling pathway,
involved in ET-1 secretion by both BMP9 and TGF-β1.
4.5 Conclusion
I report that both recombinant sEng (∆586) and sALK1-Fc can block BMP9 –, but not TGF-
β1induced ET-1 secretion. Also, I found that VEGF and not PlGF can block ET-1 secretion. In
terms of molecular pathways implicated, ALK1 and to a lesser extent p38 signalling are
important for both TGF-β1 and BMP9-induced ET-1 secretion. My studies on the effects of
VEGF on Smad phosphorylation showed that VEGF is not able to induce or block either
Smad1/5/8 or Smad2/3 phosphorylation. However, it is known that VEGF can activate the p38
signalling cascade and in certain cases compete with TGF-β1 for p38 activation. Our own
experiments suggest that VEGF may block BMP9 –induced ET-1 release at least in part by
lowering p38β phosphorylation. It is not clear why VEGF was unable to block TGF-β1 –induced
ET-1 secretion. Further studies are needed to dissect the mechanisms accounting for the
blocking by VEGF of BMP9 – but not TGF-β1 –induced signalling.
65
Moreover, in the context of preeclampsia, it appears that sEng and sFlt-1 converge on the same
pathway of ET-1 production, but have opposing effects (Figure 14). sEng can directly bind to
BMP9 molecules and remove them from circulation, blocking ALK1 signalling and ET-1
secretion. sFlt-1 on the other hand, can sequester VEGF molecules, thus interfering with the
ability of VEGF to block BMP9-induced ET-1 secretion. Therefore, induction of sEng can be
viewed as a beneficial process, aiming to counteract systemic hypertension induced by sFlt-1. In
fact if our hypothesis is correct, sEng or even sALK1-Fc, which has a much higher affinity
towards BMP9, could potentially be used as a therapy for PE. Artificial increase in sALK1-Fc
would sequester BMP9 molecules and relieve ET-1-induced hypertension. We hope that future
research will be able to further investigate role of sEng, sFlt-1 and ET-1 in preeclampsia.
66
Bibliography
1. Sibai B, Dekker G, Kupferminc M. 2005. Pre-eclampsia. Lancet 365: 785-99
2. Wang A, Rana S, Karumanchi SA. 2009. Preeclampsia: the role of angiogenic factors in
its pathogenesis. Physiology (Bethesda) 24: 147-58
3. Maynard SE, Min JY, Merchan J, Lim KH, Li J, Mondal S, Libermann TA, Morgan JP,
Sellke FW, Stillman IE, Epstein FH, Sukhatme VP, Karumanchi SA. 2003. Excess
placental soluble fms-like tyrosine kinase 1 (sFlt1) may contribute to endothelial
dysfunction, hypertension, and proteinuria in preeclampsia. J Clin Invest 111: 649-58
4. Venkatesha S, Toporsian M, Lam C, Hanai J, Mammoto T, Kim YM, Bdolah Y, Lim
KH, Yuan HT, Libermann TA, Stillman IE, Roberts D, D'Amore PA, Epstein FH, Sellke
FW, Romero R, Sukhatme VP, Letarte M, Karumanchi SA. 2006. Soluble endoglin
contributes to the pathogenesis of preeclampsia. Nat Med 12: 642-9
5. Jain A. 2012. Endothelin-1: a key pathological factor in pre-eclampsia? Reprod Biomed
Online 25: 443-9
6. Ray JG, Vermeulen MJ, Schull MJ, Redelmeier DA. 2005. Cardiovascular health after
maternal placental syndromes (CHAMPS): population-based retrospective cohort study.
Lancet 366: 1797-803
7. Raymond D, Peterson E. 2011. A critical review of early-onset and late-onset
preeclampsia. Obstet Gynecol Surv 66: 497-506
8. Wikstrom AK, Larsson A, Eriksson UJ, Nash P, Norden-Lindeberg S, Olovsson M. 2007.
Placental growth factor and soluble FMS-like tyrosine kinase-1 in early-onset and late-
onset preeclampsia. Obstet Gynecol 109: 1368-74
9. Sibai BM, Ramadan MK, Usta I, Salama M, Mercer BM, Friedman SA. 1993. Maternal
morbidity and mortality in 442 pregnancies with hemolysis, elevated liver enzymes, and
low platelets (HELLP syndrome). Am J Obstet Gynecol 169: 1000-6
10. Young BC, Levine RJ, Karumanchi SA. 2010. Pathogenesis of preeclampsia. Annu Rev
Pathol 5: 173-92
11. Noraihan MN, Sharda P, Jammal AB. 2005. Report of 50 cases of eclampsia. J Obstet
Gynaecol Res 31: 302-9
12. Sibai BM. 2005. Diagnosis, prevention, and management of eclampsia. Obstet Gynecol
105: 402-10
13. Euser AG, Cipolla MJ. 2009. Magnesium sulfate for the treatment of eclampsia: a brief
review. Stroke 40: 1169-75
67
14. Wagner LK. 2004. Diagnosis and management of preeclampsia. Am Fam Physician 70:
2317-24
15. 2002. ACOG practice bulletin. Diagnosis and management of preeclampsia and
eclampsia. Number 33, January 2002. American College of Obstetricians and
Gynecologists. Int J Gynaecol Obstet 77: 67-75
16. Duley L. 2009. The global impact of pre-eclampsia and eclampsia. Semin Perinatol 33:
130-7
17. Levine RJ, Lam C, Qian C, Yu KF, Maynard SE, Sachs BP, Sibai BM, Epstein FH,
Romero R, Thadhani R, Karumanchi SA. 2006. Soluble endoglin and other circulating
antiangiogenic factors in preeclampsia. N Engl J Med 355: 992-1005
18. Murphy SR, LaMarca BB, Cockrell K, Granger JP. 2010. Role of endothelin in mediating
soluble fms-like tyrosine kinase 1-induced hypertension in pregnant rats. Hypertension
55: 394-8
19. Kim SY, Ryu HM, Yang JH, Kim MY, Han JY, Kim JO, Chung JH, Park SY, Lee MH,
Kim DJ. 2007. Increased sFlt-1 to PlGF ratio in women who subsequently develop
preeclampsia. J Korean Med Sci 22: 873-7
20. Levine RJ, Maynard SE, Qian C, Lim KH, England LJ, Yu KF, Schisterman EF,
Thadhani R, Sachs BP, Epstein FH, Sibai BM, Sukhatme VP, Karumanchi SA. 2004.
Circulating angiogenic factors and the risk of preeclampsia. N Engl J Med 350: 672-83
21. Sibai BM. 2003. Diagnosis and management of gestational hypertension and
preeclampsia. Obstet Gynecol 102: 181-92
22. Altman D, Carroli G, Duley L, Farrell B, Moodley J, Neilson J, Smith D. 2002. Do
women with pre-eclampsia, and their babies, benefit from magnesium sulphate? The
Magpie Trial: a randomised placebo-controlled trial. Lancet 359: 1877-90
23. Nisell H, Lintu H, Lunell NO, Mollerstrom G, Pettersson E. 1995. Blood pressure and
renal function seven years after pregnancy complicated by hypertension. Br J Obstet
Gynaecol 102: 876-81
24. Irgens HU, Reisaeter L, Irgens LM, Lie RT. 2001. Long term mortality of mothers and
fathers after pre-eclampsia: population based cohort study. BMJ 323: 1213-7
25. Khong TY, De Wolf F, Robertson WB, Brosens I. 1986. Inadequate maternal vascular
response to placentation in pregnancies complicated by pre-eclampsia and by small-for-
gestational age infants. Br J Obstet Gynaecol 93: 1049-59
26. Meekins JW, Pijnenborg R, Hanssens M, McFadyen IR, van Asshe A. 1994. A study of
placental bed spiral arteries and trophoblast invasion in normal and severe pre-eclamptic
pregnancies. Br J Obstet Gynaecol 101: 669-74
68
27. Librach CL, Werb Z, Fitzgerald ML, Chiu K, Corwin NM, Esteves RA, Grobelny D,
Galardy R, Damsky CH, Fisher SJ. 1991. 92-kD type IV collagenase mediates invasion
of human cytotrophoblasts. J Cell Biol 113: 437-49
28. Bai SX, Wang YL, Qin L, Xiao ZJ, Herva R, Piao YS. 2005. Dynamic expression of
matrix metalloproteinases (MMP-2, -9 and -14) and the tissue inhibitors of MMPs
(TIMP-1, -2 and -3) at the implantation site during tubal pregnancy. Reproduction 129:
103-13
29. Lim KH, Zhou Y, Janatpour M, McMaster M, Bass K, Chun SH, Fisher SJ. 1997. Human
cytotrophoblast differentiation/invasion is abnormal in pre-eclampsia. Am J Pathol 151:
1809-18
30. Graham CH, Connelly I, MacDougall JR, Kerbel RS, Stetler-Stevenson WG, Lala PK.
1994. Resistance of malignant trophoblast cells to both the anti-proliferative and anti-
invasive effects of transforming growth factor-beta. Exp Cell Res 214: 93-9
31. Caniggia I, Mostachfi H, Winter J, Gassmann M, Lye SJ, Kuliszewski M, Post M. 2000.
Hypoxia-inducible factor-1 mediates the biological effects of oxygen on human
trophoblast differentiation through TGFbeta(3). J Clin Invest 105: 577-87
32. Caniggia I, Grisaru-Gravnosky S, Kuliszewsky M, Post M, Lye SJ. 1999. Inhibition of
TGF-beta 3 restores the invasive capability of extravillous trophoblasts in preeclamptic
pregnancies. J Clin Invest 103: 1641-50
33. Caniggia I, Taylor CV, Ritchie JW, Lye SJ, Letarte M. 1997. Endoglin regulates
trophoblast differentiation along the invasive pathway in human placental villous
explants. Endocrinology 138: 4977-88
34. Mano Y, Kotani T, Shibata K, Matsumura H, Tsuda H, Sumigama S, Yamamoto E,
Iwase A, Senga T, Kikkawa F. 2011. The loss of endoglin promotes the invasion of
extravillous trophoblasts. Endocrinology 152: 4386-94
35. Harris LK, Clancy OH, Myers JE, Baker PN. 2009. Plasma from women with
preeclampsia inhibits trophoblast invasion. Reprod Sci 16: 1082-90
36. Navarro P, Caveda L, Breviario F, Mandoteanu I, Lampugnani MG, Dejana E. 1995.
Catenin-dependent and -independent functions of vascular endothelial cadherin. J Biol
Chem 270: 30965-72
37. Zhou Y, Damsky CH, Fisher SJ. 1997. Preeclampsia is associated with failure of human
cytotrophoblasts to mimic a vascular adhesion phenotype. One cause of defective
endovascular invasion in this syndrome? J Clin Invest 99: 2152-64
38. Zhou Y, Fisher SJ, Janatpour M, Genbacev O, Dejana E, Wheelock M, Damsky CH.
1997. Human cytotrophoblasts adopt a vascular phenotype as they differentiate. A
strategy for successful endovascular invasion? J Clin Invest 99: 2139-51
69
39. Kendall RL, Wang G, Thomas KA. 1996. Identification of a natural soluble form of the
vascular endothelial growth factor receptor, FLT-1, and its heterodimerization with KDR.
Biochem Biophys Res Commun 226: 324-8
40. Koch S, Claesson-Welsh L. 2012. Signal transduction by vascular endothelial growth
factor receptors. Cold Spring Harb Perspect Med 2: a006502
41. Roncone D, Satoskar A, Nadasdy T, Monk JP, Rovin BH. 2007. Proteinuria in a patient
receiving anti-VEGF therapy for metastatic renal cell carcinoma. Nat Clin Pract Nephrol
3: 287-93
42. Luizon MR, Sandrim VC, Palei AC, Lacchini R, Cavalli RC, Duarte G, Tanus-Santos JE.
2012. Epistasis among eNOS, MMP-9 and VEGF maternal genotypes in hypertensive
disorders of pregnancy. Hypertens Res 35: 917-21
43. Feliers D, Chen X, Akis N, Choudhury GG, Madaio M, Kasinath BS. 2005. VEGF
regulation of endothelial nitric oxide synthase in glomerular endothelial cells. Kidney Int
68: 1648-59
44. He H, Venema VJ, Gu X, Venema RC, Marrero MB, Caldwell RB. 1999. Vascular
endothelial growth factor signals endothelial cell production of nitric oxide and
prostacyclin through flk-1/KDR activation of c-Src. J Biol Chem 274: 25130-5
45. Eremina V, Sood M, Haigh J, Nagy A, Lajoie G, Ferrara N, Gerber HP, Kikkawa Y,
Miner JH, Quaggin SE. 2003. Glomerular-specific alterations of VEGF-A expression
lead to distinct congenital and acquired renal diseases. J Clin Invest 111: 707-16
46. Farina A, Sekizawa A, De Sanctis P, Purwosunu Y, Okai T, Cha DH, Kang JH, Vicenzi
C, Tempesta A, Wibowo N, Valvassori L, Rizzo N. 2008. Gene expression in chorionic
villous samples at 11 weeks' gestation from women destined to develop preeclampsia.
Prenat Diagn 28: 956-61
47. Bainbridge SA, Sidle EH, Smith GN. 2005. Direct placental effects of cigarette smoke
protect women from pre-eclampsia: the specific roles of carbon monoxide and
antioxidant systems in the placenta. Med Hypotheses 64: 17-27
48. Zhou CC, Irani RA, Zhang Y, Blackwell SC, Mi T, Wen J, Shelat H, Geng YJ, Ramin
SM, Kellems RE, Xia Y. 2010. Angiotensin receptor agonistic autoantibody-mediated
tumor necrosis factor-alpha induction contributes to increased soluble endoglin
production in preeclampsia. Circulation 121: 436-44
49. Cudmore M, Ahmad S, Al-Ani B, Fujisawa T, Coxall H, Chudasama K, Devey LR,
Wigmore SJ, Abbas A, Hewett PW, Ahmed A. 2007. Negative regulation of soluble Flt-1
and soluble endoglin release by heme oxygenase-1. Circulation 115: 1789-97
50. Cudmore MJ, Ahmad S, Sissaoui S, Ramma W, Ma B, Fujisawa T, Al-Ani B, Wang K,
Cai M, Crispi F, Hewett PW, Gratacos E, Egginton S, Ahmed A. 2012. Loss of Akt
activity increases circulating soluble endoglin release in preeclampsia: identification of
inter-dependency between Akt-1 and heme oxygenase-1. Eur Heart J 33: 1150-8
70
51. Tan C, Cruet-Hennequart S, Troussard A, Fazli L, Costello P, Sutton K, Wheeler J,
Gleave M, Sanghera J, Dedhar S. 2004. Regulation of tumor angiogenesis by integrin-
linked kinase (ILK). Cancer Cell 5: 79-90
52. Okuyama H, Krishnamachary B, Zhou YF, Nagasawa H, Bosch-Marce M, Semenza GL.
2006. Expression of vascular endothelial growth factor receptor 1 in bone marrow-
derived mesenchymal cells is dependent on hypoxia-inducible factor 1. J Biol Chem 281:
15554-63
53. Gerber HP, McMurtrey A, Kowalski J, Yan M, Keyt BA, Dixit V, Ferrara N. 1998.
Vascular endothelial growth factor regulates endothelial cell survival through the
phosphatidylinositol 3'-kinase/Akt signal transduction pathway. Requirement for Flk-
1/KDR activation. J Biol Chem 273: 30336-43
54. Park JK, Jeong JW, Kang MY, Baek JC, Shin JK, Lee SA, Choi WS, Lee JH, Paik WY.
2010. Inhibition of the PI3K-Akt pathway suppresses sFlt1 expression in human placental
hypoxia models in vitro. Placenta 31: 621-9
55. Irani RA, Xia Y. 2008. The functional role of the renin-angiotensin system in pregnancy
and preeclampsia. Placenta 29: 763-71
56. Gant NF, Worley RJ, Everett RB, MacDonald PC. 1980. Control of vascular
responsiveness during human pregnancy. Kidney Int 18: 253-8
57. Wallukat G, Neichel D, Nissen E, Homuth V, Luft FC. 2003. Agonistic autoantibodies
directed against the angiotensin II AT1 receptor in patients with preeclampsia. Can J
Physiol Pharmacol 81: 79-83
58. LaMarca B, Parrish M, Ray LF, Murphy SR, Roberts L, Glover P, Wallukat G, Wenzel
K, Cockrell K, Martin JN, Jr., Ryan MJ, Dechend R. 2009. Hypertension in response to
autoantibodies to the angiotensin II type I receptor (AT1-AA) in pregnant rats: role of
endothelin-1. Hypertension 54: 905-9
59. Parrish MR, Murphy SR, Rutland S, Wallace K, Wenzel K, Wallukat G, Keiser S, Ray
LF, Dechend R, Martin JN, Granger JP, LaMarca B. 2010. The effect of immune factors,
tumor necrosis factor-alpha, and agonistic autoantibodies to the angiotensin II type I
receptor on soluble fms-like tyrosine-1 and soluble endoglin production in response to
hypertension during pregnancy. Am J Hypertens 23: 911-6
60. Zhou CC, Zhang Y, Irani RA, Zhang H, Mi T, Popek EJ, Hicks MJ, Ramin SM, Kellems
RE, Xia Y. 2008. Angiotensin receptor agonistic autoantibodies induce pre-eclampsia in
pregnant mice. Nat Med 14: 855-62
61. LaMarca BD, Gilbert J, Granger JP. 2008. Recent progress toward the understanding of
the pathophysiology of hypertension during preeclampsia. Hypertension 51: 982-8
62. Hawinkels LJ, Kuiper P, Wiercinska E, Verspaget HW, Liu Z, Pardali E, Sier CF, ten
Dijke P. 2010. Matrix metalloproteinase-14 (MT1-MMP)-mediated endoglin shedding
inhibits tumor angiogenesis. Cancer Res 70: 4141-50
71
63. Kaitu'u-Lino TJ, Palmer KR, Whitehead CL, Williams E, Lappas M, Tong S. 2012.
MMP-14 is expressed in preeclamptic placentas and mediates release of soluble endoglin.
Am J Pathol 180: 888-94
64. Walshe TE, Saint-Geniez M, Maharaj AS, Sekiyama E, Maldonado AE, D'Amore PA.
2009. TGF-beta is required for vascular barrier function, endothelial survival and
homeostasis of the adult microvasculature. PLoS One 4: e5149
65. Alt A, Miguel-Romero L, Donderis J, Aristorena M, Blanco FJ, Round A, Rubio V,
Bernabeu C, Marina A. 2012. Structural and functional insights into endoglin ligand
recognition and binding. PLoS One 7: e29948
66. Siegel PM, Massague J. 2003. Cytostatic and apoptotic actions of TGF-beta in
homeostasis and cancer. Nat Rev Cancer 3: 807-21
67. Schmierer B, Hill CS. 2007. TGFbeta-SMAD signal transduction: molecular specificity
and functional flexibility. Nat Rev Mol Cell Biol 8: 970-82
68. Gougos A, Letarte M. 1990. Primary structure of endoglin, an RGD-containing
glycoprotein of human endothelial cells. J Biol Chem 265: 8361-4
69. Lastres P, Letamendia A, Zhang H, Rius C, Almendro N, Raab U, Lopez LA, Langa C,
Fabra A, Letarte M, Bernabeu C. 1996. Endoglin modulates cellular responses to TGF-
beta 1. J Cell Biol 133: 1109-21
70. Sankar S, Mahooti-Brooks N, Centrella M, McCarthy TL, Madri JA. 1995. Expression of
transforming growth factor type III receptor in vascular endothelial cells increases their
responsiveness to transforming growth factor beta 2. J Biol Chem 270: 13567-72
71. Nolan-Stevaux O, Zhong W, Culp S, Shaffer K, Hoover J, Wickramasinghe D, Ruefli-
Brasse A. 2012. Endoglin requirement for BMP9 signaling in endothelial cells reveals
new mechanism of action for selective anti-endoglin antibodies. PLoS One 7: e50920
72. Luo K, Lodish HF. 1997. Positive and negative regulation of type II TGF-beta receptor
signal transduction by autophosphorylation on multiple serine residues. EMBO J 16:
1970-81
73. Zhang W, Jiang Y, Wang Q, Ma X, Xiao Z, Zuo W, Fang X, Chen YG. 2009. Single-
molecule imaging reveals transforming growth factor-beta-induced type II receptor
dimerization. Proc Natl Acad Sci U S A 106: 15679-83
74. Zhang W, Yuan J, Yang Y, Xu L, Wang Q, Zuo W, Fang X, Chen YG. 2010. Monomeric
type I and type III transforming growth factor-beta receptors and their dimerization
revealed by single-molecule imaging. Cell Res 20: 1216-23
75. Wieser R, Wrana JL, Massague J. 1995. GS domain mutations that constitutively activate
T beta R-I, the downstream signaling component in the TGF-beta receptor complex.
EMBO J 14: 2199-208
72
76. Batut J, Schmierer B, Cao J, Raftery LA, Hill CS, Howell M. 2008. Two highly related
regulatory subunits of PP2A exert opposite effects on TGF-beta/Activin/Nodal signalling.
Development 135: 2927-37
77. Huang F, Chen YG. 2012. Regulation of TGF-beta receptor activity. Cell Biosci 2: 9
78. Goumans MJ, Valdimarsdottir G, Itoh S, Lebrin F, Larsson J, Mummery C, Karlsson S,
ten Dijke P. 2003. Activin receptor-like kinase (ALK)1 is an antagonistic mediator of
lateral TGFbeta/ALK5 signaling. Mol Cell 12: 817-28
79. Scharpfenecker M, van Dinther M, Liu Z, van Bezooijen RL, Zhao Q, Pukac L, Lowik
CW, ten Dijke P. 2007. BMP-9 signals via ALK1 and inhibits bFGF-induced endothelial
cell proliferation and VEGF-stimulated angiogenesis. J Cell Sci 120: 964-72
80. Guerrero-Esteo M, Sanchez-Elsner T, Letamendia A, Bernabeu C. 2002. Extracellular
and cytoplasmic domains of endoglin interact with the transforming growth factor-beta
receptors I and II. J Biol Chem 277: 29197-209
81. Cheifetz S, Bellon T, Cales C, Vera S, Bernabeu C, Massague J, Letarte M. 1992.
Endoglin is a component of the transforming growth factor-beta receptor system in
human endothelial cells. J Biol Chem 267: 19027-30
82. Finnson KW, Parker WL, Chi Y, Hoemann CD, Goldring MB, Antoniou J, Philip A.
2010. Endoglin differentially regulates TGF-beta-induced Smad2/3 and Smad1/5
signalling and its expression correlates with extracellular matrix production and cellular
differentiation state in human chondrocytes. Osteoarthritis Cartilage 18: 1518-27
83. David L, Mallet C, Mazerbourg S, Feige JJ, Bailly S. 2007. Identification of BMP9 and
BMP10 as functional activators of the orphan activin receptor-like kinase 1 (ALK1) in
endothelial cells. Blood 109: 1953-61
84. Lee NY, Ray B, How T, Blobe GC. 2008. Endoglin promotes transforming growth factor
beta-mediated Smad 1/5/8 signaling and inhibits endothelial cell migration through its
association with GIPC. J Biol Chem 283: 32527-33
85. Santibanez JF, Letamendia A, Perez-Barriocanal F, Silvestri C, Saura M, Vary CP,
Lopez-Novoa JM, Attisano L, Bernabeu C. 2007. Endoglin increases eNOS expression
by modulating Smad2 protein levels and Smad2-dependent TGF-beta signaling. J Cell
Physiol 210: 456-68
86. Lebrin F, Deckers M, Bertolino P, Ten Dijke P. 2005. TGF-beta receptor function in the
endothelium. Cardiovasc Res 65: 599-608
87. Wu JW, Hu M, Chai J, Seoane J, Huse M, Li C, Rigotti DJ, Kyin S, Muir TW, Fairman
R, Massague J, Shi Y. 2001. Crystal structure of a phosphorylated Smad2. Recognition of
phosphoserine by the MH2 domain and insights on Smad function in TGF-beta signaling.
Mol Cell 8: 1277-89
73
88. Kirkbride KC, Ray BN, Blobe GC. 2005. Cell-surface co-receptors: emerging roles in
signaling and human disease. Trends Biochem Sci 30: 611-21
89. ten Dijke P, Hill CS. 2004. New insights into TGF-beta-Smad signalling. Trends
Biochem Sci 29: 265-73
90. Imamura T, Takase M, Nishihara A, Oeda E, Hanai J, Kawabata M, Miyazono K. 1997.
Smad6 inhibits signalling by the TGF-beta superfamily. Nature 389: 622-6
91. Kavsak P, Rasmussen RK, Causing CG, Bonni S, Zhu H, Thomsen GH, Wrana JL. 2000.
Smad7 binds to Smurf2 to form an E3 ubiquitin ligase that targets the TGF beta receptor
for degradation. Mol Cell 6: 1365-75
92. Shi W, Sun C, He B, Xiong W, Shi X, Yao D, Cao X. 2004. GADD34-PP1c recruited by
Smad7 dephosphorylates TGFbeta type I receptor. J Cell Biol 164: 291-300
93. Meurer SK, Esser M, Tihaa L, Weiskirchen R. 2012. BMP-7/TGF-beta1 signalling in
myoblasts: components involved in signalling and BMP-7-dependent blockage of TGF-
beta-mediated CTGF expression. Eur J Cell Biol 91: 450-63
94. Weston CR, Davis RJ. 2007. The JNK signal transduction pathway. Curr Opin Cell Biol
19: 142-9
95. Shim JH, Xiao C, Paschal AE, Bailey ST, Rao P, Hayden MS, Lee KY, Bussey C,
Steckel M, Tanaka N, Yamada G, Akira S, Matsumoto K, Ghosh S. 2005. TAK1, but not
TAB1 or TAB2, plays an essential role in multiple signaling pathways in vivo. Genes
Dev 19: 2668-81
96. Jadrich JL, O'Connor MB, Coucouvanis E. 2006. The TGF beta activated kinase TAK1
regulates vascular development in vivo. Development 133: 1529-41
97. Sorrentino A, Thakur N, Grimsby S, Marcusson A, von Bulow V, Schuster N, Zhang S,
Heldin CH, Landstrom M. 2008. The type I TGF-beta receptor engages TRAF6 to
activate TAK1 in a receptor kinase-independent manner. Nat Cell Biol 10: 1199-207
98. Hocevar BA, Brown TL, Howe PH. 1999. TGF-beta induces fibronectin synthesis
through a c-Jun N-terminal kinase-dependent, Smad4-independent pathway. EMBO J 18:
1345-56
99. Lawler S, Feng XH, Chen RH, Maruoka EM, Turck CW, Griswold-Prenner I, Derynck
R. 1997. The type II transforming growth factor-beta receptor autophosphorylates not
only on serine and threonine but also on tyrosine residues. J Biol Chem 272: 14850-9
100. Hartsough MT, Mulder KM. 1995. Transforming growth factor beta activation of
p44mapk in proliferating cultures of epithelial cells. J Biol Chem 270: 7117-24
101. Simeone DM, Zhang L, Graziano K, Nicke B, Pham T, Schaefer C, Logsdon CD. 2001.
Smad4 mediates activation of mitogen-activated protein kinases by TGF-beta in
pancreatic acinar cells. Am J Physiol Cell Physiol 281: C311-9
74
102. Lee MK, Pardoux C, Hall MC, Lee PS, Warburton D, Qing J, Smith SM, Derynck R.
2007. TGF-beta activates Erk MAP kinase signalling through direct phosphorylation of
ShcA. EMBO J 26: 3957-67
103. Zhang YE. 2009. Non-Smad pathways in TGF-beta signaling. Cell Res 19: 128-39
104. Wilkes MC, Mitchell H, Penheiter SG, Dore JJ, Suzuki K, Edens M, Sharma DK, Pagano
RE, Leof EB. 2005. Transforming growth factor-beta activation of phosphatidylinositol
3-kinase is independent of Smad2 and Smad3 and regulates fibroblast responses via p21-
activated kinase-2. Cancer Res 65: 10431-40
105. Yi JY, Shin I, Arteaga CL. 2005. Type I transforming growth factor beta receptor binds
to and activates phosphatidylinositol 3-kinase. J Biol Chem 280: 10870-6
106. Lamouille S, Derynck R. 2007. Cell size and invasion in TGF-beta-induced epithelial to
mesenchymal transition is regulated by activation of the mTOR pathway. J Cell Biol 178:
437-51
107. Remy I, Montmarquette A, Michnick SW. 2004. PKB/Akt modulates TGF-beta
signalling through a direct interaction with Smad3. Nat Cell Biol 6: 358-65
108. Miyazono K, Kamiya Y, Morikawa M. 2010. Bone morphogenetic protein receptors and
signal transduction. J Biochem 147: 35-51
109. Lai CF, Cheng SL. 2002. Signal transductions induced by bone morphogenetic protein-2
and transforming growth factor-beta in normal human osteoblastic cells. J Biol Chem
277: 15514-22
110. Li B, Yang Y, Jiang S, Ni B, Chen K, Jiang L. 2012. Adenovirus-mediated
overexpression of BMP-9 inhibits human osteosarcoma cell growth and migration
through downregulation of the PI3K/AKT pathway. Int J Oncol 41: 1809-19
111. Annes JP, Munger JS, Rifkin DB. 2003. Making sense of latent TGFbeta activation. J
Cell Sci 116: 217-24
112. Gleizes PE, Beavis RC, Mazzieri R, Shen B, Rifkin DB. 1996. Identification and
characterization of an eight-cysteine repeat of the latent transforming growth factor-beta
binding protein-1 that mediates bonding to the latent transforming growth factor-beta1. J
Biol Chem 271: 29891-6
113. Ivanovic V, Todorovic-Rakovic N, Demajo M, Neskovic-Konstantinovic Z, Subota V,
Ivanisevic-Milovanovic O, Nikolic-Vukosavljevic D. 2003. Elevated plasma levels of
transforming growth factor-beta 1 (TGF-beta 1) in patients with advanced breast cancer:
association with disease progression. Eur J Cancer 39: 454-61
114. Singer AJ, Clark RA. 1999. Cutaneous wound healing. N Engl J Med 341: 738-46
75
115. Schmid P, Cox D, Bilbe G, McMaster G, Morrison C, Stahelin H, Luscher N, Seiler W.
1993. TGF-beta s and TGF-beta type II receptor in human epidermis: differential
expression in acute and chronic skin wounds. J Pathol 171: 191-7
116. Pierce GF, Mustoe TA, Lingelbach J, Masakowski VR, Gramates P, Deuel TF. 1989.
Transforming growth factor beta reverses the glucocorticoid-induced wound-healing
deficit in rats: possible regulation in macrophages by platelet-derived growth factor. Proc
Natl Acad Sci U S A 86: 2229-33
117. Ashcroft GS, Yang X, Glick AB, Weinstein M, Letterio JL, Mizel DE, Anzano M,
Greenwell-Wild T, Wahl SM, Deng C, Roberts AB. 1999. Mice lacking Smad3 show
accelerated wound healing and an impaired local inflammatory response. Nat Cell Biol 1:
260-6
118. Amendt C, Mann A, Schirmacher P, Blessing M. 2002. Resistance of keratinocytes to
TGFbeta-mediated growth restriction and apoptosis induction accelerates re-
epithelialization in skin wounds. J Cell Sci 115: 2189-98
119. Saika S, Kono-Saika S, Ohnishi Y, Sato M, Muragaki Y, Ooshima A, Flanders KC, Yoo
J, Anzano M, Liu CY, Kao WW, Roberts AB. 2004. Smad3 signaling is required for
epithelial-mesenchymal transition of lens epithelium after injury. Am J Pathol 164: 651-
63
120. Saika S, Ikeda K, Yamanaka O, Miyamoto T, Ohnishi Y, Sato M, Muragaki Y, Ooshima
A, Nakajima Y, Kao WW, Flanders KC, Roberts AB. 2005. Expression of Smad7 in
mouse eyes accelerates healing of corneal tissue after exposure to alkali. Am J Pathol
166: 1405-18
121. Brabletz T, Pfeuffer I, Schorr E, Siebelt F, Wirth T, Serfling E. 1993. Transforming
growth factor beta and cyclosporin A inhibit the inducible activity of the interleukin-2
gene in T cells through a noncanonical octamer-binding site. Mol Cell Biol 13: 1155-62
122. Chen W, Jin W, Hardegen N, Lei KJ, Li L, Marinos N, McGrady G, Wahl SM. 2003.
Conversion of peripheral CD4+CD25- naive T cells to CD4+CD25+ regulatory T cells
by TGF-beta induction of transcription factor Foxp3. J Exp Med 198: 1875-86
123. Yoshimura A, Muto G. 2011. TGF-beta function in immune suppression. Curr Top
Microbiol Immunol 350: 127-47
124. Goumans MJ, Mummery C. 2000. Functional analysis of the TGFbeta receptor/Smad
pathway through gene ablation in mice. Int J Dev Biol 44: 253-65
125. Deng G, Curriden SA, Wang S, Rosenberg S, Loskutoff DJ. 1996. Is plasminogen
activator inhibitor-1 the molecular switch that governs urokinase receptor-mediated cell
adhesion and release? J Cell Biol 134: 1563-71
126. Finnson KW, Parker WL, ten Dijke P, Thorikay M, Philip A. 2008. ALK1 opposes
ALK5/Smad3 signaling and expression of extracellular matrix components in human
chondrocytes. J Bone Miner Res 23: 896-906
76
127. Jakowlew SB. 2006. Transforming growth factor-beta in cancer and metastasis. Cancer
Metastasis Rev 25: 435-57
128. Brown MA, Zhao Q, Baker KA, Naik C, Chen C, Pukac L, Singh M, Tsareva T, Parice
Y, Mahoney A, Roschke V, Sanyal I, Choe S. 2005. Crystal structure of BMP-9 and
functional interactions with pro-region and receptors. J Biol Chem 280: 25111-8
129. Bidart M, Ricard N, Levet S, Samson M, Mallet C, David L, Subileau M, Tillet E, Feige
JJ, Bailly S. 2012. BMP9 is produced by hepatocytes and circulates mainly in an active
mature form complexed to its prodomain. Cell Mol Life Sci 69: 313-24
130. Ricard N, Ciais D, Levet S, Subileau M, Mallet C, Zimmers TA, Lee SJ, Bidart M, Feige
JJ, Bailly S. 2012. BMP9 and BMP10 are critical for postnatal retinal vascular
remodeling. Blood 119: 6162-71
131. Johnson DW, Berg JN, Baldwin MA, Gallione CJ, Marondel I, Yoon SJ, Stenzel TT,
Speer M, Pericak-Vance MA, Diamond A, Guttmacher AE, Jackson CE, Attisano L,
Kucherlapati R, Porteous ME, Marchuk DA. 1996. Mutations in the activin receptor-like
kinase 1 gene in hereditary haemorrhagic telangiectasia type 2. Nat Genet 13: 189-95
132. David L, Mallet C, Keramidas M, Lamande N, Gasc JM, Dupuis-Girod S, Plauchu H,
Feige JJ, Bailly S. 2008. Bone morphogenetic protein-9 is a circulating vascular
quiescence factor. Circ Res 102: 914-22
133. Suzuki Y, Ohga N, Morishita Y, Hida K, Miyazono K, Watabe T. 2010. BMP-9 induces
proliferation of multiple types of endothelial cells in vitro and in vivo. J Cell Sci 123:
1684-92
134. Kim JH, Peacock MR, George SC, Hughes CC. 2012. BMP9 induces EphrinB2
expression in endothelial cells through an Alk1-BMPRII/ActRII-ID1/ID3-dependent
pathway: implications for hereditary hemorrhagic telangiectasia type II. Angiogenesis 15:
497-509
135. Lyden D, Young AZ, Zagzag D, Yan W, Gerald W, O'Reilly R, Bader BL, Hynes RO,
Zhuang Y, Manova K, Benezra R. 1999. Id1 and Id3 are required for neurogenesis,
angiogenesis and vascularization of tumour xenografts. Nature 401: 670-7
136. Cunha SI, Pardali E, Thorikay M, Anderberg C, Hawinkels L, Goumans MJ, Seehra J,
Heldin CH, ten Dijke P, Pietras K. 2010. Genetic and pharmacological targeting of
activin receptor-like kinase 1 impairs tumor growth and angiogenesis. J Exp Med 207:
85-100
137. Li JZ, Li H, Sasaki T, Holman D, Beres B, Dumont RJ, Pittman DD, Hankins GR, Helm
GA. 2003. Osteogenic potential of five different recombinant human bone morphogenetic
protein adenoviral vectors in the rat. Gene Ther 10: 1735-43
77
138. Chen C, Grzegorzewski KJ, Barash S, Zhao Q, Schneider H, Wang Q, Singh M, Pukac L,
Bell AC, Duan R, Coleman T, Duttaroy A, Cheng S, Hirsch J, Zhang L, Lazard Y,
Fischer C, Barber MC, Ma ZD, Zhang YQ, Reavey P, Zhong L, Teng B, Sanyal I, Ruben
SM, Blondel O, Birse CE. 2003. An integrated functional genomics screening program
reveals a role for BMP-9 in glucose homeostasis. Nat Biotechnol 21: 294-301
139. Park JE, Shao D, Upton PD, Desouza P, Adcock IM, Davies RJ, Morrell NW, Griffiths
MJ, Wort SJ. 2012. BMP-9 induced endothelial cell tubule formation and inhibition of
migration involves Smad1 driven endothelin-1 production. PLoS One 7: e30075
140. Star GP, Giovinazzo M, Langleben D. 2010. Bone morphogenic protein-9 stimulates
endothelin-1 release from human pulmonary microvascular endothelial cells: a potential
mechanism for elevated ET-1 levels in pulmonary arterial hypertension. Microvasc Res
80: 349-54
141. Hickey KA, Rubanyi G, Paul RJ, Highsmith RF. 1985. Characterization of a coronary
vasoconstrictor produced by cultured endothelial cells. Am J Physiol 248: C550-6
142. Inoue A, Yanagisawa M, Kimura S, Kasuya Y, Miyauchi T, Goto K, Masaki T. 1989.
The human endothelin family: three structurally and pharmacologically distinct
isopeptides predicted by three separate genes. Proc Natl Acad Sci U S A 86: 2863-7
143. Blais V, Fugere M, Denault JB, Klarskov K, Day R, Leduc R. 2002. Processing of
proendothelin-1 by members of the subtilisin-like pro-protein convertase family. FEBS
Lett 524: 43-8
144. Shimada K, Matsushita Y, Wakabayashi K, Takahashi M, Matsubara A, Iijima Y,
Tanzawa K. 1995. Cloning and functional expression of human endothelin-converting
enzyme cDNA. Biochem Biophys Res Commun 207: 807-12
145. Schmidt M, Kroger B, Jacob E, Seulberger H, Subkowski T, Otter R, Meyer T,
Schmalzing G, Hillen H. 1994. Molecular characterization of human and bovine
endothelin converting enzyme (ECE-1). FEBS Lett 356: 238-43
146. Bohm F, Pernow J. 2007. The importance of endothelin-1 for vascular dysfunction in
cardiovascular disease. Cardiovasc Res 76: 8-18
147. Arai H, Hori S, Aramori I, Ohkubo H, Nakanishi S. 1990. Cloning and expression of a
cDNA encoding an endothelin receptor. Nature 348: 730-2
148. Verhaar MC, Strachan FE, Newby DE, Cruden NL, Koomans HA, Rabelink TJ, Webb
DJ. 1998. Endothelin-A receptor antagonist-mediated vasodilatation is attenuated by
inhibition of nitric oxide synthesis and by endothelin-B receptor blockade. Circulation
97: 752-6
149. Masaki T. 2004. Historical review: Endothelin. Trends Pharmacol Sci 25: 219-24
78
150. Sanchez R, MacKenzie A, Farhat N, Nguyen TD, Stewart DJ, Mercier I, Calderone A,
Thorin E. 2002. Endothelin B receptor-mediated regulation of endothelin-1 content and
release in cultured porcine aorta endothelial cell. J Cardiovasc Pharmacol 39: 652-9
151. Miwa S, Iwamuro Y, Zhang XF, Inoki T, Okamoto Y, Okazawa M, Masaki T. 1999.
Ca2+ entry channels in rat thoracic aortic smooth muscle cells activated by endothelin-1.
Jpn J Pharmacol 80: 281-8
152. Alonso D, Radomski MW. 2003. The nitric oxide-endothelin-1 connection. Heart Fail
Rev 8: 107-15
153. Douglas SA, Sulpizio AC, Piercy V, Sarau HM, Ames RS, Aiyar NV, Ohlstein EH,
Willette RN. 2000. Differential vasoconstrictor activity of human urotensin-II in vascular
tissue isolated from the rat, mouse, dog, pig, marmoset and cynomolgus monkey. Br J
Pharmacol 131: 1262-74
154. Weitzberg E, Ahlborg G, Lundberg JM. 1993. Differences in vascular effects and
removal of endothelin-1 in human lung, brain, and skeletal muscle. Clin Physiol 13: 653-
62
155. Haynes WG, Ferro CJ, O'Kane KP, Somerville D, Lomax CC, Webb DJ. 1996. Systemic
endothelin receptor blockade decreases peripheral vascular resistance and blood pressure
in humans. Circulation 93: 1860-70
156. Kurihara Y, Kurihara H, Suzuki H, Kodama T, Maemura K, Nagai R, Oda H, Kuwaki T,
Cao WH, Kamada N, et al. 1994. Elevated blood pressure and craniofacial abnormalities
in mice deficient in endothelin-1. Nature 368: 703-10
157. Clouthier DE, Hosoda K, Richardson JA, Williams SC, Yanagisawa H, Kuwaki T,
Kumada M, Hammer RE, Yanagisawa M. 1998. Cranial and cardiac neural crest defects
in endothelin-A receptor-deficient mice. Development 125: 813-24
158. Berthiaume N, Yanagisawa M, Labonte J, D'Orleans-Juste P. 2000. Heterozygous knock-
Out of ET(B) receptors induces BQ-123-sensitive hypertension in the mouse.
Hypertension 36: 1002-7
159. Yang LL, Gros R, Kabir MG, Sadi A, Gotlieb AI, Husain M, Stewart DJ. 2004.
Conditional cardiac overexpression of endothelin-1 induces inflammation and dilated
cardiomyopathy in mice. Circulation 109: 255-61
160. Gonon AT, Gourine AV, Middelveld RJ, Alving K, Pernow J. 2001. Limitation of infarct
size and attenuation of myeloperoxidase activity by an endothelin A receptor antagonist
following ischaemia and reperfusion. Basic Res Cardiol 96: 454-62
161. Taylor RN, Varma M, Teng NN, Roberts JM. 1990. Women with preeclampsia have
higher plasma endothelin levels than women with normal pregnancies. J Clin Endocrinol
Metab 71: 1675-7
79
162. Nishikawa S, Miyamoto A, Yamamoto H, Ohshika H, Kudo R. 2000. The relationship
between serum nitrate and endothelin-1 concentrations in preeclampsia. Life Sci 67:
1447-54
163. Ajne G, Wolff K, Fyhrquist F, Carlstrom K, Hemsen-Mortberg A, Nisell H. 2003.
Endothelin converting enzyme (ECE) activity in normal pregnancy and preeclampsia.
Hypertens Pregnancy 22: 215-24
164. Fiore G, Florio P, Micheli L, Nencini C, Rossi M, Cerretani D, Ambrosini G, Giorgi G,
Petraglia F. 2005. Endothelin-1 triggers placental oxidative stress pathways: putative role
in preeclampsia. J Clin Endocrinol Metab 90: 4205-10
165. Li J, LaMarca B, Reckelhoff JF. 2012. A model of preeclampsia in rats: the reduced
uterine perfusion pressure (RUPP) model. Am J Physiol Heart Circ Physiol 303: H1-8
166. Sunderland N, Hennessy A, Makris A. 2011. Animal models of pre-eclampsia. Am J
Reprod Immunol 65: 533-41
167. Isler CM, Bennett WA, Rinewalt AN, Cockrell KL, Martin JN, Jr., Morrison JC, Granger
JP. 2003. Evaluation of a rat model of preeclampsia for HELLP syndrome characteristics.
J Soc Gynecol Investig 10: 151-3
168. Kiprono LV, Wallace K, Moseley J, Martin J, Jr., Lamarca B. 2013. Progesterone blunts
vascular endothelial cell secretion of endothelin-1 in response to placental ischemia. Am J
Obstet Gynecol 209: 44 e1- e6
169. Abdalvand A, Morton JS, Bourque SL, Quon AL, Davidge ST. 2013. Matrix
metalloproteinase enhances big-endothelin-1 constriction in mesenteric vessels of
pregnant rats with reduced uterine blood flow. Hypertension 61: 488-93
170. Roberts L, LaMarca BB, Fournier L, Bain J, Cockrell K, Granger JP. 2006. Enhanced
endothelin synthesis by endothelial cells exposed to sera from pregnant rats with
decreased uterine perfusion. Hypertension 47: 615-8
171. Alexander BT, Rinewalt AN, Cockrell KL, Massey MB, Bennett WA, Granger JP. 2001.
Endothelin type a receptor blockade attenuates the hypertension in response to chronic
reductions in uterine perfusion pressure. Hypertension 37: 485-9
172. Zhou CC, Irani RA, Dai Y, Blackwell SC, Hicks MJ, Ramin SM, Kellems RE, Xia Y.
2011. Autoantibody-mediated IL-6-dependent endothelin-1 elevation underlies
pathogenesis in a mouse model of preeclampsia. J Immunol 186: 6024-34
173. Murphy JA, Archibald ML, Chauhan BC. 2010. The role of endothelin-1 and its
receptors in optic nerve head astrocyte proliferation. Br J Ophthalmol 94: 1233-8
174. Kingman M, Ruggiero R, Torres F. 2009. Ambrisentan, an endothelin receptor type A-
selective endothelin receptor antagonist, for the treatment of pulmonary arterial
hypertension. Expert Opin Pharmacother 10: 1847-58
80
175. Taniguchi T, Muramatsu I. 2003. Pharmacological knockout of endothelin ET(A)
receptors. Life Sci 74: 405-9
176. Balconi G, Spagnuolo R, Dejana E. 2000. Development of endothelial cell lines from
embryonic stem cells: A tool for studying genetically manipulated endothelial cells in
vitro. Arterioscler Thromb Vasc Biol 20: 1443-51
177. Pece-Barbara N, Vera S, Kathirkamathamby K, Liebner S, Di Guglielmo GM, Dejana E,
Wrana JL, Letarte M. 2005. Endoglin null endothelial cells proliferate faster and are more
responsive to transforming growth factor beta1 with higher affinity receptors and an
activated Alk1 pathway. J Biol Chem 280: 27800-8
178. George EM, Granger JP. 2011. Endothelin: key mediator of hypertension in
preeclampsia. Am J Hypertens 24: 964-9
179. Scherner O, Meurer SK, Tihaa L, Gressner AM, Weiskirchen R. 2007. Endoglin
differentially modulates antagonistic transforming growth factor-beta1 and BMP-7
signaling. J Biol Chem 282: 13934-43
180. Toporsian M, Gros R, Kabir MG, Vera S, Govindaraju K, Eidelman DH, Husain M,
Letarte M. 2005. A role for endoglin in coupling eNOS activity and regulating vascular
tone revealed in hereditary hemorrhagic telangiectasia. Circ Res 96: 684-92
181. Castonguay R, Werner ED, Matthews RG, Presman E, Mulivor AW, Solban N, Sako D,
Pearsall RS, Underwood KW, Seehra J, Kumar R, Grinberg AV. 2011. Soluble endoglin
specifically binds bone morphogenetic proteins 9 and 10 via its orphan domain, inhibits
blood vessel formation, and suppresses tumor growth. J Biol Chem 286: 30034-46
182. Hunter A, Aitkenhead M, Caldwell C, McCracken G, Wilson D, McClure N. 2000.
Serum levels of vascular endothelial growth factor in preeclamptic and normotensive
pregnancy. Hypertension 36: 965-9
183. Rana S, Hacker MR, Modest AM, Salahuddin S, Lim KH, Verlohren S, Perschel FH,
Karumanchi SA. 2012. Circulating angiogenic factors and risk of adverse maternal and
perinatal outcomes in twin pregnancies with suspected preeclampsia. Hypertension 60:
451-8
184. Smith GC, Crossley JA, Aitken DA, Jenkins N, Lyall F, Cameron AD, Connor JM,
Dobbie R. 2007. Circulating angiogenic factors in early pregnancy and the risk of
preeclampsia, intrauterine growth restriction, spontaneous preterm birth, and stillbirth.
Obstet Gynecol 109: 1316-24
185. Nakao A, Imamura T, Souchelnytskyi S, Kawabata M, Ishisaki A, Oeda E, Tamaki K,
Hanai J, Heldin CH, Miyazono K, ten Dijke P. 1997. TGF-beta receptor-mediated
signalling through Smad2, Smad3 and Smad4. EMBO J 16: 5353-62
186. Xu DJ, Zhao YZ, Wang J, He JW, Weng YG, Luo JY. 2012. Smads, p38 and ERK1/2 are
involved in BMP9-induced osteogenic differentiation of C3H10T1/2 mesenchymal stem
cells. BMB Rep 45: 247-52
81
187. Akhurst RJ, Hata A. 2012. Targeting the TGFbeta signalling pathway in disease. Nat Rev
Drug Discov 11: 790-811
188. Hao J, Ho JN, Lewis JA, Karim KA, Daniels RN, Gentry PR, Hopkins CR, Lindsley CW,
Hong CC. 2010. In vivo structure-activity relationship study of dorsomorphin analogues
identifies selective VEGF and BMP inhibitors. ACS Chem Biol 5: 245-53
189. Boergermann JH, Kopf J, Yu PB, Knaus P. 2010. Dorsomorphin and LDN-193189
inhibit BMP-mediated Smad, p38 and Akt signalling in C2C12 cells. Int J Biochem Cell
Biol 42: 1802-7
190. Wang K, Feng H, Ren W, Sun X, Luo J, Tang M, Zhou L, Weng Y, He TC, Zhang Y.
2011. BMP9 inhibits the proliferation and invasiveness of breast cancer cells MDA-MB-
231. J Cancer Res Clin Oncol 137: 1687-96
191. Mitchell D, Pobre EG, Mulivor AW, Grinberg AV, Castonguay R, Monnell TE, Solban
N, Ucran JA, Pearsall RS, Underwood KW, Seehra J, Kumar R. 2010. ALK1-Fc inhibits
multiple mediators of angiogenesis and suppresses tumor growth. Mol Cancer Ther 9:
379-88
192. Schoofs K, Grittner U, Engels T, Pape J, Denk B, Henrich W, Verlohren S. 2013. The
importance of repeated measurements of the sFlt-1/PlGF ratio for the prediction of
preeclampsia and intrauterine growth restriction. J Perinat Med: 1-8
193. Chen Y, Medhora M, Falck JR, Pritchard KA, Jr., Jacobs ER. 2006. Mechanisms of
activation of eNOS by 20-HETE and VEGF in bovine pulmonary artery endothelial cells.
Am J Physiol Lung Cell Mol Physiol 291: L378-85
194. Lopez-Jaramillo P, Narvaez M, Calle A, Rivera J, Jacome P, Ruano C, Nava E. 1996.
Cyclic guanosine 3',5' monophosphate concentrations in pre-eclampsia: effects of
hydralazine. Br J Obstet Gynaecol 103: 33-8
195. Silver RK, Kupferminc MJ, Russell TL, Adler L, Mullen TA, Caplan MS. 1996.
Evaluation of nitric oxide as a mediator of severe preeclampsia. Am J Obstet Gynecol
175: 1013-7
196. Nobunaga T, Tokugawa Y, Hashimoto K, Kimura T, Matsuzaki N, Nitta Y, Fujita T,
Kidoguchi KI, Azuma C, Saji F. 1996. Plasma nitric oxide levels in pregnant patients
with preeclampsia and essential hypertension. Gynecol Obstet Invest 41: 189-93
197. Donaubauer B, Busch T, Lachmann R, Deja M, Petersen B, Francis R, Trager A, Ebsen
M, Boemke W, Kaisers U. 2006. Low-dose inhalation of an endothelin-A receptor
antagonist in experimental acute lung injury: ET-1 plasma concentration and pulmonary
inflammation. Exp Biol Med (Maywood) 231: 960-6
198. Li JL, Harris AL. 2009. Crosstalk of VEGF and Notch pathways in tumour angiogenesis:
therapeutic implications. Front Biosci (Landmark Ed) 14: 3094-110
82
199. Tam J, Duda DG, Perentes JY, Quadri RS, Fukumura D, Jain RK. 2009. Blockade of
VEGFR2 and not VEGFR1 can limit diet-induced fat tissue expansion: role of local
versus bone marrow-derived endothelial cells. PLoS One 4: e4974
200. Koolwijk P, Peters E, van der Vecht B, Hornig C, Weich HA, Alitalo K, Hicklin DJ, Wu
Y, Witte L, van Hinsbergh VW. 2001. Involvement of VEGFR-2 (kdr/flk-1) but not
VEGFR-1 (flt-1) in VEGF-A and VEGF-C-induced tube formation by human
microvascular endothelial cells in fibrin matrices in vitro. Angiogenesis 4: 53-60
201. Velasco S, Alvarez-Munoz P, Pericacho M, Dijke PT, Bernabeu C, Lopez-Novoa JM,
Rodriguez-Barbero A. 2008. L- and S-endoglin differentially modulate TGFbeta1
signaling mediated by ALK1 and ALK5 in L6E9 myoblasts. J Cell Sci 121: 913-9
202. Mallet C, Vittet D, Feige JJ, Bailly S. 2006. TGFbeta1 induces vasculogenesis and
inhibits angiogenic sprouting in an embryonic stem cell differentiation model: respective
contribution of ALK1 and ALK5. Stem Cells 24: 2420-7
203. Castanares C, Redondo-Horcajo M, Magan-Marchal N, ten Dijke P, Lamas S, Rodriguez-
Pascual F. 2007. Signaling by ALK5 mediates TGF-beta-induced ET-1 expression in
endothelial cells: a role for migration and proliferation. J Cell Sci 120: 1256-66
204. Takekawa M, Tatebayashi K, Itoh F, Adachi M, Imai K, Saito H. 2002. Smad-dependent
GADD45beta expression mediates delayed activation of p38 MAP kinase by TGF-beta.
EMBO J 21: 6473-82
205. Schiffer M, Bitzer M, Roberts IS, Kopp JB, ten Dijke P, Mundel P, Bottinger EP. 2001.
Apoptosis in podocytes induced by TGF-beta and Smad7. J Clin Invest 108: 807-16
206. Wang L, Ma R, Flavell RA, Choi ME. 2002. Requirement of mitogen-activated protein
kinase kinase 3 (MKK3) for activation of p38alpha and p38delta MAPK isoforms by
TGF-beta 1 in murine mesangial cells. J Biol Chem 277: 47257-62
207. Ferrari G, Terushkin V, Wolff MJ, Zhang X, Valacca C, Poggio P, Pintucci G, Mignatti
P. 2012. TGF-beta1 induces endothelial cell apoptosis by shifting VEGF activation of
p38(MAPK) from the prosurvival p38beta to proapoptotic p38alpha. Mol Cancer Res 10:
605-14
208. Abecassis L, Rogier E, Vazquez A, Atfi A, Bourgeade MF. 2004. Evidence for a role of
MSK1 in transforming growth factor-beta-mediated responses through p38alpha and
Smad signaling pathways. J Biol Chem 279: 30474-9