-
*Molecular Cell Biology, Leiden University Medical Center,
Postbus 9600, 2300 RC Leiden, The Netherlands.Institute of Human
Genetics, International Centre for Life, Central Parkway, Newcastle
University, NE1 3BZ, UK.emails: [email protected];
[email protected]:10.1038/nrm2262Published online 26
September 2007
Extracellular control of TGF signalling in vascular development
and diseasePeter ten Dijke* and Helen M. Arthur
Abstract | The intracellular mechanism of transforming growth
factor (TGF) signalling via kinase receptors and SMAD effectors is
firmly established, but recent studies of human cardiovascular
syndromes such as Marfan syndrome and preeclampsia have refocused
attention on the importance of regulating the availability of
active extracellular TGF. It seems that elastic extracellular
matrix (ECM) components have a crucial role in controlling TGF
signalling, while soluble and membrane bound forms of TGF
coreceptors add further layers of regulation. Together, these
extracellular interactions determine the final bioavailability of
TGF to vascular cells, and dysregulation is associated with an
increasing number of vascular pathologies.
Transforming growth factor1 (TGF1) is the prototypic member of a
large family of evolutionarily conserved pleio-tropic secreted
cytokines, which also includes the activins and bone morphogenetic
proteins (BMPs). Individual family members have crucial roles in
multiple processes throughout development and in the maintenance of
tissue homeostasis in adult life1,2. Not surprisingly, therefore,
subversion of signalling by TGF family members has been implicated
in many human diseases, including cancer, fibrosis, autoimmune and
vascular diseases3.
The TGF family of ligands mediate their effects by binding
specific transmembrane type I and type II Ser/Thr kinase receptors.
The type I receptors act downstream of type II receptors and
determine the signalling specificity within the receptor complex.
Upon ligandinduced heteromeric complex formation, the type II
receptor transphosphorylates and activates the type I receptor,
which subsequently propagates the signal by phosphorylating
specific receptorregulated (R) SMAD transcription factors at the
two Cterminal Ser residues (FIG. 1). On activation, RSMADs form
heteromeric complexes with a related partner molecule, the CoSMAD
(SMAD4 in mammals), and accumulate in the nucleus where they
participate in the transcriptional control of target genes1,2.
Despite the large number and distinct functions of TGF family
members (33 in mammals), there is an enormous convergence in
signalling to only five type II receptors, seven type I receptors
(also described as activin receptorlike kinases (ALKs)) and two
main SMAD intracellular pathways1,2 (FIG. 1). So, how are
signalling specificity and diversity generated, especially as the
signalling receptors and SMADs are broadly coexpressed?
Extracellular
regulation of TGF signalling by coreceptors (which control the
access of ligands to signalling receptors) and interplay between
TGF ligands and extracellular molecules that regulate the activity
of these ligands could be important mechanisms that control
signalling specificity. Consistent with these potential mechanisms,
all three TGF isoforms (TGF1, TGF2 and TGF3 in mammals) are
secreted in latent forms that need to be activated before they can
bind to signalling receptors4. Furthermore, the biological
activities of TGF family members are regulated by specific secreted
inhibitors (see below) that sequester ligands and block binding to
receptors.
One of the compelling themes to emerge in recent years is the
crucial role of elastic extracellular matrix (ECM) proteins in
regulating TGF bioavailability in the vascular system5. Defects in
the ECM, which were initially thought to affect the physical
properties of vessel walls and thereby compromise the vasculature,
are now linked to enhanced TGFSMAD signalling. In this Review, we
focus on recent insights into the mechanisms that regulate
extracellular TGF activity in an appropriate spatio temporal manner
and how this contributes to its pivotal role in vascular
development and disease. For reviews addressing other related
aspects of TGF family signalling at the intracellular level and
dysregulated signalling in disease, see REFS 13, 68.
Secretion and ECM interactionSecretion and storage in the
ECM.The three TGF isoforms in mammals TGF1, TGF2 and TGF3 are
multifunctional and act in an autocrine, paracrine and sometimes
endocrine manner to regulate diverse
Pleiotropic Influencing multiple different traits.
R E V I E W S
NATURE REvIEwS | molecular cell biology vOLUME 8 | NOvEMBER 2007
| 857 2007 Nature Publishing Group
-
Nature Reviews | Molecular Cell Biology
EndoglinTGF
TGFBR2
ALK5 ALK1,2,3,6
Sol-Endo Sol-EndoBetaglycan
BMPR2SMAD2/3 SMAD1/5/8
SMAD4
P P
PP
PP
PP
SMAD7 SMAD6
Geneexpression
PP Gene
expression
Plasma membrane
Nucleus
BMP
Convertase family of endoproteases A group of enzymes that make
an internal cut in a polypeptide chain to convert it from an
inactive to an active form.
FibrillinAn extracellular matrix glycoprotein that is a
structural component of microfibrils.
MicrofibrilA fibre component (10 nm in diameter) of the
extracellular matrix that is essential for the integrity of elastic
fibres, which are particularly abundant in the aorta.
developmental processes and maintain tissue homeostasis in the
adult. The bioavailability of active TGF is regulated at multiple
levels, including secretion and interaction with ECM components,
and each step in the activation pathway is under tight control
(FIG. 2). TGFs are synthesized as precursor proteins that are
proteolytically processed. The signal peptide is removed from the
preproTGF during transit through the rough endoplasmic reticulum
and, following dimerization, another cleavage occurs by the
convertase family of endoproteases9. These proteases cleave the
precursor into the Cterminal mature peptide and the Nterminal
precursor remnant (also known as latency associated peptide (LAP))
within the secretory vesicles or in the extracellular space10.
Control and/or localization of convertase activity may represent an
important level of regulation of TGF ligands. After cleavage,
mature TGF
and LAP remain attached via noncovalent bonds to form the small
latent complex (SLC). LAP shields the receptorinteracting epitopes
in the mature protein and this keeps TGF in its latent form4.
The SLC can covalently attach to the large latent TGFbinding
protein (LTBP) to form the large latent complex (LLC)11,12 (FIG.
2). Most cell types secrete TGF as part of LLC, although some cells
(such as the bone cell line UMR106) secrete SLC13. Four different
LTPBs have been identified, of which LTBP1, LTBP3 and, to a lesser
degree, LTBP4 covalently bind to LAPs of all three TGF isoforms14.
LTBPs contain multiple epidermalgrowth factor (EGF)like repeats and
Cysrich domains that are also found in fibrillins, which are
extracellular proteins that are required for elastic fibre
formation. The Cterminal region of LTBP1 binds to the Nterminal
region of fibrillin1, linking LLC to elastic microfibrils15. After
secretion, LLC binds to the ECM via the Nterminal domain of LTBP
and this interaction is further supported by covalent
transglutaminaseinduced crosslinks16. Antibodies to LTBP1 and
inhibitors of transglutaminase activity inhibit the activation of
latent TGF, which demonstrates that localization of LTBP to the ECM
is required for effective TGF activation16,17.
LTBPs might have multiple functions: they target the latent TGF
complex to specific sites, including structural components within
the elastic fibres, where they may be stored for later use. This
targeting is determined by binding to ECM components, such as
fibronectin, via the variable Nterminal regions of LTBPs. In
addition, LTPBs stabilize latent TGF complexes and regulate their
activation at the cell surface (discussed below)18. Important
evidence for these roles comes from mouse models that are either
deficient in Ltbp3 or have a hypomorphic mutation in Ltbp4. These
mice are viable but have multiple phenotypes that are related to
decreased TGF signalling and defects in elastic fibres19,20 (TABLES
1,2). These results indicate an important role for LTBPs in
connective tissue deposition and as a local regulator of TGF
availability.
Release from microfibrils and ECM.In order for latent TGF to
become activated and function at adjacent and neighbouring cells,
the LLC must be liberated from microfibrils and ECM (FIG. 2). A
recent study revealed that LLC release can be initiated with the
displacement of LTBP bound to fibrillin1 (REF. 21). Degradation of
microfibrils by inflammatory proteolytic enzymes (such as elastase)
releases fragments of fibrillin1, including an internal fibrillin
fragment that efficiently binds to the Nterminal region of
fibrillin1 and displaces LTBP. This releases LLC from microfibrils
and contributes to local TGF activation21.
Several proteases, including plasmin, mastcell chymase and
thrombin, release LLC from the ECM4. Cleavage of LTBP1 occurs in a
sensitive hinge region such that the Nterminal fragment remains
bound to ECM, but the remainder of LLC is released (FIG. 2). In a
recent study, BMP1like matrix metalloproteases (MMPs) were shown to
cleave LTBP1 at two specific sites in the hinge region to release
LLC22, and facilitate the subsequent MMPdependent LAP cleavage. In
the absence of BMP1,
Figure 1 | Signal transduction by TgF family members. Canonical
signalling by transforminggrowth factor (TGF) superfamily members
can be divided into two main intracellular pathways according to
the SMAD mediators: either SMAD2/3 or SMAD1/5/8. Members of the TGF
family bind to specific Ser/Thr kinase type II and type I
receptors; in most cells, TGF signals via TGFBR2 and ALK5 (also
known as TGF receptor1; TGFBR1), and bone morphogenetic proteins
(BMPs) signal via the BMP type II receptor (BMPR2) and ALK1, 2, 3
and 6. The accessory receptors betaglycan and endoglin can modulate
signalling via the type II and type I receptors. Betaglycan
enhances TGF2 binding to TGF receptors, whereas endoglin may
perform a similar function for selected TGF family members and
their receptors. Soluble endoglin (SolEndo) is thought to sequester
ligand and thereby inhibits receptor binding95; however, the exact
mechanism through which this occurs is not known as endoglin
requires TGFBR2 for TGF binding113. Activated type I receptors
induce the phosphorylation of specific receptor regulated (R)
SMADs, which are the intracellular effectors of TGF family members.
In most cell types, TGF induces SMAD2/3 phosphorylation and BMPs
induce SMAD1/5/8 phosphorylation. Activated RSMADs form heteromeric
complexes with SMAD4 that accumulate in the nucleus, where they
regulate the expression of target genes such as SERPINE1
(plasminogen activator inhibitor) and ID1 (inhibitor of DNA
binding1) in cooperation with transcription factors, coactivators
and corepressors1,2. Inhibitory SMADs, such as SMAD6 and SMAD7, can
antagonize TGF signalling by inhibiting the activation of
RSMADs.
R E V I E W S
858 | NOvEMBER 2007 | vOLUME 8 www.nature.com/reviews/molcellbio
2007 Nature Publishing Group
-
Nature Reviews | Molecular Cell Biology
N
N
C
C
C
FurinElastase
Fibrillin-1
Emilin
LTBP
N
N
1 Pre-pro-TGF 5
6
7
8
2 Pro-TGF
3 SLC
4 LLC
LLC boundto ECMand elasticmicrofibrils
ECM(fibronectin)
Hingeregion
Matureprotein
LAP
a Synthesis and secretion b Activation and receptor binding
BMP1
MMP2
(and otherfibrillin fragments)
Cytoskeleton
Plasmamembrane
Plasmamembrane
LAPremnants
TGFBR2
Intracellular signaling
ALK5
Integrin
cells have increased fibrillar structures that contain LTBP1 and
show reduced TGF1 activity22. Consistent with this interaction,
impairment in TGF signalling or deficiency in BMP1 result in
similar defects in the frontal skull bones of mice23.
Activation and receptor binding. In order for TGF to bind to its
cognate cellsurface receptors, the mature peptide needs to be
released from LAP (FIG. 2). The mechanism by which latent TGF is
converted into active TGF varies according to cell type and
context, but all activating mechanisms directly target LAP4.
Invitro, physical conditions such as exposure to extremes of pH or
high temperature denature LAP but not mature TGF4. Invivo, it is
thought that binding of the multifunctional secreted
thrombospondin1 (THBS1) to LAP disrupts the noncovalent
interactions between LAP and mature TGF24. Activation can also
occur by proteases that cleave LAP to release bioactive TGF. These
processes couple matrix turnover to the production of a strong
inducer of ECM accumulation and may be crucial in maintaining a
balance of matrix components.
Another important activation mechanism for latent TGF is through
binding of v6 and v8 integrins to the RGD sequence in LAP25. The
mechanism is unclear, but interaction with the RGD domain of LAP
may induce a conformational change that leads to liberation or
exposure of TGF. Knockin mice, in which the RGD sequence in TGF1LAP
was replaced by a nonfunctional ArgGlyGlu sequence, recapitulate
all the main features of Tgfb1null mice, pointing to the importance
of RGD in TGF1 activation26. It should be noted that the LAP of
TGF2 does not contain an RGD sequence and may be activated by
another mechanism that remains to be determined25. Activation of
latent TGF by v8 integrins involves the membrane type I (MT1)
matrix metalloprotease (MMP), and is therefore sensitive to MMP
inhibitors; however, it is currently unclear whether it requires
LTBP27. By contrast, v6mediated activation is resistant to MMP
inhibitors and requires a direct interaction of fibronectin with
LTBP1 that targets LLC to the ECM28. As a result, latent TGF is
inefficiently activated in cells that lack fibronectin or its
receptor, 51 (REF. 28).
In line with these roles, inactivation of some of the murine
genes that encode putative activators leads to defects that are
reminiscent of mice deficient in TGF signalling components (TABLES
1,2). For example, Thbs1deficient mice have a similar, albeit
milder, phenotype to the Tgfb1 knockout24. Also, mice that lack v
or 8 integrins have defects in vascular and palatal development
that phenocopy the defects of Tgfb1 and Tgfb3knockout mice,
respectively29,30. These phenotypes support the idea that
extracellular activation of TGF is a major controlling step before
receptor binding invivo.
Cell-surface receptors. Once the active TGF family member is
released from the ECM, it signals via specific complexes of type I
and type II Ser/Thr kinase receptors (FIG. 1). The type I and type
II receptors are structurally similar with small Cysrich
extracellular domains, single transmembranespanning regions and
intracellular parts that mainly comprise kinase domains. TGF
signals via TGF type II receptor (TGFBR2) and TGFBR1 (also known as
ALK5; a type I receptor). Other combinations
Figure 2 | regulation of TgF bioavailability. Schematic model of
synthesis, secretion and matrix deposition of transforming growth
factor (TGF) (a) and activation and TGF receptor binding (b). TGF
is synthesized as a preproprotein, which undergoes proteolytic
processing in the rough endoplasmic reticulum (1). Two monomers of
TGF dimerize through disulphide bridges (2). The proTGF dimer is
then cleaved by furin convertase to yield the small latent TGF
complex (SLC), in which the latencyassociated peptide (LAP; orange)
and the mature peptide (red) are connected by noncovalent bonds
(3). This processing step is inhibited by emilin1. The large latent
TGF complex (LLC) is formed by covalent attachment of the large
latent TGF binding protein (LTBP, shown in blue; 4). The Nterminal
and hinge region of LTBP interact with extracellular matrix (ECM)
components such as fibronectin; this interaction can be covalent
owing to crosslinking by transglutaminase. The Cterminal region of
LTBP (blue) interacts noncovalently with the Nterminal region of
fibrillin1 (green; 4). As part of TGF activation and receptor
binding (b), an internal fragment of fibrillin1 (indicated in
purple in 5) can be released by proteolysis (mediated by elastases
at sites indicated by black arrowheads in 5) and interacts with
Nterminal region of fibrillin1 to displace LTBP and release LLC
(6). The LLC can be targeted to the cell surface by binding to
integrins via RGD sequences (blue regions) in LAP (6). Bone
morphogenetic protein1 (BMP1) can cleave two sites in the hinge
region of LTBP (arrowheads in 6), which results in the release of
LLC (7). Matrix metalloprotease2 (MMP2) (and other proteases) can
cleave LAP (black arrowheads in 7) to release the mature TGF (red).
Mature TGF can then bind to its cognate receptors, TGFBR2 and ALK5
(8).
R E V I E W S
NATURE REvIEwS | molecular cell biology vOLUME 8 | NOvEMBER 2007
| 859 2007 Nature Publishing Group
-
Matrix metalloprotease One of a family of structurally related
extracellular Ca2+-dependent zinc-containing proteases involved in
tissue remodelling and ECM degradation.
RGD sequence An amino acid sequence (Arg-Gly-Asp) found in
extracellular matrix proteins that directly binds to integrins.
of type I and type II receptors respond to other TGF family
ligands, for example, the BMP type II receptor (BMPR2) responds to
BMP ligands in combination with ALK1, 2, 3 or 6 type I
receptors.
TGF receptors are expressed on endothelial cells (ECs) and
vascular smooth muscle cells (vSMCs), but also on many other cell
types31,32. This, together with the highly contextdependent
responses, generates a multifaceted role for TGF family members in
vascular biology33. In ECs, TGF can transduce signals via ALK1 in
addition to ALK5, and is thought to contribute to the bifunctional
effects of TGF on angiogenesis34. ALK5 is required for TGFALK1
activation, whereas ALK1 inhibits intracellular ALK5SMAD
signalling. The differential activation of these two distinct type
I receptor pathways by TGF provides the ECs with an intricate
mechanism to precisely regulate, and even switch between,
TGFinduced biological responses. For example, TGFALK1 activation
leads to stimulation of EC proliferation and migration, whereas
TGFALK5 activation inhibits these responses34.
Type III (or auxiliary) receptors such as betaglycan and
endoglin (ENG) present another mechanism through which signalling
specificity is regulated. TGF2, which has a low intrinsic affinity
for TGFBR2, requires betaglycan for efficient signalling35. In
addition, betaglycan can be shed by cells upon proteolysis in the
juxtamembrane region to sequester mature TGF and inhibit TGF
signalling36. Proliferating ECs express endoglin, which is required
for efficient TGFALK1 signalling37,38; endoglin may also be
proteolytically shed (see below). Thus, different combinations of
ligands, receptors and coreceptors, either on the membrane or shed
from it, may result in complex patterns of TGF activity.
TGF regulates vascular developmentNew blood vessels develop by a
combination of vasculogenesis and angiogenesis39 (BOX 1). These
processes are regulated by cytokines and growth factors in a highly
orchestrated manner during embryogenesis. The vital importance of
TGF signalling in vascular development was recognized following the
identification of mutations
Table 1 | Human syndromes and animal models associated with
inactivation or misexpression of TGF signalling components*gene
Human syndrome animal models refs
Extracellular regulation of TGF signallingBMP1 Unknown KO:
reduced skull ossification, abnormal collagen fibrils in amnion,
die at birth. 23
EMILIN1 Hypertension KO: reduced arterial diameter, increased
vascular resistance, emphysema, increased TGF signalling in
vascular wall.
93
FBN1 (fibrillin1) MFS1 KO: ruptured aortic aneurysm, impaired
pulmonary function, die at birth. 83, 84
Het for missense mutation (C1039G): aortic aneurysm,
emphysema.
FBN2 (fibrillin2) Contractural arachnodactyly
KO: bilateral syndactyly. 109
FN1 (fibronectin1) EhlersDanlos syndrome, type X
KO: deformed heart and embryonic vessels, defective
extraembryonic vasculature.
114
EFEMP2 (fibulin4) Cutis laxa KO: perinatal lethality, arterial
narrowing, aortic rupture. 90, 91
Hypomorphic: tortuous aorta, aortic aneurysm, increased TGF
activation.FBLN5 (fibulin5, DANCE) Cutis laxa KO: tortuosity and
elongation of the aorta, loose skin. 88, 89
LTBP1 Unknown KO: persistent truncus arteriosus. 124
LTBP3 Unknown KO: craniofacial malformations, osteosclerosis and
osteoarthritis. 19
LTBP4 Unknown Hypomorphic: emphysema, cardiomyopathy and
colorectal cancer. 20
THBS1 (thrombospondin1) Unknown KO: pneumonia, alveolar
haemorrhage, vSMC hyperplasia. 24, 115
ITGB8 (integrin, 8) Unknown KO: embryonic lethal with vascular
defects and/or perinatal lethality and cerebral haemorrhage.
29
ITGAV (integrin, V) Unknown KO: embryonic lethal placental
defects or perinatal lethal with intracerebral and intestinal
haemorrhages and cleft palate.
30
ITGB6 (integrin, 6) Unknown KO: inflammation of lungs and skin.
116TGFB1 CamuratiEngelmann
disease||KO: embryonic lethal with vascular defects or postnatal
lethality from autoimmune disease (phenotype is
modifierdependent).
40, 44
KI of RGD/E mutation: recapitulates KO phenotypes. 26
TGFB2 Unknown KO: aortic arch defects, cardiac septal defects,
perinatal lethality. 40
TGFB3 Unknown KO: cleft palate, delayed lung maturation, die
shortly after birth. 40
*Continued in Table 2. Animal models are murine, unless
otherwise indicated, and may be heterozygous (Het) or homozygous
for the targeted allele (knockout (KO) for null allele; knockin
(KI) for hypomorphic/mutant allele) or transgenic. Cutis laxa maps
to several genes and is characterized by pendulous inelastic skin,
which is sometimes combined with aortic aneurysm and tortuous
arteries; however, the specific genes associated with the
cardiovascular features are not yet clear. ||CamuratiEngelmann
disease is caused by an activating mutation in TGF1. BMP1, bone
morphogenetic protein1; LTBP, large latent TGFbinding protein;
MFS1, Marfan syndrome type 1; TGF, transforming growth factor;
vSMC, vascular smooth muscle cell.
R E V I E W S
860 | NOvEMBER 2007 | vOLUME 8 www.nature.com/reviews/molcellbio
2007 Nature Publishing Group
-
in TGF receptor genes in familial vascular pathologies
(discussed below). Furthermore, studies in mouse models showed
that, in the absence of key TGF receptors, angiogenesis stalls in
the yolk sac at an early stage with fatal consequences40 (FIG.
3).
Role of TGF ligands in angiogenesis.The requirement for TGF
receptors for angiogenesis raises the question of which TGF
ligand(s) regulates new vessel formation. Of the three mammalian
TGF isoforms, TGF1 is localized to ECs during embryogenesis41,
suggesting that it is the most likely of the three to be involved
in angiogenesis. TGF2 and TGF3 seem to be less important than TGF1
in angiogenesis because mice that lack either of these ligands show
relatively normal angiogenesis42,43. However, there may be
redundancy between all three TGF ligands, and multiple or
conditional knockout mice are required to investigate this
possibility. Interestingly, loss of Tgfb1 in a null mouse can
result in different phenotypes depending on the genetic background.
For example, in C57BL/6 mice, angiogenesis is abnormal and embryos
die at embryonic day (E)10.5, whereas in the NIH background the
cardiovascular
system tends to develop normally but neonates die from an
uncontrolled autoimmune assault40. The dramatic influence of
genetic background results from three unlinked genetic modifiers44.
These genes and their functions are unknown, but one possibility is
that one or more may regulate TGF activation.
Invitro studies have demonstrated the importance of regulating
the availability of active TGF. Low extracellular TGF1
concentrations promote the cell proliferation and migration that is
associated with the active proliferation of new vessels in
angiogenesis37. By contrast, high levels of extracellular TGF1 lead
to cytostasis and synthesis of ECM proteins that are associated
with mature or differentiating vessels. TGF family members can also
act in a paracrine manner by stimulating the production of
proangiogenic cytokines, such as vascular endothelial growth factor
(vEGF), TGF and monocyte chemoattractant protein1 (MCP1)45,46,47.
Moreover, TGFs can affect the function of other factors; for
example, TGF converts the prosurvival function of vEGF into an
apoptotic factor for ECs48. The resultant interplay results in a
refined and cellcontextspecific response to TGF stimulation.
Table 2 | Human syndromes and animal models associated with
inactivation or misexpression of TGF signalling components*gene
Human syndrome animal models refs
TGF auxiliary receptorsENG (endoglin) HHT1 Het: vascular lesions
similar to HHT. 40, 59
KO: embryonic lethal, reduced vSMC differentiation, heart
defects.
Elevated soluble endoglin Preeclampsia None 95
TGFBR3 (betaglycan) Unknown KO: poorly formed cardiac septa,
incomplete compaction of ventricular walls. 117
TGF signalling receptorsACVRL1 (ALK1, vbg) HHT2 (rarely PAH)
Het: vascular lesions similar to HHT. 60, 118
KO: embryonic lethal, reduced vSMC differentiation, dilated
vessels, AVMs. 40, 64
Zebrafish KO: dilated vessels, AVMs. 51
TGFBR2 MFS2, LDS KO: embryonic lethal, vascular defects. 40,
123
TGFBR1 (ALK5) LDS KO: embryonic lethal, angiogenesis defects.
119
BMPR2 PAH KO: preangiogenesis lethality. 120, 121
Het: mild pulmonary hypertension.
Transgenic inducible BMPR2mutant allele: pulmonary hypertension.
74
Intracellular TGF signalling moleculesSMAD1 Unknown KO:
preangiogenesis lethality, defects in chorionallantoic circulation.
40
SMAD2 Unknown KO: preangiogenesis lethality. 40
SMAD3 Unknown KO: viable. Impaired immunity, colon cancer.
40
SMAD4 Juvenile polyposis +/ HHT
KO: preangiogenesis lethality. 40, 122
Het: polyposis of the glandular stomach and duodenum.
SMAD5 Unknown KO: embryonic lethal, angiogenesis defects. 40
SMAD6 Unknown KO: cardiac defects, aortic ossification, elevated
blood pressure. 40
MAP3K7 (TAK1) Unknown KO: embryonic lethal, reduced numbers of
vSMCs, dilated vessels, AVMs. 50
*Continued from Table 1. Animal models are murine, unless
otherwise indicated, and may be heterozygous (Het) or homozygous
for the targeted allele (knockout (KO) for null allele; knockin
(KI) for hypomorphic/mutant allele) or transgenic. ALK1, activin
receptorlike kinase1; AVMs, arteriovenous malformations; BMPR2,
bone morphogenetic protein receptor2; HHT, hereditary haemorrhagic
telangiectasia; LDS, LoeysDietz syndrome; MFS2, Marfan syndrome
type 2; PAH, pulmonary arterial hypertension; TGF, transforming
growth factor; TAK1, TGF activated kinase1; TGFBR, TGF receptor;
vSMC, vascular smooth muscle cell.
R E V I E W S
NATURE REvIEwS | molecular cell biology vOLUME 8 | NOvEMBER 2007
| 861 2007 Nature Publishing Group
-
Intracellular TGF signalling in angiogenesis.Because canonical
TGF signalling occurs through SMAD activation, one might expect
similarities in the phenotypes of the SMADknockout and TGF
receptorknockout mice. In fact, SMAD1, SMAD2 and SMAD4 are required
during early embryonic development before angiogenesis presumably
because they are central to multiple TGF family ligands40. SMAD3
regulates mucosal immunity and does not appear to be required for
cardiovascular development49. Similarities in the phenotypes of the
Smad5null mouse and TGF receptor knockouts point to a
proangiogenesis role for SMAD5. The most striking similarities have
been observed in TGF activated kinase1 (TAK1) and TGF receptor
mutant mouse phenotypes50. TAK1 (also known as MAP3K7) is activated
by TGF and inflammatory cytokines and is central to several
signalling pathways that control immune and stress responses.
Mutations in acvrl1(also known as vbg, the zebrafish homologue of
ALK1) result in the formation of dilated cranial vessels that
contain increased numbers of ECs51, whereas morpholino knockdown
experiments indicate a synergistic interaction between the
zebrafish homologues of TAK1 and ALK1 (REF. 50). This invivo data
from both mouse and zebrafish models suggest that TGF signalling
through TAK1 is important in the regulation of vascular
development.
TGF and blood-vessel myogenesis.One of the best understood roles
of TGF signalling in angiogenesis is that of promoting vessel
muscularization. vSMCs (arteries and veins) or pericytes
(capillaries) cover new vessels and are required for both vessel
stability and function. Muscularization is achieved when ECs
promote paracrine TGF1 signalling to neighbouring mesenchymal cells
to promote SMC or pericyte differentiation. Establishment of
cellcell contacts, at the level of gapjunction communication
between endothelial and mesenchymal cells, is required for TGF
activation and SMC differentiation52. TGF signalling between ECs
and vSMCs is also dependent on endoglin53. Ultimately, TGF
activation stimulates the differentiation of mesenchymal cells
to SMCs through the combinatorial activation of several SMC
genes such as SMactin (ACTA2) and transgelin (TAGLN; also known as
SM22)54. In line with their vascular myogenesis role, specific loss
of the TGF receptors, TGFBR2 or ALK5, in vSMCs results in vascular
defects and embryonic lethality55.
Importantly, TGF signalling in vascular development is not
confined to SMCs. Recent analysis of mice with ECspecific loss of
either Tgfbr2 or Tgfbr1(which encodes ALK5) has clearly shown that
both TGF receptors also have essential angiogenesisrelated roles in
ECs56,53. The role of TGF receptors in ECs may be crucial in
development earlier than in vSMCs, as mutants with ECspecific loss
of Tgfbr2 die from vascular defects 2 days sooner (at E10.5) than
those with vSMCspecific loss (E12.5)55. However, a recent report
suggests that Tgfbr1 is not expressed in ECs57. This discrepancy
may be explained by the limitations of knockin lacZ reporters for
monitoring endogenous protein expression, and further work is
required to determine the relative contributions of TGF signalling
receptors to EC function invivo. Nonetheless, the central
importance of properly regulated TGF signalling in the vasculature
is clearly demonstrated by the growing number of familial
cardiovascular disorders that are associated with mutations
affecting TGF family signalling.
Vascular disease and TGF signallingOver recent years, numerous
familial vascular diseases have been mapped to receptors of the TGF
family and, in many cases, further understanding has been reached
through studies of mouse models58. A common theme is that TGF
signalling is important in bloodvessel morphogenesis and stability.
Recent data also provide insights into the role of TGF signalling
in hypertension and preeclampsia. Furthermore, recognition of the
importance of TGF signalling in angiogenesis has opened up new
possibilities for targeting cancer by modulating this pathway in
the tumour vasculature.
Hereditary haemorrhagic telangiectasia (HHT). The vascular
disorder hereditary haemorrhagic telangiectasia type 1 (HHT1)
results from mutations in ENG that lead tohaploinsufficiency of
endoglin59. The closely related disorder HHT type 2 (HHT2) is
caused by loss of function or dominant negative mutations in ACVRL1
(which encodes ALK1)60,61. Endoglin is expressed in ECs of all
developing blood vessels, and ALK1 is primarily expressed in
arterial ECs57,62, which points to defects in ECs as the primary
cause of HHT.
The main clinical features of HHT are bleeding from small
vascular lesions (telangiectases) in the mucocutaneous tissues, and
the presence of arteriovenous malformations (AvMs) in the lung,
liver and/or cerebral vasculature. The typical onset of symptoms
(usually major bleeding from nasal telangiectases) is during
puberty, and AvMs may be present that become larger and symptomatic
as patients age. Several possible mechanisms have been proposed to
explain the development of AvMs, including loss of vSMCs,
dysregulated vascular tone or apoptosis of capillary ECs (FIG.
4).
Box 1 | Vasculogenesis and angiogenesis
Vasculogenesis is the earliest step in the development of new
blood vessels and involves the differentiation of angioblasts into
endothelial cells (ECs) and their assembly into a primary vascular
plexus. Angiogenesis is the process during which blood vessels
develop from existing capillaries by sprouting, pruning and/or
splitting (intussusception). These processes are driven by a
complex interaction of growth factors (for example, vascular
endothelial growth factor (VEGF), fibroblast growth factor (FGF)
and transforming growth factor (TGF)) and their receptors. Live
imaging of angiogenesis in zebrafish mutants suggests that the
layout of the primitive vascular network is genetically determined.
The directionality of new vessel growth is driven by leading
endothelial tip cells in response to guidance molecules, whereas
the lagging ECs form lumens by intercellular fusion of endothelial
vacuoles112. During maturation, pericytes or vascular smooth muscle
cells (vSMCs) are recruited to capillaries and larger vessels,
respectively. Plateletderived growth factor (PDGF) signalling is
important for the initial recruitment of mesenchymal cells that
differentiate to vSMCs in response to TGF signalling. As the animal
grows, the vascular network continues to remodel by pruning and
branching in response to a combination of growth factors, hypoxic
triggers and blood flow to form a complex treelike structure of
arteries, veins and capillaries.
MorpholinoChemically synthesized oligonucleotide analogues used
to knock down gene expression by specifically binding to target
transcripts to inhibit RNA splicing or translation.
Vessel muscularization The development of smooth muscle cells
around a vessel to support and stabilize it.
PericyteA smooth muscle-like cell that is intimately associated
with endothelial cells of small blood vessels.
Mesenchymal cellA member of a heterogeneous multipotent cell
population that arises mainly from embryonic mesoderm.
Hypertension Elevated blood pressure.
Arteriovenous malformation(AVM). Abnormal communication between
an artery and a vein producing dilated vessels.
Intussusceptive angiogenesis The process of blood vessel growth
by splitting the wall of an existing blood vessel extends into the
lumen to split a single vessel in two.
R E V I E W S
862 | NOvEMBER 2007 | vOLUME 8 www.nature.com/reviews/molcellbio
2007 Nature Publishing Group
-
Nature Reviews | Molecular Cell Biology
Angioblast cells
Primitive vascular plexus
Intussusceptiveangiogenesis
Vasculogenesis
Angiogenesis
Tip cellVacuole fusion and lumen
formation
Sproutingangiogenesis
Progression blocked in yolk sacin the absence ofTGF1, TGFBR2,
endoglin,ALK1, ALK5 or SMAD5
Mesenchymalcell
Gap-junction formation,TGF1 activation,SMC differentiation
andmutual inhibition ofproliferation
Endothelial cell
PDGF
PDGF
Sprouting angiogenesis The process by which endothelial cells
migrate and proliferate into the surrounding matrix to form new
vessel branches in response to an angiogenic stimulus.
To investigate abnormalities in ECs derived from patients with
HHT, cultured circulating endothelial progenitor cells were used to
generate blood outgrowth ECs. These had a disorganized actin
cytoskeleton that may contribute to vascular fragility in HHT63.
Further investigations in mouse models of HHT suggest that there
are reduced levels of TGF signalling from ECs to vSMCs or
pericytes, which results in decreased SMC differentiation.
This defect generates weaker vessels that are prone to bleeding,
a main characteristic of HHT53. In addition, development of AvMs
early in embryogenesis may be due to loss of arteriovenous
identity. Arterialspecific and venousspecific signalling molecules
such as ephrin B2 (EFNB2) and EPHB4, respectively, are thought to
ensure that these vessel types remain distinct and separate during
development. AvMs frequently develop in Acvrl1null mouse embryos,
although to a lesser extent in Engnull mice, and may be due to
downregulation of EFNB2 in the Acvrl1/ mice64,65.
There may also be systemic defects in TGF signalling in HHT1, as
reduced endoglin levels have been shown to cause decreased levels
of plasma TGF in both mouse models and patients with HHT1 (REF.
66). However, this remains to be clarified because raised TGF1
levels have also been reported67. The exact molecular changes
leading to HHT are not yet clear. New insights are anticipated
following the recent identification of BMP9 as a ligand for both
endoglin and ALK1 (REF. 68), and newly derived conditional knockout
Eng and Acvrl1 mouse models69 (S.P. Oh, unpublished
observations).
Pulmonary arterial hypertension (PAH).The devastating lung
disease pulmonary arterial hypertension (PAH) is characterized by
pathological changes that include hypertrophy (enlargement) of the
medial smooth muscle layers and intimal thickening in the
precapillary arterioles (FIG. 4). This leads to increased pulmonary
artery resistance that can ultimately result in rightsided heart
failure. In addition, plexiform lesions that comprise multiple
capillary channels and proliferating ECs may develop near occluded
regions. The average age of onset is in the third decade of life in
women and the fourth decade in men, but can occur at any age.
Familial PAH (fPAH) is associated with lossoffunction mutations
in the BMP receptor2 gene (BMPR2). Reduced levels of this receptor
in pulmonary artery SMCs and ECs in patients with PAH lead to
reduced SMAD1 activation and/or increased activation of the p38
mitogenactivated protein kinase (MAPK) that may contribute to vSMC
hyperproliferation70,71. It appears that BMPR2deficient cells may
respond abnormally to BMP ligands by redirecting signalling through
the type II activin responsive receptor ActRIIA (in association
with either ALK2 or ALK3)72. In addition, the long cytoplasmic tail
of the BMPR2 protein, which is truncated in many patients with
fPAH, is required for proper regulation of the cytoskeleton, and
misregulation may contribute to the aetiology of PAH73. Mouse
models of fPAH with inactivated Bmpr2 have yielded a lack of PAH
symptoms in normoxic conditions. However, postnatal expression of a
dominantnegative BMPR2mutation in vSMCs leads to pulmonary
hypertension and modest muscularization of distal arteries. These
data suggest that aberrant BMPR2 signalling in vSMCs is sufficient
to produce the pulmonary hypertensive phenotype and provides a
useful animal model for further investigation74.
The link between dysregulated BMP signalling and progression of
this devastating disease remains to be determined, but an important
clue may come from the
Figure 3 | TgF signalling in vasculogenesis and angiogenesis.
Vasculogenesis and two types of angiogenesis are shown:
intussusceptive and sprouting angiogenesis (BOX 1). Vasculogenesis
involves the differentiation of endothelial cells (ECs) from
precursor angioblast cells to form a primitive plexus of
capillaries, which remodel and grow by angiogenesis.
Intussusceptive angiogenesis involves the splitting and growing of
vessels in situ in a metabolically efficient manner, and is found,
for example, in the developing yolk sac and lung. Vessel splitting
occurs by the formation of translumen pillars (arrowheads) but the
molecular mechanisms are not well understood. In sprouting
angiogenesis, endothelial cells proliferate behind the tip cell of
a growing branch in response to cytokines such as vascular
endothelial growth factor (VEGF) and lumens can form by vacuole
fusion. Both forms of angiogenesis require the recruitment of
smooth muscle cells (SMCs) to stabilize the nascent vessels.
Neighbouring mesenchymal cells migrate towards the neovessel in
response to plateletderived growth factor (PDGF) and then
differentiate into vascular SMCs (vSMCs) in response to
transforming growth factor (TGF) signalling. The continued tight
intercellular association between ECs and vSMCs promotes sustained
TGF activation and mutual inhibition of cell proliferation. When
TGF signalling is defective, for example in TGFreceptor knockout
mice, smooth muscle differentiation fails to proceed and
angiogenesis stalls. ALK, activin receptorlike kinase; TGFBR2, TGF
type II receptor.
R E V I E W S
NATURE REvIEwS | molecular cell biology vOLUME 8 | NOvEMBER 2007
| 863 2007 Nature Publishing Group
-
Nature Reviews | Molecular Cell Biology
Terminal bronchiole and alveoli
Pre-capillary pulmonary arteriole
Pulmonary venule
Capillaries
ArterioleVenule
Site of partialocclusion
Endothelial cellSmooth muscle cellStromal cell
a b
c
d e
fact that patients with HHT who have ACVRL1mutations are
predisposed to the development of PAH75. As ALK1 is expressed in
pulmonary ECs, this finding suggests that aberrations in TGF family
signalling in the endothelium can also be the primary defect in
PAH. Recent identification of BMP9 as an ALK1 ligand suggests that
reduced ALK1 levels may lead to altered BMP9 signalling68, although
any link to PAH remains to be determined.
Enhanced TGF signalling and aortic aneurysms.Aortic aneurysms,
predominantly originating in the aortic root (FIG. 5), are
important features of Marfan syndrome (MFS) and a clinically
overlapping disorder that has been variously termed MFS type 2
(MFS2), familial
thoracic aneurysm disorder (AAT3) and LoeysDietz syndrome
(LDS)7678. MFS is primarily associated with mutations in
FBN1(fibrillin1), whereas LDS is associated with mutations in
TGFBR1or TGFBR2. Aortic aneurysms are insidious malformations with
a variable age of onset which, if left untreated, carry the risk of
aortic dissection, rupture and sudden death. The vulnerability of
the aortic root in these disorders may relate to the two different
developmental origins (neural crest and secondary heart field) of
vSMCs at the region of the aortic valve, which have different
responses to TGF signalling79. This results in two closely abutting
rings of different vSMC populations and it is possible that the
junction between them represents a susceptible site for aortic
dissection80 (FIG. 5).
Anastomosis A naturally occurring arteriovenous connection that
may be dynamically regulated and is particularly frequent in
thermoregulatory vascular beds.
Figure 4 | Vascular remodelling in PaH and HHT. a | Terminal
alveoli in pulmonary arterial hypertension (PAH) showing the
approximate site of occlusion in the precapillary arteriole. b |
Crosssection of a normal precapillary pulmonary arteriole, which
shows a typical vessel structure of a single layer of endothelial
cells surrounded by supporting smooth muscle cells. c |
Crosssection of a partially occluded precapillary arteriole in PAH
with proliferating vascular smooth muscle cells (vSMCs) and
endothelial cells (ECs). A plexiform lesion that comprises multiple
vascular channels is shown to the left of the occluded arteriole. d
| The normal capillary network that is present in most vascular
beds and interconnects an arteriole and a venule. e | Formation of
an arteriovenous malformation in hereditary haemorrhagic
telangiectasia (HHT), which may develop by one or more different
mechanisms: the loss of arterial and venous identity during
development would disrupt the normal separation of arteries and
veins leading to arteriovenous connections; abnormal vascular
remodelling and dilation following local inflammation or trauma may
fail to resolve; apoptosis of the capillary ECs in regions of
hypoxia would remove the natural capillary bed that separates
arteries and veins; or gradual dilation of a naturally occurring
anastomosis may occur as a result of loss of SMCs and/or loss of
vessel tone leading to capillary regression due to lack of blood
flow.
R E V I E W S
864 | NOvEMBER 2007 | vOLUME 8 www.nature.com/reviews/molcellbio
2007 Nature Publishing Group
-
Nature Reviews | Molecular Cell Biology
Tunica intimaTunica mediaTunica adventitia
a b c
d
LTBP
LAPTGF1
Microfibrils Few microfibrils orsecondary proteolysis
e
MFS
Normal elastin fibres Fragmented elastin fibres
PT
It was initially thought that aortic aneurysms in MFS were due
to structural defects in the aorta resulting from a failure to
stabilize elasticfibre structure when fibrillin1 was limiting. The
resultant elastin fragmentation in the aortic wall would make the
aorta susceptible to injury from haemodynamic forces. Recently, it
has become clear that an important function of fibrillin1 is to
control TGF bioavailability (FIG. 2). Reduced fibrillin1 may result
in incorrect LLC sequestration and excessive activation of TGF
signalling, a major contributory factor in the vascular pathology
of MFS81,82. Mice deficient in Fbn1,but not mice overexpressing
mutant Fbn1, develop MFS phenotypic manifestations83,84. This
suggests that a deficiency in microfibrils, as opposed to a
transdominant effect of mutant fibrillin1, is a major determinant
of MFS. It has also been proposed that the progressive proteolytic
damage that is characteristic of MFS may lead to the formation of
fibrillin fragments that further increase the bioavailability of
TGF21. In the case of LDS, when TGFBR2 levels are reduced owing to
pathologic mutation, compensatory mechanisms may lead to an
overshoot of TGF signalling in the aortic media. For example,
increased TGF expression and increased fibrosis have been observed
in transgenic mice that overexpress a dominant negative form of
Tgfbr2(REF. 85). Similar, but as yet unknown, mechanisms may
explain the upregulation of TGFSMAD signalling and fibrosis that is
seen in the aortae of patients with LDS, who carry heterozygous
inactivating mutations in either TGFBR2 or TGFBR1(REF. 77).
whatever the cause of enhanced TGF signalling, work using animal
models of MFS clearly showed that aortic aneurysms were prevented
by administration of neutralizing antibodies to TGF, confirming the
contribution of excess TGF signalling to the vascular pathology82.
In the absence of approved antiTGF clinical therapies, attention
turned to angiotensin, a potent vasoconstrictor that interacts with
angiotensin receptors on blood vessels and induces SMAD2/3
phosphorylation86. This change of focus was rewarded when the
angiotensin type I receptor antagonist, losartan (Cozaar; Merck),
showed a strikingly similar protective effect in the development of
aortic aneurysms in animal models of MFS82. However, the mechanism
by which losartan protects against aortic aneurysm and reduces
SMAD2 activation is not fully understood. One possibility is the
involvement of THBS1, an important target downstream of angiotensin
signalling87 that can activate latent TGF. whatever the mechanism,
losartan has rapidly moved to Phase III clinical trials in patients
with MFS (see the US National Marfan Foundation web site).
Taken together, the evidence indicates that misregulation of TGF
signalling owing to defects in extracellular proteins is centrally
important to the development of aortic aneurysms, and there is now
a realistic hope for patient therapies. This view has now replaced
the previous idea that aortic aneurysms were simply due to a
structural deficiency of the elastin matrix in the aorta. It is
notable in this context that mice lacking the extracellular protein
fibulin5 have fragmented elastin and tortuous elongated aortae, but
do not develop aortic
Figure 5 | TgF-associated defects in marfan and loeysDietz
syndromes. a | An aorta with two seams of vascular smooth muscle
cells (vSMCs) at the aortic root and pulmonary trunk (PT) that are
derived from different developmental origins cardiac
neuralcrestderived SMCs (dark pink), secondary heartfieldderived
SMCs (purple) and secondary heartfieldderived myocardial muscle
(blue). These areas may be sites that are vulnerable to dissection.
Heart and pulmonary arteries are shown as dotted lines. b | An
aortic root aneurysm that is typical of Marfan syndrome (MFS) and
LoeysDietz syndrome. c | Ascending aortic aneurysm. d | Descending
aortic aneurysm. e | A crosssection of aorta shows the three layers
(tunics) of the vessel: tunica adventitia (which contains collagen,
fibroblasts, nerves and capillaries), tunica media (which contains
SMCs, elastic fibres/microfibrils, collagen and proteoglycan) and
tunica intima (which contains endothelial cells and basal lamina).
Fragmented elastin fibres in the aorta are seen in MFS and are
associated with release of the large latent transforming growth
factor (TGF) complex and active TGF1. LAP, latencyassociated
peptide; LTBP, large latent TGF binding protein.
R E V I E W S
NATURE REvIEwS | molecular cell biology vOLUME 8 | NOvEMBER 2007
| 865 2007 Nature Publishing Group
-
Placental syncytiotrophoblast cellA multinucleated cell found at
the boundary of the fetal and maternal layers of the placenta.
aneurysms or dissections88,89. It may be that the aortae of
these mice have normal TGF activity and, hence, do not progress to
aneurismal changes. By contrast, mutant mice that have lost the
related Efemp2 gene (which encodes fibulin4) show a striking aortic
rupture phenotype and perinatal lethality90. This defect precluded
analysis of this gene in mature vasculature, but mice that are
homozygous for a hypomorphic allele with reduced fibulin4
expression survive to adulthood with a milder phenotype. These mice
have aortic dilatation, aneurysms and aortic valve dysfunction
associated with abnormal TGF signalling in the aorta91. These
different mouse models offer the opportunity to unravel the complex
interaction between aortic integrity and ECM regulation of TGF
activity.
TGF and hypertension.A link between increased levels of
circulating TGF and hypertension has been known for some time92,
but new findings are throwing light on the nature of this
interaction. Emilin1, an ECM protein that is associated with the
microfibrils of the elastic matrix in the aortic media, regulates
TGF availability and arterial diameter93. By interacting with
unprocessed TGF, emilin1 protects it from proteolytic processing by
the endoprotease furin convertase (FIG. 2). Loss of emilin1
therefore results in the increased conversion of proTGF to the
mature form and a subsequent increase in TGF signalling. This leads
to a reduction in the arterial lumen diameter with a resultant
increase in vascular resistance and hypertension93. A possible
mechanism is that excessive levels of active TGF causes premature
cytostasis of the vSMCs, and subsequently restricts vessel size.
Alternatively, reduced arterial diameter may be a secondary
consequence of vascular remodelling in these mice94. This phenotype
can be rescued genetically by crossing in a single Tgfb1null allele
to reduce TGF levels.
A related disease that involves raised blood pressure is
preeclampsia, a serious disease of late pregnancy that threatens
the health of both mother and baby. Patients with preeclampsia have
increased circulating levels of a soluble form of vEGF receptor1
(vEGFR1; also known as sFLT1), which is thought to function as a
sink for vEGF ligands and reduce the access of vEGF ligands to
vEGFRs on ECs. Recently, elevated levels of soluble endoglin were
also found to cooperate with soluble vEGFR1 in the pathogenesis of
preeclampsia95. Soluble endoglin is probably formed by proteolytic
cleavage of fulllength endoglin, and is possibly derived from
placental syncytiotrophoblast cells. This soluble form may function
in an analogous way to soluble vEGFR1 by binding circulating TGF,
which therefore inhibits TGFBR2 and ALK5dependent signalling95.
This is thought to reduce the level of nitric oxide synthase3
(NOS3, eNOS) activation and attenuate vasorelaxation responses,
leading to preeclampsia.
The mechanisms that regulate the production of soluble endoglin
and a detailed understanding of its molecular role in patients with
preeclampsia remain to be determined. A recent report showed that
haem oxygenase1, an antiinflammatory enzyme required
for successful pregnancy, inhibits the release of soluble vEGFR1
and soluble endoglin, a role that might be used in the future
treatment of preeclampsia96. However, circulating levels of soluble
endoglin may already be a useful diagnostic marker to prioritize
patients for treatment before the onset of symptoms97.
Once again, the role of TGF in the regulation of hypertension
appears to be contradictory. Increased levels of activated TGF in
emilin1-null mice result in hypertension, whereas reduced TGF
signalling in ECs may contribute to increased blood pressure in
preeclampsia93,95. However, the underlying mechanisms are
different. There is no detectable change in vasoregulatory
responses in arteries in emilin1-null mice instead, there is a
reduction in vessel size. By contrast, in preeclampsia there is a
transient defect in vasoregulatory mechanisms, but the blood
pressure returns to normal levels once the fetus is born. In
addition, there may be other ligands involved in the preeclampsia
model, as endoglin binds other ligands of the TGF family, including
BMP9 (REF. 68).
TGF and tumour angiogenesis.The importance of understanding the
regulation of angiogenesis by TGF family members is crucial because
many of these processes are recapitulated in a disorganized fashion
in neoplastic disease. Several smallmolecule TGF receptor kinase
inhibitors have been developed for cancer treatment98. However,
there may also be unwanted effects of TGF receptor inhibition, such
as the possibility that the rate of progression of some cancers
will increase when the tumoursuppressive effects of TGF are
inhibited99. In light of this possibility, it is promising that
there appears to be no predisposition to cancer in patients with
LDS who have reduced TGF receptor levels100. Rather, the
paradoxical enhancement of TGF signalling in aortic aneurysms of
patients with LDS raises the possibility that inhibiting TGF
receptors may lead to a signalling imbalance.
A recent study took advantage of the effects of a TGF receptor
inhibitor on the neovasculature that mirrored the vascular defects
seen in TGFreceptor knockout mice. Shortterm treatment with an ALK5
inhibitor in mice that had intractable solid tumours led to reduced
vSMC coverage in neovessels, making them leaky and promoting the
accumulation of anticancer drugs in the tumour tissue101. Parallel
experiments targeting endoglin, which is strongly upregulated in
angiogenic tumour ECs, suggests that endoglin may also represent a
useful antiangiogenesis target for cancer treatment102,103,104.
Equally important is the development of proangiogenic therapies
used to treat ischaemic disease; manipulating TGF signalling may
represent a valuable therapeutic approach and preclinical data have
shown that endoglin promotes tissue repair by proangiogenic
circulating blood cells105.
Conclusions and future perspectivesIn the pursuit of molecular
mechanisms that underlie the vascular pathology of MFS and related
diseases, an important additional role for elastic ECM components
in controlling growth factor signalling was discovered5.
R E V I E W S
866 | NOvEMBER 2007 | vOLUME 8 www.nature.com/reviews/molcellbio
2007 Nature Publishing Group
-
Truncus arteriosusA single vessel that forms early in
development and then septates to form the aorta and pulmonary
trunk. A persistent truncus arteriosus is one that has failed to
septate, compromising the separation of pulmonary and systemic
circulations.
Aortopulmonary septation The process whereby the pulmonary trunk
and aorta separate during development.
Besides being structural scaffolds that provide tissue
integrity, elastic microfibril components have an instructive role
in regulating TGF processing and bioavailability. Latent TGF
complexes, bound via LTBPs to fibrillin1, are sequestered and
provide latent embedded signals that can be rapidly released by
proteases in response to tissue perturbations. This allows for
rapid and localized changes in TGF activity without the need for
denovo protein synthesis106. Dysregulation of TGF family signalling
in MFS may provide an explanation for clinical variability, as
disease pathogenesis may differ in a genotype and contextdependent
manner5. In addition, dysregulated TGF signalling may underlie the
pathology of other vascular disorders. For example, mutations in
the facilitative glucose transporter GLUT10 were found to cause
arterial tortuosity syndrome (ATS), and were associated with
upregulation of TGFSMAD signalling in the arterial wall107.
The instructive role for microfibrils has been extended to other
TGF family members such as BMP7, which binds via its prodomain
directly to fibrillin1, potentially targeting it to the ECM108.
Mutations in fibrillin2 (FBN2) are associated with disorganized
microfibrils and the Marfanlike disorder, congenital contractural
arachnodactyly (TABLE 1), and Fbn2 null mice display bilateral
syndactyly of forelimbs and hindlimbs109. Interestingly, mice that
are double heterozygous for null Fbn2 and Bmp7 alleles, which alone
are phenotypically silent, have impaired digit formation,
suggesting a functional interaction between fibrillin2 and BMP7
(REF. 109). The extracellular regulation of BMP signalling by ECM
components has been largely overlooked and needs further
investigation.
It is possible that fibrillins have dual and opposing roles in
regulating TGF activity. They may act positively by concentrating
ligand at sites of function, and negatively by sequestering the LLC
and inhibiting TGF activation. It is also interesting to note that
mice lacking Ltbp1 show persistent truncus arteriosus, a phenotype
that is also seen after neuralcrestspecific ablation of TGF
signalling110 (TABLE 1). This may reflect a failure to provide a
sequestered source of TGF in the absence of LTBP1 during
aortopulmonary septation.
Recognizing the importance of extracellular regulation of TGF
activity in aortic muscle cells in MFS and LDS has opened up
exciting possibilities for treatments of further muscle disorders.
Skeletal muscle weakness is an additional feature of MFS that
results from increased TGF signalling, which inhibits regeneration
of muscle cells and stimulates fibrosis. In parallel to the
beneficial effects of losartan on aortic function in a mouse model
of MFS, losartan also improves muscle morphology and enhances
satellitecell activation and muscle regeneration. This led to the
consideration that these benefits might ameliorate another
progressive muscle weakness disorder, Duchennes muscular dystrophy,
which is caused by familial mutations in dystrophin (DMD). Losartan
treatment lowered the abnormally high intramuscular TGF levels,
improved muscle histopathology and reduced fibrosis in a
Dmd-deficient mouse model of this disease, raising hopes that this
treatment may also benefit patients with Duchennes muscular
dystrophy111.
The complexities of TGF signalling are challenging to scientists
and clinicians alike. However, it is clear that TGF components
represent valuable therapeutic targets, even though the changes in
blood vessel architecture and blood pressure that are associated
with dysregulated TGF signalling indicate the necessity for careful
monitoring of these parameters in patients. The effectiveness of
losartan for treating dysregulated TGF signalling means that the
interaction between angiotensin and TGF signalling pathways is an
important area of investigation. There is also an urgent need to
improve our understanding of the extracellular regulation of TGF
ligand activation and the context dependence of cellular responses.
Furthermore, the phenotypes of mouse models with mutations in genes
that are associated with extracellular regulation of TGF signalling
(TABLE 1) raise the possibility that further human cardiovascular
syndromes may eventually be linked to some of these genes. Advances
in all of these areas will inform current treatment strategies and
open up possibilities for developing new therapies for a range of
vascular pathologies.
1. Massague, J. & Gomis, R. R. The logic of TGF signaling.
FEBS Lett. 580, 28112820 (2006).
2. Feng, X. H. & Derynck, R. Specificity and versatility in
TGF- signaling through Smads. Annu. Rev. Cell Dev. Biol. 21, 659693
(2005).
3. Blobe, G. C., Schiemann, W. P. & Lodish, H. F. Role of
transforming growth factor in human disease. N. Engl. J. Med. 342,
13501358 (2000).
4. Annes, J. P., Munger, J. S. & Rifkin, D. B. Making sense
of latent TGF activation. J. Cell Sci. 116, 217224 (2003).
5. Robinson, P. N. et al. The molecular genetics of Marfan
syndrome and related disorders. J. Med. Genet. 43, 769787
(2006).
6. Grainger, D. J. TGF- and atherosclerosis in man. Cardiovasc.
Res. 74, 213222 (2007).
7. Ruiz-Ortega, M., Rodriguez-Vita, J., Sanchez-Lopez, E.,
Carvajal, G. & Egido, J. TGF- signaling in vascular fibrosis.
Cardiovasc. Res. 74, 196206 (2007).
8. Bierie, B. & Moses, H. L. TGF- and cancer. Cytokine
Growth Factor Rev. 17, 2940 (2006).
9. Dubois, C. M., Laprise, M. H., Blanchette, F., Gentry, L. E.
& Leduc, R. Processing of transforming growth factor 1
precursor by human furin convertase. J. Biol. Chem. 270, 1061810624
(1995).
10. Beck, S. et al. Extraembryonic proteases regulate Nodal
signalling during gastrulation. Nature Cell Biol. 4, 981985
(2002).
11. Kanzaki, T. et al. TGF-1 binding protein: a component of the
large latent complex of TGF-1 with multiple repeat sequences. Cell
61, 10511061 (1990).
12. Saharinen, J., Taipale, J. & Keski-Oja, J. Association
of the small latent transforming growth factor- with an eight
cysteine repeat of its binding protein LTBP-1. EMBO J. 15, 245253
(1996).
13. Dallas, S. L. et al. Characterization and autoregulation of
latent transforming growth factor (TGF ) complexes in
osteoblast-like cell lines. Production of a latent complex lacking
the latent TGF -binding protein. J. Biol. Chem. 269, 68156821
(1994).
14. Saharinen, J. & Keski-Oja, J. Specific sequence motif of
8-Cys repeats of TGF- binding proteins, LTBPs, creates a
hydrophobic interaction surface for binding of small latent TGF-.
Mol. Biol. Cell 11, 26912704 (2000).
15. Isogai, Z. et al. Latent transforming growth factor -binding
protein 1 interacts with fibrillin and is a microfibril-associated
protein. J. Biol. Chem. 278, 27502757 (2003).
16. Nunes, I., Gleizes, P. E., Metz, C. N. & Rifkin, D. B.
Latent transforming growth factor- binding protein domains involved
in activation and transglutaminase-dependent cross-linking of
latent transforming growth factor-. J. Cell Biol. 136, 11511163
(1997).
17. Flaumenhaft, R. et al. Role of the latent TGF- binding
protein in the activation of latent TGF- by co-cultures of
endothelial and smooth muscle cells. J. Cell Biol. 120, 9951002
(1993).
18. Rifkin, D. B. Latent transforming growth factor- (TGF-)
binding proteins: orchestrators of TGF- availability. J. Biol.
Chem. 280, 74097412 (2005).
19. Dabovic, B. et al. Bone abnormalities in latent TGF-[]
binding protein (Ltbp)-3-null mice indicate a role for Ltbp-3 in
modulating TGF- bioavailability. J. Cell Biol. 156, 227232
(2002).
20. Sterner-Kock, A. et al. Disruption of the gene encoding the
latent transforming growth factor- binding protein 4 (LTBP-4)
causes abnormal lung development, cardiomyopathy, and colorectal
cancer. Genes Dev. 16, 22642273 (2002).
21. Chaudhry, S. S. et al. Fibrillin-1 regulates the
bioavailability of TGF1. J. Cell Biol. 176, 355367 (2007).
R E V I E W S
NATURE REvIEwS | molecular cell biology vOLUME 8 | NOvEMBER 2007
| 867 2007 Nature Publishing Group
-
An internal proteolytic fragment of fibrillin1 is shown to
regulate the bioavailability of TGF by inducing the release of the
large latent complex bound to microfibrils.
22. Ge, G. & Greenspan, D. S. BMP1 controls TGF1 activation
via cleavage of latent TGF-binding protein. J. Cell Biol. 175,
111120 (2006).Bone morphogenetic protein1 (BMP1)like
metalloprotease is shown to cleave large latent TGF binding protein
1 (LTBP1) at two specific sites, thereby liberating the large
latent TGF complex (LLC) from the extracellular matrix (ECM).
23. Pappano, W. N., Steiglitz, B. M., Scott, I. C., Keene, D. R.
& Greenspan, D. S. Use of Bmp1/Tll1 doubly homozygous null mice
and proteomics to identify and validate in vivo substrates of bone
morphogenetic protein 1/tolloid-like metalloproteinases. Mol. Cell
Biol. 23, 44284438 (2003).
24. Crawford, S. E. et al. Thrombospondin-1 is a major activator
of TGF-1 in vivo. Cell 93, 11591170 (1998).
25. Sheppard, D. Integrin-mediated activation of latent
transforming growth factor . Cancer Metastasis Rev. 24, 395402
(2005).
26. Yang, Z. et al. Absence of integrin-mediated TGF1 activation
in vivo recapitulates the phenotype of TGF1-null mice. J. Cell
Biol. 176, 787793 (2007).
27. Mu, D. et al. The integrin v8 mediates epithelial
homeostasis through MT1-MMP-dependent activation of TGF-1. J. Cell
Biol. 157, 493507 (2002).
28. Fontana, L. et al. Fibronectin is required for integrin
alphav6-mediated activation of latent TGF- complexes containing
LTBP-1. FASEB J. 19, 17981808 (2005).
29. Zhu, J. et al. 8 integrins are required for vascular
morphogenesis in mouse embryos. Development 129, 28912903
(2002).
30. Bader, B. L., Rayburn, H., Crowley, D. & Hynes, R. O.
Extensive vasculogenesis, angiogenesis, and organogenesis precede
lethality in mice lacking all alpha v integrins. Cell 95, 507519
(1998).
31. Goumans, M. J. et al. Balancing the activation state of the
endothelium via two distinct TGF- type I receptors. EMBO J. 21,
17431753 (2002).
32. Massague, J., Cheifetz, S., Boyd, F. T. & Andres, J. L.
TGF- receptors and TGF- binding proteoglycans: recent progress in
identifying their functional properties. Ann. N. Y. Acad. Sci. 593,
5972 (1990).
33. Lebrin, F., Deckers, M., Bertolino, P. & ten Dijke, P.
TGF- receptor function in the endothelium. Cardiovasc. Res. 65,
599608 (2005).
34. Goumans, M. J. et al. Activin receptor-like kinase (ALK)1 is
an antagonistic mediator of lateral TGF /ALK5 signaling. Mol. Cell
12, 817828 (2003).
35. Sankar, S., Mahooti-Brooks, N., Centrella, M., McCarthy, T.
L. & Madri, J. A. Expression of transforming growth factor type
III receptor in vascular endothelial cells increases their
responsiveness to transforming growth factor 2. J. Biol. Chem. 270,
1356713572 (1995).
36. Lopez-Casillas, F., Payne, H. M., Andres, J. L. &
Massague, J. Betaglycan can act as a dual modulator of TGF- access
to signaling receptors: mapping of ligand binding and GAG
attachment sites. J. Cell Biol. 124, 557568 (1994).
37. Lebrin, F. et al. Endoglin promotes endothelial cell
proliferation and TGF-/ALK1 signal transduction. EMBO J. 23,
40184028 (2004).
38. Blanco, F. J. et al. Interaction and functional interplay
between endoglin and ALK-1, two components of the endothelial
transforming growth factor- receptor complex. J. Cell Physiol. 204,
574584 (2005).
39. Adams, R. H. & Alitalo, K. Molecular regulation of
angiogenesis and lymphangiogenesis. Nature Rev. Mol. Cell Biol. 8,
464478 (2007).Excellent review on the molecular mechanisms that
underlie the formation and function of blood and lymphatic
vessels.
40. Goumans, M. J. & Mummery, C. Functional analysis of the
TGF receptor/Smad pathway through gene ablation in mice. Int. J.
Dev. Biol. 44, 253265 2000).
41. Akhurst, R. J., Lehnert, S. A., Faissner, A. & Duffie,
E. TGF in murine morphogenetic processes: the early embryo and
cardiogenesis. Development 108, 645656 (1990).
42. Bartram, U. et al. Double-outlet right ventricle and
overriding tricuspid valve reflect disturbances of looping,
myocardialization, endocardial cushion differentiation, and
apoptosis in TGF-2-knockout mice. Circulation 103, 27452752
(2001).
43. Kaartinen, V. et al. Abnormal lung development and cleft
palate in mice lacking TGF-3 indicates defects of
epithelial-mesenchymal interaction. Nature Genet. 11, 415421
(1995).
44. Tang, Y. et al. Epistatic interactions between modifier
genes confer strain-specific redundancy for Tgfb1 in developmental
angiogenesis. Genomics 85, 6070 (2005).
45. Deckers, M. M. et al. Bone morphogenetic proteins stimulate
angiogenesis through osteoblast-derived vascular endothelial growth
factor A. Endocrinology 143, 15451553 (2002).
46. Vinals, F. & Pouyssegur, J. Transforming growth factor 1
(TGF-1) promotes endothelial cell survival during in vitro
angiogenesis via an autocrine mechanism implicating TGF- signaling.
Mol. Cell Biol. 21, 72187230 (2001).
47. Ma, J., Wang, Q., Fei, T., Han, J. D. & Chen, Y. G.
MCP-1 mediates TGF- -induced angiogenesis by stimulating vascular
smooth muscle cell migration. Blood 109, 987994 (2007).
48. Ferrari, G. et al. VEGF, a prosurvival factor, acts in
concert with TGF-1 to induce endothelial cell apoptosis. Proc. Natl
Acad. Sci. USA 103, 1726017265 (2006).
49. Yang, X. et al. Targeted disruption of SMAD3 results in
impaired mucosal immunity and diminished T cell responsiveness to
TGF-. EMBO J. 18, 12801291 (1999).
50. Jadrich, J. L., OConnor, M. B. & Coucouvanis, E. The TGF
activated kinase TAK1 regulates vascular development in vivo.
Development 133, 15291541 (2006).
51. Roman, B. L. et al. Disruption of acvrl1 increases
endothelial cell number in zebrafish cranial vessels. Development
129, 30093019 (2002).
52. Hirschi, K. K., Burt, J. M., Hirschi, K. D. & Dai, C.
Gap junction communication mediates transforming growth factor-
activation and endothelial-induced mural cell differentiation.
Circ. Res. 93, 429437 (2003).
53. Carvalho, R. L. et al. Defective paracrine signalling by TGF
in yolk sac vasculature of endoglin mutant mice: a paradigm for
hereditary haemorrhagic telangiectasia. Development 131, 62376247
(2004).
54. Kumar, M. S. & Owens, G. K. Combinatorial control of
smooth muscle-specific gene expression. Arterioscler. Thromb. Vasc.
Biol. 23, 737747 (2003).
55. Carvalho, R. et al. Compensatory mechanisms activated during
vasculogenesis in mice by TGF-receptor deletion. J. Cell Sci. (in
the press).
56. Jiao, K. et al. Tgf signaling is required for
atrioventricular cushion mesenchyme remodeling during in vivo
cardiac development. Development 133, 45854593 (2006).
57. Seki, T., Hong, K. H. & Oh, S. P. Nonoverlapping
expression patterns of ALK1 and ALK5 reveal distinct roles of each
receptor in vascular development. Lab. Invest. 86, 116129
(2006).
58. Harradine, K. A. & Akhurst, R. J. Mutations of TGF
signaling molecules in human disease. Ann. Med. 38, 403414
(2006).
59. McAllister, K. A. et al. Endoglin, a TGF- binding protein of
endothelial cells, is the gene for hereditary haemorrhagic
telangiectasia type 1. Nature Genet. 8, 345351 (1994).
60. Johnson, D. W. et al. Mutations in the activin receptor-like
kinase 1 gene in hereditary haemorrhagic telangiectasia type 2.
Nature Genet. 13, 189195 (1996).
61. Gu, Y. et al. Functional analysis of mutations in the kinase
domain of the TGF- receptor ALK1 reveals different mechanisms for
induction of hereditary hemorrhagic telangiectasia. Blood 107,
19511954 (2006).
62. Jonker, L. & Arthur, H. M. Endoglin expression in early
development is associated with vasculogenesis and angiogenesis.
Mech. Dev. 110, 193196 (2002).
63. Fernandez, L. A. et al. Blood outgrowth endothelial cells
from hereditary haemorrhagic telangiectasia patients reveal
abnormalities compatible with vascular lesions. Cardiovasc. Res.
68, 235248 (2005).
64. Urness, L. D., Sorensen, L. K. & Li, D. Y. Arteriovenous
malformations in mice lacking activin receptor-like kinase-1.
Nature Genet. 26, 328331 (2000).
65. Sorensen, L. K., Brooke, B. S., Li, D. Y. & Urness, L.
D. Loss of distinct arterial and venous boundaries in mice lacking
endoglin, a vascular-specific TGF coreceptor. Dev. Biol. 261,
235250 (2003).
66. Letarte, M. et al. Reduced endothelial secretion and plasma
levels of transforming growth factor-1 in patients with hereditary
hemorrhagic telangiectasia type 1. Cardiovasc. Res. 68, 155164
(2005).
67. Sadick, H. et al. Patients with hereditary hemorrhagic
telangiectasia have increased plasma levels of vascular endothelial
growth factor and transforming growth factor-1 as well as high ALK1
tissue expression. Haematologica 90, 818828 (2005).
68. Scharpfenecker, M. et al. BMP-9 signals via ALK1 and
inhibits bFGF-induced endothelial cell proliferation and
VEGF-stimulated angiogenesis. J. Cell Sci. 120, 964972 (2007).
69. Allinson, K., Carvalho, R. L. C., van den Brink, S.,
Mummery, C. L. & Arthur, H. M. Generation of a floxed allele of
the mouse endoglin gene. Genesis 45, 391395 (2007).
70. Morrell, N. W. et al. Altered growth responses of pulmonary
artery smooth muscle cells from patients with primary pulmonary
hypertension to transforming growth factor- (1) and bone
morphogenetic proteins. Circulation 104, 790795 (2001).
71. Yang, X. et al. Dysfunctional Smad signaling contributes to
abnormal smooth muscle cell proliferation in familial pulmonary
arterial hypertension. Circ. Res. 96, 10531063 (2005).
72. Yu, P. B., Beppu, H., Kawai, N., Li, E. & Bloch, K. D.
Bone morphogenetic protein (BMP) type II receptor deletion reveals
BMP ligand-specific gain of signaling in pulmonary artery smooth
muscle cells. J. Biol. Chem. 280, 2444324450 (2005).
73. Foletta, V. C. et al. Direct signaling by the BMP type II
receptor via the cytoskeletal regulator LIMK1. J. Cell Biol. 162,
10891098 (2003).
74. West, J. et al. Pulmonary hypertension in transgenic mice
expressing a dominant-negative BMPRII gene in smooth muscle. Circ.
Res. 94, 11091114 (2004).
75. Harrison, R. E. et al. Molecular and functional analysis
identifies ALK-1 as the predominant cause of pulmonary hypertension
related to hereditary haemorrhagic telangiectasia. J. Med. Genet.
40, 865871 (2003).
76. Mizuguchi, T. et al. Heterozygous TGFBR2 mutations in Marfan
syndrome. Nature Genet. 36, 855860 (2004).
77. Loeys, B. L. et al. A syndrome of altered cardiovascular,
craniofacial, neurocognitive and skeletal development caused by
mutations in TGFBR1 or TGFBR2. Nature Genet. 37, 275281 (2005).
78. Pannu, H. et al. Mutations in transforming growth factor-
receptor type II cause familial thoracic aortic aneurysms and
dissections. Circulation 112, 513520 (2005).
79. Topouzis, S. & Majesky, M. W. Smooth muscle lineage
diversity in the chick embryo. Two types of aortic smooth muscle
cell differ in growth and receptor-mediated transcriptional
responses to transforming growth factor-. Dev. Biol. 178, 430445
(1996).
80. Waldo, K. L. et al. Secondary heart field contributes
myocardium and smooth muscle to the arterial pole of the developing
heart. Dev. Biol. 281, 7890 (2005).
81. Neptune, E. R. et al. Dysregulation of TGF- activation
contributes to pathogenesis in Marfan syndrome. Nature Genet. 33,
407411 (2003).
82. Habashi, J. P. et al. Losartan, an AT1 antagonist, prevents
aortic aneurysm in a mouse model of Marfan syndrome. Science 312,
117121 (2006).Aortic aneurysm in a mouse model of MFS is shown to
be associated with increased TGF signalling and can be prevented by
TGFneutralizing antibody or the angiotensin II type 1 receptor
(AT1) blocker, losartan.
83. Carta, L. et al. Fibrillins 1 and 2 perform partially
overlapping functions during aortic development. J. Biol. Chem.
281, 80168023 (2006).
84. Judge, D. P. et al. Evidence for a critical contribution of
haploinsufficiency in the complex pathogenesis of Marfan syndrome.
J. Clin. Invest. 114, 172181 (2004).
85. Denton, C. P. et al. Fibroblast-specific expression of a
kinase-deficient type II transforming growth factor (TGF) receptor
leads to paradoxical activation of TGF signaling pathways with
fibrosis in transgenic mice. J. Biol. Chem. 278, 2510925119
(2003).
86. Wang, W. et al. Essential role of Smad3 in angiotensin
II-induced vascular fibrosis. Circ. Res. 98, 10321039 (2006).
87. Zhou, Y., Poczatek, M. H., Berecek, K. H. &
Murphy-Ullrich, J. E. Thrombospondin 1 mediates angiotensin II
induction of TGF- activation by cardiac and renal cells under both
high and low glucose conditions. Biochem. Biophys. Res. Commun.
339, 633641 (2006).
R E V I E W S
868 | NOvEMBER 2007 | vOLUME 8 www.nature.com/reviews/molcellbio
2007 Nature Publishing Group
-
88. Yanagisawa, H. et al. Fibulin-5 is an elastin-binding
protein essential for elastic fibre development in vivo. Nature
415, 168171 (2002).
89. Nakamura, T. et al. Fibulin-5/DANCE is essential for
elastogenesis in vivo. Nature 415, 171175 (2002).
90. McLaughlin, P. J. et al. Targeted disruption of fibulin-4
abolishes elastogenesis and causes perinatal lethality in mice.
Mol. Cell Biol. 26, 17001709 (2006).
91. Hanada, K. et al. Perturbations of vascular homeostasis and
aortic valve abnormalities in fibulin-4 deficient mice. Circ. Res.
100, 738746 (2007).
92. Cambien, F. et al. Polymorphisms of the transforming growth
factor- 1 gene in relation to myocardial infarction and blood
pressure. The Etude Cas-Temoin de lInfarctus du Myocarde (ECTIM)
Study. Hypertension 28, 881887 (1996).
93. Zacchigna, L. et al. Emilin1 links TGF- maturation to blood
pressure homeostasis. Cell 124, 929942 (2006).Mice lacking emilin1,
a Cysrich secreted glycoprotein that prevents maturation of proTGF
precursor by furin convertases, display elevated blood pressure due
to increased TGF signalling in the vasculature.
94. August, P. & Suthanthiran, M. Transforming growth factor
signaling, vascular remodeling, and hypertension. N. Engl. J. Med.
354, 27212723 (2006).
95. Venkatesha, S. et al. Soluble endoglin contributes to the
pathogenesis of preeclampsia. Nature Med. 12, 642649 (2006).Soluble
endoglin is shown to lead to dysregulated TGF signalling in the
vasculature and may act in concert with VEGFR1 to induce severe
preeclampsia.
96. Cudmore, M. et al. Negative regulation of soluble Flt-1 and
soluble endoglin release by heme oxygenase-1. Circulation 115,
17891797 (2007).
97. Lopez-Novoa, J. M. Soluble endoglin is an accurate predictor
and a pathogenic molecule in pre-eclampsia. Nephrol. Dial.
Transplant. 22, 712714 (2007).
98. Akhurst, R. J. Large- and small-molecule inhibitors of
transforming growth factor- signaling. Curr. Opin. Investig. Drugs
7, 513521 (2006).
99. Derynck, R., Akhurst, R. J. & Balmain, A. TGF- signaling
in tumor suppression and cancer progression. Nature Genet. 29,
117129 (2001).
100. Akhurst, R. J. TGF signaling in health and disease. Nature
Genet. 36, 790792 (2004).
101. Kano, M. R. et al. Improvement of cancer-targeting therapy,
using nanocarriers for intractable solid tumors by inhibition of
TGF- signaling. Proc. Natl Acad. Sci. USA 104, 34603465
(2007).Application of a smallmolecule TGF type I receptor inhibitor
at a low dose decreased pericyte coverage of the endothelium
specifically in the tumour neovasculature, and promoted
accumulation of anticancer nanocarriers in tumours.
102. Jerkic, M. et al. Reduced angiogenic responses in adult
endoglin heterozygous mice. Cardiovasc. Res. 69, 845854 (2006).
103. Duff, S. E., Li, C., Garland, J. M. & Kumar, S. CD105
is important for angiogenesis: evidence and potential applications.
FASEB J. 17, 984992 (2003).
104. Tan, G. H. et al. Combination of low-dose cisplatin and
recombinant xenogeneic endoglin as a vaccine induces synergistic
antitumor activities. Int. J. Cancer 112, 701706 (2004).
105. van Laake, L. W. et al. Endoglin has a crucial role in
blood cell-mediated vascular repair. Circulation 114, 22882297
(2006).
106. Ramirez, F. & Rifkin, D. B. Cell signaling events: a
view from the matrix. Matrix Biol. 22, 101107 (2003).
107. Coucke, P. J. et al. Mutations in the facilitative glucose
transporter GLUT10 alter angiogenesis and cause arterial tortuosity
syndrome. Nature Genet. 38, 452457 (2006).
108. Gregory, K. E. et al. The prodomain of BMP-7 targets the
BMP-7 complex to the extracellular matrix. J. Biol. Chem. 280,
2797027980 (2005).
109. Arteaga-Solis, E. et al. Regulation of limb patterning by
extracellular microfibrils. J. Cell Biol. 154, 275281 (2001).
110. Choudhary, B. et al. Cardiovascular malformations with
normal smooth muscle differentiation in neural crest-specific type
II TGF receptor (Tgfbr2) mutant mice. Dev. Biol. 289, 420429
(2006).
111. Cohn, R. D. et al. Angiotensin II type 1 receptor blockade
attenuates TGF--induced failure of muscle regeneration in multiple
myopathic states. Nature Med. 13, 204210 (2007).
112. Kamei, M. et al. Endothelial tubes assemble from
intracellular vacuoles in vivo. Nature 442, 453456 (2006).
113. Barbara, N. P., Wrana, J. L. & Letarte, M. Endoglin is
an accessory protein that interacts with the signaling receptor
complex of multiple members of the transforming growth factor-
superfamily. J. Biol. Chem. 274, 584594 (1999).
114. George, E. L., Georges-Labouesse, E. N., Patel-King, R. S.,
Rayburn, H. & Hynes, R. O. Defects in mesoderm, neural tube and
vascular development in mouse embryos lacking fibronectin.
Development 119, 10791091 (1993).
115. Lawler, J. et al. Thrombospondin-1 is required for normal
murine pulmonary homeostasis and its absence causes pneumonia. J.
Clin. Invest. 101, 982992 (1998).
116. Huang, X. Z. et al. Inactivation of the integrin 6 subunit
gene reveals a role of epithelial integrins in regulating
inflammation in the lung and skin. J. Cell Biol. 133, 921928
(1996).
117. Stenvers, K. L. et al. Heart and liver defects and reduced
transforming growth factor 2 sensitivity in transforming growth
factor type III receptor-deficient embryos. Mol. Cell Biol. 23,
43714385 (2003).
118. Srinivasan, S. et al. A mouse model for hereditary
hemorrhagic telangiectasia (HHT) type 2. Hum. Mol. Genet. 12,
473482 (2003).
119. Larsson, J. et al. Abnormal angiogenesis but intact
hematopoietic potential in TGF- type I receptor-deficient mice.
EMBO J. 20, 16631673 (2001).
120. Beppu, H. et al. BMPR-II heterozygous mice have mild
pulmonary hypertension and an impaired pulmonary vascular
remodeling response to prolonged hypoxia. Am. J. Physiol. Lung
Cell. Mol. Physiol. 287, L1241L1247 (2004).
121. Lane, K. B. et al. Heterozygous germline mutations in
BMPR2, encoding a TGF- receptor, cause familial primary pulmonary
hypertension. The International PPH Consortium. Nature Genet. 26,
8184 (2000).
122. Gallione, C. J. et al. A combined syndrome of juvenile
polyposis and hereditary haemorrhagic telangiectasia associated
with mutations in MADH4 (SMAD4). Lancet 363, 852859 (2004).
123. Loeys, B.L. et al. Aneurysm syndromes caused by mutations
in the TGF- receptor. N. Engl. J. Med. 355, 788798 (2006).Mutations
in either TGFBR1 or TGFBR2 are shown to predispose patients to
aggressive and widespread vascular disease.
124. Todorovic, V. et al. Long form of latent TGF- binding
protein 1 (Ltbp1L) is essential for cardiac outflow tract septation
and remodeling. Development (in the press).
AcknowledgementsResearch in our laboratories is supported by
grants from the Dutch Cancer Society, the EC (Angiotargeting and
Tumour Host Genomics), the Ludwig Institute for Cancer Research,
the Netherlands Organization for Scientific Research, the British
Heart Foundation, Newcastle Hospital Trustees, the Cookson
Foundation and the Wellcome Trust. We are grateful to our
colleagues for valuable discussion, and apologize to those whose
contributions have not been cited because of space constraints.
DATABASESEntrez Gene:
http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=geneACVRL1 |
BMPR2OMIM:
http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=OMIMarterial
tortuosity syndrome | Duchennes muscular dystrophy | familial
thoracic aneurysm disorder | hereditary haemorrhagic telangiectasia
type 1 | HHT type 2 | LoeysDietz syndrome | Marfan syndrome | MFS
type 2 | pulmonary arterial hypertensionUniProtKB:
http://ca.expasy.org/sprotALK1 | betaglycan | BMP1 | emilin1 |
endoglin | fibrillin1 | fibrillin2 | SMAD1 | SMAD2 | SMAD3 | SMAD4
| SMAD5 | TGF1 | TGF2 | TGF3 | TGFBR1 | TGFBR2
FURTHER INFORMATIONPeter ten Dijkes homepage:
http://www.lumc.nl/1050/research/signaaltransductie.htmlHelen
Arthurs homepage:
http://www.ncl.ac.uk/ihg/staff/profile/helen.arthurHHT Foundation
International web site: http://www.hht.org/The US National Marfan
Foundation web site: http://www.marfan.org
all linkS are acTiVe in THe online PDF
R E V I E W S
NATURE REvIEwS | molecular cell biology vOLUME 8 | NOvEMBER 2007
| 869 2007 Nature Publishing Group