RESEARCH ARTICLE 4029
Development 137, 4029-4038 (2010) doi:10.1242/dev.050591
© 2010. Published by The Company of Biologists Ltd
Cooperation of two ADAMTS metalloproteases in closure of
the mouse palate identifies a requirement for versican
proteolysis in regulating palatal mesenchyme proliferation
Hiroyuki Enomoto*, Courtney M. Nelson*, Robert P. T. Somerville, Katrina Mielke, Laura J. Dixon,
Kimerly Powell and Suneel S. Apte†
SUMMARY
We have identified a role for two evolutionarily related, secreted metalloproteases of the ADAMTS family, ADAMTS20 and
ADAMTS9, in palatogenesis. Adamts20 mutations cause the mouse white-spotting mutant belted (bt), whereas Adamts9 is
essential for survival beyond 7.5 days gestation (E7.5). Functional overlap of Adamts9 with Adamts20 was identified using
Adamts9+/–;bt/bt mice, which have a fully penetrant cleft palate. Palate closure was delayed, although eventually completed, in
both Adamts9+/–;bt/+ and bt/bt mice, demonstrating cooperation of these genes. Adamts20 is expressed in palatal mesenchyme,
whereas Adamts9 is expressed exclusively in palate microvascular endothelium. Palatal shelves isolated from Adamts9+/–;bt/bt mice
fused in culture, suggesting an intact epithelial TGF3 signaling pathway. Cleft palate resulted from a temporally specific delay in
palatal shelf elevation and growth towards the midline. Mesenchyme of Adamts9+/–;bt/bt palatal shelves had reduced cell
proliferation, a lower cell density and decreased processing of versican (VCAN), an extracellular matrix (ECM) proteoglycan and
ADAMTS9/20 substrate, from E13.5 to E14.5. Vcan haploinsufficiency led to greater penetrance of cleft palate in bt mice, with a
similar defect in palatal shelf extension as Adamts9+/–;bt/bt mice. Cell density was normal in bt/bt;Vcanhdf/+ mice, consistent with
reduced total intact versican in ECM, but impaired proliferation persisted in palate mesenchyme, suggesting that ADAMTScleaved versican is required for cell proliferation. These findings support a model in which cooperative versican proteolysis by
ADAMTS9 in vascular endothelium and by ADAMTS20 in palate mesenchyme drives palatal shelf sculpting and extension.
INTRODUCTION
Closure of the secondary palate (palatogenesis) separates the oral
and nasal cavities, permitting compartmentalization of digestive
and respiratory processes. Failure of palatogenesis leads to cleft
palate, one of the commonest human developmental malformations
(Cox, 2004). Although cleft palate can occur as an isolated
phenomenon in humans, it is frequently associated with cleft lip or
other anomalies (Cox, 2004). Palatogenesis results from complex
interactions involving pharyngeal ectoderm (palatal epithelium)
and mesenchyme derived primarily from craniofacial neural crest
cells (NCCs) (Dudas et al., 2007; Kaufman, 1992). During mouse
embryogenesis, the secondary palate develops from two sagittally
oriented palatal shelves that arise from the maxillary primordia and
project inferiorly on each side of the tongue at 12.5 days of
gestation (E12.5). They undergo elevation to a horizontal plane at
E13.5-14 and grow rapidly towards each other to meet in the
midline at E14.5-E15. Upon contact, the medial edge epithelium
overlying the apposed surfaces undergoes apoptosis or epithelialmesenchymal transformation, resulting in the formation of a
seamless bridge of mesenchyme by E16 (Dudas et al., 2007),
within which the palatine bone forms.
Department of Biomedical Engineering, Lerner Research Institute, Cleveland Clinic
Foundation, 9500 Euclid Avenue, Cleveland OH 44195, USA.
*These authors contributed equally to this work
†
Author for correspondence ([email protected])
Accepted 27 September 2010
Although only a few single gene mutations causing cleft palate
have been described in humans (Cox, 2004; Jugessur and Murray,
2005), the inactivation of several mouse genes has led to cleft palate
and identified key regulatory pathways in palatogenesis (Murray and
Schutte, 2004). Inherited cleft palate in mice is always recessive, as
a complete cleft of the secondary palate is incompatible with
survival. Mouse genetics has clearly demonstrated how signaling by
members of the TGF superfamily (Proetzel et al., 1995; Taya et al.,
1999), FGFs (Rice et al., 2004; Trokovic et al., 2003), PDGFs (Ding
et al., 2004) and EGF receptor ligands (Miettinen et al., 1999)
mediates epithelial-mesenchymal interactions that drive palatal
morphogenesis. In particular, midline fusion of the apposed medial
epithelial edge of the palatal shelves is specifically dependent on
TGF3 and its downstream signaling pathways (Dudas et al., 2004a;
Taya et al., 1999). Extensive extracellular matrix (ECM) remodeling
during palatogenesis and experimental evidence using
metalloprotease inhibitors support a role for metalloproteases in
palatogenesis (Blavier et al., 2001; Brown et al., 2002). However,
there is only one prior report of metalloprotease mutations causing
cleft palate. Combinatorial deletion of Mmp14 and Mmp16, which
encode transmembrane proteases, leads to severe craniofacial
dysmorphism and extensive skeletal anomalies, with cleft palate
observed in 80% of double-null embryos (Shi et al., 2008). These
defects were attributed to diminished turnover of the ECM protein
collagen I, and cleft palate could be attributed, at least in part, to the
global craniofacial anomalies (Shi et al., 2008), rather than to an
exclusive and specific interference with palatal shelves.
Here, we provide unequivocal genetic evidence that two
members of the ADAMTS (a disintegrin-like and metalloprotease
domain with thrombospondin type 1 motif) protease family,
DEVELOPMENT
KEY WORDS: ADAMTS, Cleft palate, Versican, Mouse
ADAMTS9 and ADAMTS20, act locally in palate closure and
cooperate in proteolysis of versican (VCAN), an abundant, large
aggregating proteoglycan found in association with hyaluronan in
ECM. Combinatorial genetic evidence using mice lacking these
ADAMTS genes and a mouse Vcan mutant strongly suggests that
defective versican proteolysis is the definitive mechanism
underlying the observed cleft palate phenotype. The data further
suggest that ADAMTS9 and ADAMTS20 are not required solely
for versican clearance, but that they might generate versican
proteolytic products that influence palate mesenchyme
proliferation.
ADAMTS designates a family of secreted metalloproteases that
contain thrombospondin type 1 repeats (TSRs). Of the 19
mammalian ADAMTS proteases, some are of relatively recent
evolutionary origin (Huxley-Jones et al., 2005) and may have
highly specialized functions (Apte, 2009). ADAMTS9 and
ADAMTS20 are the only mammalian ADAMTS with a closely
related non-chordate ortholog, named gon-1 in the nematode C.
elegans, and are presumed to have arisen by duplication and
continued evolution of a shared ancestral gene. gon-1 has an
essential role in cell migration during morphogenesis of the
nematode gonad (Blelloch and Kimble, 1999). Notable conserved
features of GON-1, ADAMTS9 and ADAMTS20 include
numerous TSRs (18 in GON-1 and 15 in ADAMTS9 and
ADAMTS20) and a unique C-terminal domain (Llamazares et al.,
2003; Somerville et al., 2003). gon-1 mutants are partially rescued
by ADAMTS9 (Hesselson et al., 2004), suggesting that their
proteolytic mechanism is conserved. Although ADAMTS9 lacks a
membrane anchor, it binds to the cell surface (Somerville et al.,
2003), where it is proteolytically active (Koo et al., 2006; Koo et
al., 2007). Thus, it is likely to be involved in cell surface or
pericellular ECM proteolysis. ADAMTS9 and ADAMTS20 belong
to an ADAMTS subgroup termed the proteoglycanases that
includes ADAMTS1, ADAMTS4 and ADAMTS5, the ability of
which to cleave large, aggregating proteoglycans, such as versican
and aggrecan, is crucial in physiological and disease processes
(Apte, 2009).
Adamts20 is mutated in the recessive mouse white-spotting
mutant belted (bt), which has an unpigmented belt on the lumbar
torso (Rao et al., 2003). Adamts20 promotes melanoblast survival
(Silver, 2008) and is expressed by dermal mesenchymal cells (Rao
et al., 2003). However, the expression of Adamts20 in craniofacial
mesenchyme (K. A. Jungers and S.S.A., unpublished) suggests
additional developmental roles. Adamts9 is widely expressed
during embryogenesis and – of relevance to the present
investigation – its mRNA is found in craniofacial mesoderm and in
microvascular endothelial cells of the embryo and adult organs
(Jungers et al., 2005; Koo et al., 2010). Recent work suggests that
Adamts9 might be anti-angiogenic in adult microvascular
endothelial cells (Koo et al., 2010). We considered the possibility
that mice lacking both ADAMTS9 and ADAMTS20 would show
novel defects. Although Adamts9–/– mice do not survive beyond
E7.5, we used them to obtain genetic evidence of cooperation with
Adamts20 in closure of the secondary palate and we investigated
the underlying mechanisms to identify a novel pathway operational
in palatogenesis.
MATERIALS AND METHODS
Mouse strains and genotyping
All mouse work was performed under a protocol approved by the
Cleveland Clinic Institutional Animal Care and Use Committee.
Adamts9lacZ/+ mice (referred to here as Adamts9+/–), btbei mice (referred to
Development 137 (23)
here as bt) and the Vcan mutant heart defect (hdf, referred to as Vcanhdf),
as well as null alleles of Adamts4 and Adamts5 and their genotyping
protocols have been described previously (McCulloch et al., 2009a;
McCulloch et al., 2009b; Rao et al., 2003; Silver, 2008). All mice were
extensively crossed into the C57Bl/6 strain. Bmp4 and Pdgfra mutant mice
in C57Bl/6 were kindly provided by Dr Jan Christian (Oregon Health and
Science University, Portland, OR, USA) and Dr Phillipe Soriano (Mount
Sinai Medical Center, NY, USA), respectively. For timed pregnancies the
date of the vaginal plug was designated E0.5. All analyses were performed
using littermate controls of the appropriate genotype. In the crosses that
were used to obtain Adamts9+/–;bt/bt mice, it was not possible to obtain
wild-type littermates; instead, littermates having the most appropriate
alternative genotype (bt/+) were used as controls.
Micro-computed tomography (mCT) analysis and scanning
electron microscopy (SEM)
Newborn mice were fixed in 10% buffered formalin for 48 hours and
transferred to 70% ethanol. mCT scanning of the skull was performed as
previously described (Le Goff et al., 2006). For SEM, the lower jaw was
removed, the heads were fixed, sputter-coated with gold and viewed from
the inferior aspect by SEM as previously described (Oblander et al., 2005).
At least five newborn mice encompassing all the illustrated genotypes were
examined by mCT and two litters of embryos representing the various
genotypes were analyzed by SEM.
Histology, immunohistochemistry (IHC), -gal histochemistry and
in situ hybridization (ISH)
Embryos were fixed in 4% paraformaldehyde (PFA) for durations
appropriate to their gestational stage, followed by embedding in paraffin
and Hematoxylin and Eosin (H&E) staining. Over 20 mice of each
genotype (Adamts9+/–;bt/bt and bt/+) were examined histologically at
E13.5 and E14.5. pSmad2/3 immunohistochemistry of palatal shelves was
performed with a polyclonal antibody (no. 3101, Cell Signaling
Technology, Danvers, MA, USA) using an indirect immunoperoxidase
method. Prior to pSmad2/3 staining, the sections were pretreated with 10
mM sodium citrate buffer (pH 6.0) for 10 minutes at 95°C. Cleavage of the
ADAMTS-processed Glu441-Ala442 peptide bond in versican (V1 splice
isoform enumeration) was detected by immunofluorescence staining using
a neo-epitope antibody that specifically detects versican cleaved at this site
but not intact versican (Affinity BioReagents, Golden, CO, USA) (Sandy
et al., 2001), and an Alexa 488-labeled secondary antibody. For
immunofluorescence, four mice of each of the Adamts9+/–;bt/bt and control
(bt/+) genotypes were analyzed at each time point (E13.5 and E14.5), with
consistent results.
For cell proliferation analysis, the phospho-histone H3 antibody (antipH3; Upstate Biotechnology-Millipore, Billerica, MA, USA) was applied
using an indirect immunoperoxidase method. The proliferation index in
palatal shelf mesenchyme was quantified by counting at least 500 nuclei in
sections from five mice each of the Adamts9+/–;bt/bt and control (bt/+)
genotypes at each time point (E13.5 and E14.5), using four sections from
each mouse. Cell density within the mesenchymal component of the palate
was quantified from photographic images of the shelves by overlaying
them with a counting grid in the center of the palatal shelf (the area of
interest is indicated in Fig. 4C), which consistently showed a difference
between Adamts9+/–;bt/bt and control (bt/+) palates. Four sections each
from five embryos of each genotype at each of the two time points were
used in this analysis. Statistical analysis of the quantitative data for cell
proliferation and cell density was performed using an unpaired, two-tailed
Student’s t-test.
ISH of paraffin sections from E14.5 embryos using Adamts9 and
Adamts20 35S-UTP-labeled antisense cRNA probes was performed as
previously described (Jungers et al., 2005; Rao et al., 2003). The
corresponding sense probes did not give signal above background, as
previously shown (Jungers et al., 2005; Rao et al., 2003). -gal staining of
Adamts9+/– mice followed by endomucin immunohistochemistry to
identify vascular endothelium (n6) was described previously (Koo et al.,
2010).
DEVELOPMENT
4030 RESEARCH ARTICLE
RESEARCH ARTICLE 4031
ADAMTS proteases and cleft palate
Palatal shelf organ culture
Using a published procedure (LaGamba et al., 2005) palatal shelves were
dissected from E14.5 Adamts9+/–;bt/bt embryos (n4) and placed on the
surface of a 0.9% agarose gel in DMEM/F12 supplemented with 10% fetal
bovine serum (FBS) and antibiotics (100 U/ml penicillin and 100 mg/ml
streptomycin), with their medial edge epithelia in contact. The medium was
replaced daily. After 3 days in culture, the shelves were fixed in 10%
neutral-buffered formalin and examined histologically by H&E staining of
coronal sections. Affigel beads (Bio-Rad, Hercules, CA, USA) were
soaked for 1 hour in conditioned medium containing the previously
identified N-terminal versican fragment (G1-DPEAAE441) expressed in
HEK293F cells. As a control, beads were soaked in conditioned medium
from HEK293F cells transfected with vector alone. E13.5 Adamts9+/–;bt/bt
maxillary arches containing the palate shelves were isolated and pre-soaked
beads were inserted into the palate shelves and placed onto 0.1-m pore
size Nuclepore filters (Whatman, Florham Park, NJ, USA). The explants
were cultured in a 24-well tissue culture dish at the air-liquid interface in
1:1 DMEM:Ham’s F12, plus 0.1% FBS. Explants were incubated for 14
hours at 37°C in 5% CO2 followed by fixation in 4% PFA and IHC using
anti-pH3 on paraffin sections.
Fig. 1. Cleft palate in Adamts9+/–;bt/bt mice. (A)External
appearance of newborn mice of the indicated genotypes. (B)Cleft
palate in Adamts9+/–;bt/bt mice (right, arrows). (C)mCT of the skull in
three projections shows that all skull bones are present and similar in
each genotype, other than abnormal formation of the palatine
(asterisk) and basi-sphenoid, the anomalous development of which is
presumed secondary to abnormal palatogenesis in Adamts9+/–;bt/bt
mice. The pterygoid processes of the sphenoid bone are indicated by
arrows; their wider spacing in Adamts9+/–;bt/bt mice is typical of
unfused palates.
All newborn Adamts9+/–;bt/bt mice had horizontally inclined,
unfused palatal shelves that were overlaid with an intact epithelium
of normal thickness. Analysis of palatogenesis in Adamts9+/–;bt/bt
embryos from E12.5-15.5 illustrated that the palatal shelves were
always formed correctly from the maxillary primordia by E13.5
(Fig. 2A). In Adamts9+/–;bt/bt mice, palate shelf growth towards
the midline from E14.5 onwards was always retarded, and in some
Adamts9+/–;bt/bt mice, palatal shelf elevation to the horizontal
orientation was also delayed relative to the controls (Fig. 2A). In
addition, palatal shelf sculpting (remodeling) in Adamts9+/–;bt/bt
embryos was consistently abnormal as, unlike in the other
genotypes obtained, the shelves remained broad, with a rounded
Table 1. Frequency of genotypes arising from crosses between Adamts9+/–;bt/+ and Adamts9+/+;bt/bt mice
Genotype
Stage
E9.5-18.5
Postnatal day 10
Adamts9+/+;bt/+
Adamts9+/–;bt/+
Adamts9+/+;bt/bt
Adamts9+/–;bt/bt
56 (28)
68 (35)
47 (24)
68 (35)
41 (21)
58 (30)
53 (27)
–
Shown are the number (percentage in parentheses) of mice of each genotype recovered at the indicated stages.
DEVELOPMENT
RESULTS
Perinatal lethality and cleft palate in
Adamts9+/–;bt/bt mice
In crosses between bt/bt and Adamts9+/–;bt/+ mice, or intercrosses
of Adamts9+/–;bt/+ mice, we did not obtain viable
Adamts9+/–;bt/bt mice upon genotyping between 8 and 10 days
after birth, whereas other expected genotypes were present (Table
1). However, Adamts9+/–;bt/bt embryos were identified in E18.5
litters at the expected frequency (Table 1). A few viable
Adamts9+/–;bt/bt mice were identified shortly after birth but died
within 24 hours with conspicuous aerogastria (air in the stomach),
suggesting a cleft palate, but were otherwise externally normal
(Fig. 1A). A complete cleft of the secondary palate (Fig. 1B)
was found in 100% of Adamts9+/–;bt/bt mice. By contrast,
bt/bt;Adamts5–/–, bt/bt;Adamts4–/–, Adamts5–/–;Adamts9+/– and
Adamts4–/–;Adamts5–/–;bt/+ mice were viable and did not have
cleft palate (data not shown), indicating a specific functional
association of ADAMTS20 with ADAMTS9, but not with
ADAMTS4 or ADAMTS5 in palatogenesis. Consistent with these
observations, Adamts4 and Adamts5 expression, as determined by
-gal staining in knockout mice (McCulloch et al., 2009a;
McCulloch et al., 2009b), is undetectable in palatal shelves during
development (McCulloch et al., 2009a) (data not shown). Thus, of
the proteoglycanase-encoding genes tested, only Adamts9 and
Adamts20 participate in palatogenesis.
mCT analysis of newborn Adamts9+/–;bt/bt mice consistently
demonstrated abnormal palatine and basi-sphenoid bones (Fig. 1C),
which have been previously described in mouse strains with a cleft
of the secondary palate (Trokovic et al., 2003; Zhang et al., 2002)
and are secondary to failed palatal closure. All other craniofacial
skeletal elements were consistently present and appeared normal
(Fig. 1C). The mandibles and tongue of Adamts9+/–;bt/bt mice were
similar to those of littermates (Fig. 1C), arguing against mechanical
interference with palatal shelf elevation (Dudas et al., 2004b).
4032 RESEARCH ARTICLE
Development 137 (23)
free edge and broader base. Unlike Fgf10 and Fgfr2b deficient
mice (Rice et al., 2004), adhesion of the palatal shelves to
surrounding structures was not seen in Adamts9+/–;bt/bt mice.
Independent contributions of Adamts9 and
Adamts20 to mouse palatogenesis
Comparative histology of the various genotypes between E12.5 and
E15.5 suggested that the elevation and growth of the palatal shelves
were also delayed in bt/bt and Adamts9+/–;bt/+ embryos (Fig. 2A),
although their shelves did not have the abnormal shape of
Adamts9+/–;bt/bt shelves. We used SEM to visualize the complete
anterior-posterior extent of the palatal shelves. We confirmed the
delay in shelf extension in Adamts9+/–;bt/+ and bt/bt mice and, as
expected, the consistently defective palatal closure in
Adamts9+/–;bt/bt mice (Fig. 2B). The delay of shelf extension
affected the anterior and posterior halves equally in the affected
genotypes, unlike in several other cleft palate mutants (He et al.,
2008; Yu et al., 2005; Zhang et al., 2002). Palatal extension to the
midline was delayed to a greater extent in bt/bt than in
Adamts9+/–;bt/+ mice (Fig. 2A,B). Despite the delay, palatogenesis
was successfully completed in E15.5 Adamts9+/–;bt/+ mice and
bt/bt mice as viewed by SEM (Fig. 2B). Adamts9+/–;bt/+ mice
survived at the predicted Mendelian frequency and did not develop
cleft palate, whereas bt/bt mice were later found to have a low
incidence of cleft palate (3%). These data strongly supported an
independent as well as a cooperative participation of ADAMTS9
and ADAMTS20 in palatogenesis.
Palatal epithelial fusion is a TGF3-dependent metalloproteasemediated process (Blavier et al., 2001; Taya et al., 1999) and
defects in TGF3 or its downstream signaling pathway result in a
specific defect in epithelial fusion (Cui et al., 2003; Cui et al.,
2005). Phospho (p) Smad2/3 immunohistochemistry, which reports
TGF receptor binding and activation of downstream Smad
signaling in palatal epithelium, provided a strong nuclear signal in
the prospective medial edge epithelium of Adamts9+/–;bt/bt mice
at E14.5 (which corresponds to the time point at which palatal
shelves normally begin fusion) despite these palatal shelves being
widely separated (see Fig. S1A in the supplementary material). To
determine whether Adamts9+/–;bt/bt palatal shelves were fusion
competent, they were dissected at E13.5 and placed with their
medial edge epithelia in contact in organ culture, where they fused
successfully, with disappearance of the epithelium (see Fig. S1B in
the supplementary material). Thus, a deficiency of Adamts9 and
Adamts20 does not impair this key pathway in palatogenesis, and
the in vitro data suggest that these palatal shelves would have fused
had they been able to meet in the midline.
Since a failure of NCCs to populate the branchial arches has
been described as a cause of cleft palate (Trokovic et al., 2003) and
Adamts9 is expressed in branchial arches (Jungers et al., 2005), we
used SEM to evaluate early branchial arch development. Externally
normal branchial arches in E10.5 Adamts9+/–;bt/bt embryos (see
Fig. S2 in the supplementary material) and, subsequently, the lack
of abnormalities in newborn mice in the skeletal elements
originating from the branchial arches (Fig. 1C), excluded abnormal
DEVELOPMENT
Fig. 2. Adamts9 and Adamts20 participate
independently and cooperatively in palatal shelf
extension towards the midline. (A)The palatal
fusion defect in Adamts9+/–;bt/bt mice results from
defective shelf elevation and extension to midline
between E13.5 and E14.5. Coronal sections at various
stages of palatal development are shown for the
indicated genotypes. Palatal shelves are indicated by
asterisks. There is reduced sculpting of the palatal
shelves from bt/bt and Adamts9+/–:bt/+ mice, as well as
Adamts9+/–;bt/bt mice. Scale bar: 100m. (B)Scanning
electron microscopy of palates (inferior view) of various
genotypes was obtained at the indicated gestational
ages. The arrows indicate the location of the medial
edge of the palatal shelf. Note the delayed closure at
E14.5 in Adamts9+/–;bt/+ and bt/bt mice, and failure to
close by E15.5 in Adamts9+/–;bt/bt mice.
ADAMTS proteases and cleft palate
RESEARCH ARTICLE 4033
endomucin immunostaining at E13.5 did not demonstrate a
difference in palatal vasculature between Adamts9+/–;bt/bt and bt/+
(control) mice (Fig. 3E,F).
branchial arch development as an underlying mechanism. Taken
together with the chronological analysis of palate development
from E13.5-15.5, this identified a temporally specific defect in
palatal shelf growth and/or remodeling involving mesenchyme.
Adamts9 and Adamts20 mRNAs are expressed in
distinct cell lineages in the palate
We determined expression of Adamts9 and Adamts20 by ISH in the
extending palatal shelves of non-transgenic embryos to examine
whether these genes acted locally and to identify the expressing
cells. At E13.5 (not shown) and E14.5 (Fig. 3A,B), both Adamts9
and Adamts20 mRNAs were expressed in the palatal shelf
mesenchyme. However, each mRNA had a distinct distribution,
with Adamts9 mRNA being associated with palatal capillaries and
Adamts20 being broadly expressed in mesenchyme (Fig. 3A,B).
Lineage tracing of -gal-stained Wnt1-Cre-R26R mice has
demonstrated that the vast majority of palate mesenchymal cells,
with the notable exception of vascular endothelium, which arises
from mesoderm, are of NCC origin (Yoshida et al., 2008). -gal
staining of E13.5 and E14.5 Adamts9+/– palatal shelves followed
by endomucin immunostaining (to identify vascular endothelium)
demonstrated that Adamts9 was exclusively expressed by capillary
endothelium (Fig. 3C,D). Thus, Adamts9 and Adamts20 cooperate
in palatogenesis but are expressed by distinct cell populations with
different embryonic origins. Visualization of palatal capillaries by
Versican processing is reduced in mutant palatal
shelves
The relative abundance of ECM in the Adamts9+/–;bt/bt shelves led
us to investigate the role of a previously identified substrate of
ADAMTS9 and ADAMTS20, versican (Silver, 2008; Somerville
et al., 2003), the proteolysis of which was recently implicated in
ADAMTS developmental phenotypes, i.e. white spotting (Silver,
2008), soft-tissue syndactyly (McCulloch et al., 2009b) and cardiac
dysmorphogenesis (Kern et al., 2007; Kern et al., 2006; Kern et al.,
2010; Stankunas et al., 2008). Versican is a highly hydrated
proteoglycan that forms large complexes with hyaluronan and thus
has space-filling properties. Immunostaining of versican showed it
to be abundant in palatal shelf mesenchyme (Fig. 5A). No
difference in staining of intact versican was detected between
Adamts9+/–;bt/bt and bt/+ palates (Fig. 5A). A neo-epitope
antibody that specifically recognizes the cleaved Glu441-Ala442
peptide bond in the versican core protein (V1 isoform
enumeration), but not intact versican (Sandy et al., 2001),
demonstrated cleaved versican diffusely within wild-type and bt/+
palatal mesenchyme (Fig. 5B). However, in Adamts9+/–;bt/bt mice
there was a striking decrease in the staining intensity and
distribution of cleaved versican in the palatal mesenchyme.
Cleaved versican was primarily detected around capillaries in
Adamts9+/–;bt/bt shelves, consistent with expression of the intact
Adamts9 allele in vascular endothelium and loss of ADAMTS20 in
mesenchyme (Fig. 5B). These results strongly associated the cleft
palate phenotype with decreased versican processing.
Genetic interaction of Vcan with Adamts20
implies an active role for versican proteolysis in
palatogenesis
To determine whether versican was merely a space-occupying
substrate that needed clearance by ADAMTS proteases during
palatal development (implying a passive role for versican), or
whether versican proteolysis was required to facilitate
mesenchymal proliferation (implying an active role for processed
DEVELOPMENT
Fig. 3. Adamts9 and Adamts20 are expressed locally in palatal
shelf, but in distinct cell types. (A,B)In situ hybridization (ISH) of
Adamts9 (A) and Adamts20 (B) mRNA in the palatal shelf (asterisk) at
E13.5 (coronal sections). ISH signal is red and nuclei blue (DAPI). Note
that expression of both Adamts9 and Adamts20 is confined to
mesenchyme, but in distinct distributions. (C,D)Combined -gal
histochemistry (blue) and endomucin immunohistochemistry (red) of
Adamts9+/– mice showing that Adamts9 expression is confined to
microvascular endothelial cells. (C)An overview of staining in an E13.5
palatal shelf (D) Palatal capillaries (from C, boxed) at higher
magnification. Note that endomucin-positive cytoplasm surrounds gal-stained nuclei. (E,F)Endomucin immunostaining shows comparable
capillary density and distribution in Adamts9+/–;bt/bt and bt/+ palatal
shelf at E13.5. Scale bars: 50m.
Mesenchymal cell proliferation is reduced in
Adamts9+/–;bt/bt palatal shelves
Immunostaining for pH3, a marker for cycling cells (Hendzel et al.,
1997; Yu et al., 2005), showed a statistically significant reduction of
mesenchymal proliferation at E13.5 and E14.5 (Fig. 4A,B), the time
points at which growth retardation was evident during palatogenesis
in Adamts9+/–;bt/bt mice. We therefore attributed the reduced growth
of the palatal shelves to decreased cell proliferation. Mesenchymal
cell density was reduced in Adamts9+/–;bt/bt palatal shelves
compared with those of bt/+ mice (Fig. 4C), which is consistent with
reduced cell proliferation and is also suggestive of a relative excess
of ECM (i.e. defective sculpting) in the mutant palatal shelves. We
evaluated cell death in palatal shelves of Adamts9+/–;bt/bt mice
because melanoblast survival was previously shown to be reduced in
bt/bt mice and because bt/bt;Adamts5–/– mice have impaired
apoptosis during interdigital web regression. However, few apoptotic
cells were seen in palatal mesenchyme of Adamts9+/–;bt/bt or bt/+
(control) mice (see Fig. S3 in the supplementary material), consistent
with apoptosis not being a crucial feature of palatogenesis until shelf
fusion (Mori et al., 1994), nor an underlying mechanism of this cleft
palate phenotype.
4034 RESEARCH ARTICLE
Development 137 (23)
Fig. 4. Decreased cell proliferation and
reduced cell density in Adamts9+/–;bt/bt
palatal shelves. (A)Anti-pH3
immunostaining (brown) in palatal shelves.
Note the blue staining (-gal) in capillaries of
Adamts9+/–;bt/bt palatal shelves. The
genotype and gestational ages of the mice are
indicated. The proliferation index was
determined in the area between the dotted
line and the medial edge of the shelf. Note
the rounded shape of the Adamts9+/–;bt/bt
palatal shelves. (B)Proliferation index (pH3stained nuclei/total nuclei counted) at E13.5
and E14.5 (n5 for each genotype at each
age). A statistically significant difference was
observed at both ages (*, P<0.05; #,
P<0.005). (C)Cell density was quantified in
the center of the palatal shelf by counting
nuclei in sections from Adamts9+/–;bt/bt and
bt/+ palatal shelves (n5 for each genotype) in
the area indicated by the box. (D)A
statistically significant reduction in cell density
(*P<0.05) at E13.5 and E14.5 was noted in
Adamts9+/–;bt/bt palatal shelves. Scale bars:
50m.
Liu et al., 2005; Zhang et al., 2002). We obtained viable
bt/bt;Pdgfra+/–, bt/bt;Bmp4+/–, Adamts9+/–;Pdgfra+/– and
Adamts9+/–;Bmp4+/– mice with unimpaired palate closure (data not
shown). The lack of interaction of Adamts9 and Adamts20 with
Bmp4 and Pdgfra demonstrated that the genetic interaction with
Vcan was specific, and suggested that cleft palate in Adamts9+/–;bt/bt
mice might result from interference with a novel mechanism
involving versican that was operational in the palatal mesenchyme.
The cleft secondary palate in affected bt/bt;Vcanhdf/+ embryos
(Fig. 6B) resulted from a similar delay in palatal shelf extension as
in Adamts9+/–:bt/bt embryos (Fig. 6C, Fig. 2A), with no other
apparent anomalies. However, bt/bt;Vcanhdf/+ palate shelves (Fig.
6C) did not have the blunt, rounded appearance of Adamts9+/–:bt/bt
Fig. 5. Versican is present normally in the palate but its processing is reduced in Adamts9+/–;bt/bt palatal shelves.
(A)Immunofluorescence microscopy of intact versican using a polyclonal antibody that recognizes the GAG- domain in the V1 and V0 isoforms
shows widespread versican localization throughout the palate mesenchyme. The boxed regions are shown at higher magnification to the right.
There was no discernible difference in versican localization between the two genotypes. (B)Immunofluorescence microscopy of cleaved versican
(using anti-DPEAAE) illustrates its broad distribution within the growing end of the bt/+ (control) palatal shelf. There is greatly reduced staining in
the intermesenchymal ECM of the Adamts9+/–;bt/bt palatal shelf, with residual staining present only around capillaries. The boxed regions are
shown at higher magnification to the right. Scale bars: 50m.
DEVELOPMENT
versican), we asked whether a reduction of Vcan gene dosage
would rescue delayed palatal shelf growth in bt/bt mice. The Vcan
mutant heart defect (hdf) dies at E10 with severe cardiac
development defects, but Vcanhdf/+ mice are viable and apparently
normal (Mjaatvedt et al., 1998).
Analysis of progeny resulting from crosses of bt/+;Vcanhdf/+ and
bt/bt mice showed that 65% of bt/bt;Vcanhdf/+ mice developed cleft
palate (Fig. 6A,B), despite having intact Adamts9 alleles. By
contrast, only 3% of bt/bt mice had cleft palate (Fig. 6A). To evaluate
the specificity of this genetic interaction, we crossed the Adamts9+/–
and bt/bt mice to mice with inactivated Pdgfra and Bmp4. Cleft
palate has been previously described in mice lacking these genes, as
a consequence of impaired palatal shelf growth (Ding et al., 2004;
ADAMTS proteases and cleft palate
RESEARCH ARTICLE 4035
Fig. 6. High incidence of cleft palate in
bt/bt;Vcanhdf/+ mice. (A)The incidence of cleft palate
in crosses between bt/+;Vcanhdf/+ and bt/bt mice.
(B)Complete cleft of the secondary palate in
bt/bt;Vcanhdf/+ mice. In the bt/bt;Vcanhdf/+ palate, the
edges of the palatal shelves are indicated by arrows.
The appearance of the palate is similar to that of the
Adamts9+/–;bt/bt shelves in Fig. 1B. (C)Histology of
palatal closure (coronal sections stained with
Hematoxylin and Eosin) in the indicated genotypes.
Note the retarded growth of palatal shelves of
bt/bt;Vcanhdf/+ mice, but their relatively normal shape
compared with Adamts9+/–;bt/bt shelves (see Fig. 2A),
and the failure to establish contact (asterisk at E16.5).
Scale bar: 100m.
lower production of the former. However, the higher penetrance of
cleft palate in bt/bt mice with Vcan haploinsufficiency than in bt/bt
mice suggested an additional, active role for proteolysed versican.
Immunostaining for intact versican confirmed a reduction in
versican content in the bt/bt;Vcanhdf/+ palatal shelves (Fig. 8A), and
staining for cleaved versican identified considerably weaker
mesenchymal and perivascular ECM staining compared with bt/+
shelves (Fig. 8B). The lack of ADAMTS20 is held to be
responsible for the lack of intermesenchymal processed versican,
and the persistence of versican processing around capillaries (Fig.
8B) is consistent with intact Adamts9 alleles in these mice.
Taken together, the findings are consistent with a model in
which decreased Vcan dosage results in a reduction of the overall
amount of processed versican in bt/bt;Vcanhdf/+ mice, which has
Fig. 7. Decreased cell proliferation but
normal cell density in bt/bt;Vcanhdf/+
palatal shelves. (A)Anti-pH3
immunostaining in palatal shelves. The
genotype and gestational ages of the mice
are indicated. The proliferation index was
determined in the area between the dotted
line and the medial edge of the shelf.
(B)Quantification of proliferation (pH3stained nuclei/total nuclei counted) at E13.5
and E14.5. *, P<0.05 (n4). (C)Cell density
was quantified in the center of the palatal
shelf by counting all nuclei in sections from
bt/bt;Vcanhdf/+ and bt/+ palatal shelves in the
area indicated by the box. (D)There is no
significant alteration in cell density at E13.5
or E14.5 in bt/bt;Vcanhdf/+ palatal shelves
(n4 for each genotype). Scale bars: 50m.
DEVELOPMENT
shelves (Fig. 2A, Fig. 4A). pH3 immunostaining demonstrated a
statistically significant reduction in cell proliferation at E14.5, but
not at E13.5, suggesting a similar mechanism as in
Adamts9+/–;bt/bt mice (Fig. 7A,B). However, mesenchymal density
was unaltered in bt/bt;Vcanhdf/+ palatal shelves compared with bt/+
shelves. Thus, unlike Adamts9+/–;bt/bt shelves, reduced cell density
was not seen in bt/bt;Vcanhdf/+ shelves. This observation is
consistent with a relative excess of ECM in Adamts9+/–;bt/bt mice
(i.e. reduced cell density), which are expected to have normal
versican production but decreased versican clearance with the
deficiency of two versican-clearing proteases. Reduced versican
production in bt/bt;Vcanhdf/+ shelves might well explain the lack of
a difference in cell density compared with bt/+ mice because the
potential accumulation of uncleaved versican could be offset by the
4036 RESEARCH ARTICLE
Development 137 (23)
Fig. 8. Versican GAG- and anti-DPEAAE
immunofluorescence reflects both
reduced Vcan gene dosage and reduced
versican processing in bt/bt;Vcanhdf/+
palatal shelves. (A)Immunofluorescence
microscopy of intact versican using a
polyclonal antibody that recognizes the
GAG- domain in the V1 and V0 isoforms,
showing reduced staining in bt/bt;Vcanhdf/+
palatal shelves, consistent with Vcan
haploinsufficiency. (B)Immunofluorescence
microscopy of cleaved versican (using antiDPEAAE) illustrates greatly reduced staining
in the bt/bt;Vcanhdf/+ palatal shelves. In bt/+
palatal shelves, note the bright signal
around palatal capillaries and its reduction
in the bt/bt;Vcanhdf/+ palatal shelf. Scale
bars: 50m.
significant effect on cell proliferation in palatal mesenchyme
compared with a control bead (see Fig. S4 in the supplementary
material).
DISCUSSION
Early embryonic lethality of the Adamts9-null mice precluded the
generation of double-null mice (with bt) for complete determination
of overlapping functions. Nevertheless, we examined the effect of
reduced Adamts9 gene dosage in a bt/bt background by generating
Adamts9+/–;bt/bt mice. These mice lack severe craniofacial
anomalies other than cleft palate, suggesting that reducing the gene
dosage of Adamts9 on the bt/bt background did not broadly impair
craniofacial neural crest-related functions and that palatogenesis in
bt/bt mice is sensitive to Adamts9 haploinsufficiency. Cleft palate in
Adamts9+/–;bt/bt mice results from a temporally restricted defect that
Fig. 9. Models of versican and ADAMTS function in the mesenchyme. The models summarize the experimental observations in wild-type,
Adamts9+/–;bt/bt (at E13.5 and E14.5) and bt/bt;Vcanhdf/+ (at E14.5) palatal shelves, and the proposed underlying mechanisms. In wild-type and
Adamts9+/–;bt/bt mice, a deep shade of blue indicates normal versican levels in the ECM, whereas bt/bt;Vcanhdf/+ mice have half the normal amount
(light-blue). The wild-type model indicates that cleaved versican is normally present in both the pericapillary and intermesenchymal ECM. In both
Adamts9+/–;bt/bt and bt/bt;Vcanhdf/+ palates there is a paucity of intermesenchymal cleaved versican because ADAMTS20 is absent. In
Adamts9+/–;bt/bt palates there is also a paucity of cleaved versican around capillaries because of Adamts9 haploinsufficiency. In bt/bt;Vcanhdf/+
palates there is less cleaved versican around capillaries owing to Vcan haploinsufficiency, although Adamts9 dosage is not reduced. We propose
that the amount of cleaved versican affects mesenchymal cell proliferation. In addition, the model depicts reduced cell density (i.e. the cell-to-matrix
ratio) in the Adamts9+/–;bt/bt palate, which results from both reduced clearance of versican (i.e. increased ECM) and decreased cell proliferation.
Note that these models do not include the epithelium, as the observed mechanism seems to be wholly operational in mesenchyme.
DEVELOPMENT
the same end result as the absence of ADAMTS20 and reduced
dosage of ADAMTS9 in Adamts9+/–;bt/bt mice (Fig. 9). In this
model, proteolysis of versican by ADAMTS9 (acting in
endothelium) and ADAMTS20 (made by mesenchymal cells)
contributes to the overall level of cleaved versican. Thus, these
proteases are postulated to work cooperatively in the palatal
shelf to enable mesenchymal proliferation. Interestingly,
Adamts9+/–;Vcanhdf/+ mice did not develop cleft palate, suggesting
that intact Adamts20 alleles and the remaining Adamts9 allele
provided sufficient versican proteolysis in mesenchyme to
compensate for reduced versican. To further extend this model
experimentally, we introduced a bead soaked in conditioned
medium containing the N-terminal versican fragment that
spans residues 1-441 (G1-DPEAAE441) into palates of E13.5
Adamts9+/–;bt/bt mice. However, this fragment did not have a
is localized to mesenchyme from E13.5-14.5. Crucial developmental
events that occur before or after this interval appeared to be
unaffected, as the branchial arches were morphologically normal,
craniofacial skeletal patterning was unaffected, and the palatal
shelves formed from the maxillary shelves and could fuse in vitro.
The reduced outgrowth of the palatal shelf in the Adamts9+/–;bt/bt
mutant mice resulted in part from reduced cell proliferation. A lack
of sculpting of the shelves and a failure of elongation, as a likely
consequence of impaired versican remodeling, are likely to be
contributory factors to the failure to meet in the midline.
NCCs make a significant contribution to craniofacial mesenchyme
and their interactions with ECM are believed to have a major
influence on their fate (Bronner-Fraser, 1993; Francis-West et al.,
1998). By contrast, the contribution of craniofacial mesoderm, which
is highly concentrated in the core of developing branchial arches, and
which is the source of the branchial arch vasculature and craniofacial
capillaries (Yoshida et al., 2008), is less well appreciated. Adamts9
expression in the palatal shelves was restricted to microvascular
endothelial cells, which are derived from mesoderm, whereas
Adamts20 is expressed in NCC-derived mesenchyme.
The versican-rich ECM of the palatal shelves is typical of the
provisional matrix present during embryogenesis. Versican has been
implicated in the regulation of neural crest migration and
cardiovascular development (Dutt et al., 2006; Henderson et al.,
1997; Landolt et al., 1995; Perissinotto et al., 2000; Perris et al.,
1996). Splotch, a Pax3 mutant, is believed to result in part from the
overexpression of versican, which results in skin pigmentation and
craniofacial defects (Henderson et al., 1997). However, once
organogenesis is completed, the provisional ECM requires clearance
to allow replacement by definitive ECM. ADAMTS proteases
appear to be crucial for the clearance of versican in provisional ECM
(Kern et al., 2007; Kern et al., 2006; Kern et al., 2010; McCulloch
et al., 2009b; Stankunas et al., 2008). However, this process, and the
specific proteases involved in remodeling the versican-rich ECM,
have not previously been investigated in palatogenesis.
Reduced versican proteolysis in Adamts9+/–;bt/bt palatal shelves,
together with altered shelf sculpting and reduced cell density,
strongly implicate defective versican turnover in the cleft palate. A
model of gene dosage-dependent cooperative proteolysis of
versican in the palate is supported by the ability of ADAMTS9 and
ADAMTS20 to cleave versican (Silver, 2008; Somerville et al.,
2003), by their expression during palatal shelf extension, and by
the delayed palate closure in Adamts9+/–;bt/+ and bt/bt mice.
Genetic interaction with Vcan, but not Bmp4 or Pdgfra, strongly
suggests that the observed interaction with Vcan was specific.
Taken together, the experimental findings suggest that Adamts9 and
Adamts20 act cooperatively via versican proteolysis in a novel
pathway that ensures closure of the secondary palate. This work
also identifies three new candidate genes for cleft palate in humans:
ADAMTS9, ADAMTS20 and VCAN.
Recently, we demonstrated that Adamts9 cooperates with
Adamts5 and Adamts20 in the regression of interdigital webs, a
process in which versican proteolysis has a crucial role (McCulloch
et al., 2009b). Specifically, combinatorial mutants of these genes
led to soft-tissue syndactyly, and reduced versican processing was
observed in the mutant webs (McCulloch et al., 2009b).
Coincidentally, versicanolysis appears to be required in
palatogenesis and web regression at approximately the same
gestational age (E14.5), and bt/bt;Vcanhdf/+ mice also develop softtissue syndactyly (McCulloch et al., 2009b). Insertion of a bead
containing the recombinant G1-DPEAAE441 versican fragment led
to enhanced apoptosis in Adamts5–/–;bt/bt webs (McCulloch et al.,
RESEARCH ARTICLE 4037
2009b). The present work suggests a similar bioactive function for
cleaved versican, but in the maintenance of mesenchymal
proliferation rather than in apoptosis. However, introduction of the
G1-DPEAAE441 versican fragment into mutant palates did not
affect cell proliferation. Therefore, although the genetic evidence
strongly implicates versican fragmentation in regulating cell
proliferation, the precise fragment or fragments that mediate the
effect remain to be identified, and it is likely that the versican core
protein undergoes ADAMTS-mediated proteolysis at multiple sites.
To explain the cooperative action of three ADAMTS proteases in
web regression, we previously proposed a model of cooperative
versicanolysis for the generation of a critical functional threshold of
a bioactive fragment (McCulloch et al., 2009b). Versican is widely
expressed during embryogenesis and in adult tissues, and versicandegrading ADAMTS proteases overlap considerably in their
expression patterns, such as in the cardiovascular system. Previous
work identified a significant role for ADAMTS1 and ADAMTS9 in
versican processing during myocardial compaction and
valvulogenesis (Kern et al., 2007; Kern et al., 2006; Kern et al.,
2010; Stankunas et al., 2008). In addition, recent work (C. Kern,
personal communication) has identified decreased versican
processing associated with failed endocardial cushion remodeling in
Adamts5–/– mice. In these mice, versican haploinsufficiency (i.e.
Adamts5–/–;Vcanhdf/+) led to substantial rescue of the valvular defect,
suggesting that the versican clearance function of ADAMTS5 is
highly significant in this developmental setting. Thus, depending on
the context, ADAMTS proteases may primarily mediate versican
clearance or contribute to both versican clearance and the generation
of bioactive versican fragments. These significant findings render it
necessary to expand investigation of the biological impact of
versican proteolysis by ADAMTS proteases, and to obtain additional
insights into the mechanisms of versican processing.
Acknowledgements
We thank David R. Beier for providing bt mice; Dr Christine Kern, Dr Corey
Mjaatvedt and Roche Pharmaceuticals for providing hdf mice; Dr Phillipe
Soriano and Dr Michelle Tallquist for Pdgfra mice; Dr Jan Christian and Dr
Brigid Hogan for Bmp4 mice; Dr Dietmar Vestweber for anti-endomucin
antibody; Amanda Allamong and Michael Braun for technical support for
histology; and Craig Bennetts for mCT. This work was supported by NIH award
AR49930 (to S.S.A.). Histology and mCT analyses were supported by NIH Core
Center award AR50953. Deposited in PMC for release after 12 months.
Competing interests statement
The authors declare no competing financial interests.
Supplementary material
Supplementary material for this article is available at
http://dev.biologists.org/lookup/suppl/doi:10.1242/dev.050591/-/DC1
References
Apte, S. S. (2009). A disintegrin-like and metalloprotease (reprolysin-type) with
thrombospondin type 1 motif (ADAMTS) superfamily: functions and mechanisms.
J. Biol. Chem. 284, 31493-31497.
Blavier, L., Lazaryev, A., Groffen, J., Heisterkamp, N., DeClerck, Y. A. and
Kaartinen, V. (2001). TGF-beta3-induced palatogenesis requires matrix
metalloproteinases. Mol. Biol. Cell 12, 1457-1466.
Blelloch, R. and Kimble, J. (1999). Control of organ shape by a secreted
metalloprotease in the nematode Caenorhabditis elegans. Nature 399, 586590.
Bronner-Fraser, M. (1993). Neural crest cell migration in the developing embryo.
Trends Cell Biol. 3, 392-397.
Brown, N. L., Yarram, S. J., Mansell, J. P. and Sandy, J. R. (2002). Matrix
metalloproteinases have a role in palatogenesis. J. Dent. Res. 81, 826-830.
Cox, T. C. (2004). Taking it to the max: the genetic and developmental mechanisms
coordinating midfacial morphogenesis and dysmorphology. Clin. Genet. 65, 163176.
Cui, X. M., Chai, Y., Chen, J., Yamamoto, T., Ito, Y., Bringas, P. and Shuler, C. F.
(2003). TGF-beta3-dependent SMAD2 phosphorylation and inhibition of MEE
proliferation during palatal fusion. Dev. Dyn. 227, 387-394.
DEVELOPMENT
ADAMTS proteases and cleft palate
Cui, X. M., Shiomi, N., Chen, J., Saito, T., Yamamoto, T., Ito, Y., Bringas, P., Chai,
Y. and Shuler, C. F. (2005). Overexpression of Smad2 in Tgf-beta3-null mutant
mice rescues cleft palate. Dev. Biol. 278, 193-202.
Ding, H., Wu, X., Bostrom, H., Kim, I., Wong, N., Tsoi, B., O’Rourke, M., Koh, G.
Y., Soriano, P., Betsholtz, C. et al. (2004). A specific requirement for PDGF-C in
palate formation and PDGFR-alpha signaling. Nat. Genet. 36, 1111-1116.
Dudas, M., Nagy, A., Laping, N. J., Moustakas, A. and Kaartinen, V. (2004a).
Tgf-beta3-induced palatal fusion is mediated by Alk-5/Smad pathway. Dev. Biol.
266, 96-108.
Dudas, M., Sridurongrit, S., Nagy, A., Okazaki, K. and Kaartinen, V. (2004b).
Craniofacial defects in mice lacking BMP type I receptor Alk2 in neural crest cells.
Mech. Dev. 121, 173-182.
Dudas, M., Li, W. Y., Kim, J., Yang, A. and Kaartinen, V. (2007). Palatal fusionwhere do the midline cells go? A review on cleft palate, a major human birth
defect. Acta Histochem. 109, 1-14.
Dutt, S., Kleber, M., Matasci, M., Sommer, L. and Zimmermann, D. R. (2006).
Versican V0 and V1 guide migratory neural crest cells. J. Biol. Chem. 281, 1212312131.
Francis-West, P., Ladher, R., Barlow, A. and Graveson, A. (1998). Signalling
interactions during facial development. Mech. Dev. 75, 3-28.
He, F., Xiong, W., Yu, X., Espinoza-Lewis, R., Liu, C., Gu, S., Nishita, M., Suzuki,
K., Yamada, G., Minami, Y. et al. (2008). Wnt5a regulates directional cell
migration and cell proliferation via Ror2-mediated noncanonical pathway in
mammalian palate development. Development 135, 3871-3879.
Henderson, D. J., Ybot-Gonzalez, P. and Copp, A. J. (1997). Over-expression of
the chondroitin sulphate proteoglycan versican is associated with defective neural
crest migration in the Pax3 mutant mouse (splotch). Mech. Dev. 69, 39-51.
Hendzel, M. J., Wei, Y., Mancini, M. A., Van Hooser, A., Ranalli, T., Brinkley, B.
R., Bazett-Jones, D. P. and Allis, C. D. (1997). Mitosis-specific phosphorylation of
histone H3 initiates primarily within pericentromeric heterochromatin during G2
and spreads in an ordered fashion coincident with mitotic chromosome
condensation. Chromosoma 106, 348-360.
Hesselson, D., Newman, C., Kim, K. W. and Kimble, J. (2004). GON-1 and fibulin
have antagonistic roles in control of organ shape. Curr. Biol. 14, 2005-2010.
Huxley-Jones, J., Apte, S. S., Robertson, D. L. and Boot-Handford, R. P. (2005).
The characterisation of six ADAMTS proteases in the basal chordate Ciona
intestinalis provides new insights into the vertebrate ADAMTS family. Int. J.
Biochem. Cell Biol. 37, 1838-1845.
Jugessur, A. and Murray, J. C. (2005). Orofacial clefting: recent insights into a
complex trait. Curr. Opin. Genet. Dev. 15, 270-278.
Jungers, K. A., Le Goff, C., Somerville, R. P. and Apte, S. S. (2005). Adamts9 is
widely expressed during mouse embryo development. Gene Expr. Patterns 5, 609617.
Kaufman, M. H. (1992). The Atlas Of Mouse Development. London: Academic Press.
Kern, C. B., Twal, W. O., Mjaatvedt, C. H., Fairey, S. E., Toole, B. P., IruelaArispe, M. L. and Argraves, W. S. (2006). Proteolytic cleavage of versican during
cardiac cushion morphogenesis. Dev. Dyn. 235, 2238-2247.
Kern, C. B., Norris, R. A., Thompson, R. P., Argraves, W. S., Fairey, S. E., Reyes,
L., Hoffman, S., Markwald, R. R. and Mjaatvedt, C. H. (2007). Versican
proteolysis mediates myocardial regression during outflow tract development. Dev.
Dyn. 236, 671-683.
Kern, C. B., Wessels, A., McGarity, J., Dixon, L. J., Alston, E., Argraves, W. S.,
Geeting, D., Nelson, C. M., Menick, D. R. and Apte, S. S. (2010). Reduced
versican cleavage due to Adamts9 haploinsufficiency is associated with cardiac and
aortic anomalies. Matrix Biol. 29, 304-316.
Koo, B. H., Longpre, J. M., Somerville, R. P., Alexander, J. P., Leduc, R. and
Apte, S. S. (2006). Cell-surface processing of pro-ADAMTS9 by furin. J. Biol.
Chem. 281, 12485-12494.
Koo, B. H., Longpre, J. M., Somerville, R. P., Alexander, J. P., Leduc, R. and
Apte, S. S. (2007). Regulation of ADAMST9 secretion and enzymatic activity by its
propeptide. J. Biol. Chem. 282, 16146-16154.
Koo, B. H., Coe, D. M., Dixon, L. J., Somerville, R. P., Nelson, C. M., Wang, L. W.,
Young, M. E., Lindner, D. J. and Apte, S. S. (2010). ADAMTS9 is a cellautonomously acting, anti-angiogenic metalloprotease expressed by microvascular
endothelial cells. Am. J. Pathol. 176, 1494-1504.
LaGamba, D., Nawshad, A. and Hay, E. D. (2005). Microarray analysis of gene
expression during epithelial-mesenchymal transformation. Dev. Dyn. 234, 132-142.
Landolt, R. M., Vaughan, L., Winterhalter, K. H. and Zimmermann, D. R. (1995).
Versican is selectively expressed in embryonic tissues that act as barriers to neural
crest cell migration and axon outgrowth. Development 121, 2303-2312.
Le Goff, C., Somerville, R. P., Kesteloot, F., Powell, K., Birk, D. E., Colige, A. C.
and Apte, S. S. (2006). Regulation of procollagen amino-propeptide processing
during mouse embryogenesis by specialization of homologous ADAMTS proteases:
insights on collagen biosynthesis and dermatosparaxis. Development 133, 15871596.
Liu, W., Sun, X., Braut, A., Mishina, Y., Behringer, R. R., Mina, M. and Martin, J.
F. (2005). Distinct functions for Bmp signaling in lip and palate fusion in mice.
Development 132, 1453-1461.
Llamazares, M., Cal, S., Quesada, V. and Lopez-Otin, C. (2003). Identification and
characterization of ADAMTS-20 defines a novel subfamily of metalloproteinases-
Development 137 (23)
disintegrins with multiple thrombospondin-1 repeats and a unique GON domain. J.
Biol. Chem. 278, 13382-13389.
McCulloch, D. R., Goff, C. L., Bhatt, S., Dixon, L. J., Sandy, J. D. and Apte, S. S.
(2009a). Adamts5, the gene encoding a proteoglycan-degrading metalloprotease,
is expressed by specific cell lineages during mouse embryonic development and in
adult tissues. Gene Expr. Patterns 9, 314-323.
McCulloch, D. R., Nelson, C. M., Dixon, L. J., Silver, D. L., Wylie, J. D., Lindner,
V., Sasaki, T., Cooley, M. A. W. S. A. and Apte, S. S. (2009b). ADAMTS
proteases generate active versican fragments that regulate interdigital web
regression. Dev. Cell 17, 687-698.
Miettinen, P. J., Chin, J. R., Shum, L., Slavkin, H. C., Shuler, C. F., Derynck, R.
and Werb, Z. (1999). Epidermal growth factor receptor function is necessary for
normal craniofacial development and palate closure. Nat. Genet. 22, 69-73.
Mjaatvedt, C. H., Yamamura, H., Capehart, A. A., Turner, D. and Markwald, R.
R. (1998). The Cspg2 gene, disrupted in the hdf mutant, is required for right
cardiac chamber and endocardial cushion formation. Dev. Biol. 202, 56-66.
Mori, C., Nakamura, N., Okamoto, Y., Osawa, M. and Shiota, K. (1994).
Cytochemical identification of programmed cell death in the fusing fetal mouse
palate by specific labelling of DNA fragmentation. Anat. Embryol. (Berl.) 190, 2128.
Murray, J. C. and Schutte, B. C. (2004). Cleft palate: players, pathways, and
pursuits. J. Clin. Invest. 113, 1676-1678.
Oblander, S. A., Zhou, Z., Galvez, B. G., Starcher, B., Shannon, J. M., Durbeej,
M., Arroyo, A. G., Tryggvason, K. and Apte, S. S. (2005). Distinctive functions
of membrane type 1 matrix-metalloprotease (MT1-MMP or MMP-14) in lung and
submandibular gland development are independent of its role in pro-MMP-2
activation. Dev. Biol. 277, 255-269.
Perissinotto, D., Iacopetti, P., Bellina, I., Doliana, R., Colombatti, A., Pettway,
Z., Bronner-Fraser, M., Shinomura, T., Kimata, K., Morgelin, M. et al. (2000).
Avian neural crest cell migration is diversely regulated by the two major
hyaluronan-binding proteoglycans PG-M/versican and aggrecan. Development
127, 2823-2842.
Perris, R., Perissinotto, D., Pettway, Z., Bronner-Fraser, M., Morgelin, M. and
Kimata, K. (1996). Inhibitory effects of PG-H/aggrecan and PG-M/versican on
avian neural crest cell migration. FASEB J. 10, 293-301.
Proetzel, G., Pawlowski, S. A., Wiles, M. V., Yin, M., Boivin, G. P., Howles, P.
N., Ding, J., Ferguson, M. W. and Doetschman, T. (1995). Transforming
growth factor-beta 3 is required for secondary palate fusion. Nat. Genet. 11, 409414.
Rao, C., Foernzler, D., Loftus, S. K., Liu, S., McPherson, J. D., Jungers, K. A.,
Apte, S. S., Pavan, W. J. and Beier, D. R. (2003). A defect in a novel ADAMTS
family member is the cause of the belted white-spotting mutation. Development
130, 4665-4672.
Rice, R., Spencer-Dene, B., Connor, E. C., Gritli-Linde, A., McMahon, A. P.,
Dickson, C., Thesleff, I. and Rice, D. P. (2004). Disruption of Fgf10/Fgfr2bcoordinated epithelial-mesenchymal interactions causes cleft palate. J. Clin. Invest.
113, 1692-1700.
Sandy, J. D., Westling, J., Kenagy, R. D., Iruela-Arispe, M. L., Verscharen, C.,
Rodriguez-Mazaneque, J. C., Zimmermann, D. R., Lemire, J. M., Fischer, J.
W., Wight, T. N. et al. (2001). Versican V1 proteolysis in human aorta in vivo
occurs at the Glu441-Ala442 bond, a site that is cleaved by recombinant
ADAMTS-1 and ADAMTS-4. J. Biol. Chem. 276, 13372-13378.
Shi, J., Son, M. Y., Yamada, S., Szabova, L., Kahan, S., Chrysovergis, K., Wolf,
L., Surmak, A. and Holmbeck, K. (2008). Membrane-type MMPs enable
extracellular matrix permissiveness and mesenchymal cell proliferation during
embryogenesis. Dev. Biol. 313, 196-209.
Silver, D. L., Hou, L., Somerville, R., Young, M. E., Apte, S. S. and Pavan, W. J.
(2008). The secreted metalloprotease ADAMTS20 is required for melanoblast
survival. PLoS Genet. 4, 1-15.
Somerville, R. P., Longpre, J. M., Jungers, K. A., Engle, J. M., Ross, M., Evanko,
S., Wight, T. N., Leduc, R. and Apte, S. S. (2003). Characterization of ADAMTS9 and ADAMTS-20 as a distinct ADAMTS subfamily related to Caenorhabditis
elegans GON-1. J. Biol. Chem. 278, 9503-9513.
Stankunas, K., Hang, C. T., Tsun, Z. Y., Chen, H., Lee, N. V., Wu, J. I., Shang, C.,
Bayle, J. H., Shou, W., Iruela-Arispe, M. L. et al. (2008). Endocardial Brg1
represses ADAMTS1 to maintain the microenvironment for myocardial
morphogenesis. Dev. Cell 14, 298-311.
Taya, Y., O’Kane, S. and Ferguson, M. W. (1999). Pathogenesis of cleft palate in
TGF-beta3 knockout mice. Development 126, 3869-3879.
Trokovic, N., Trokovic, R., Mai, P. and Partanen, J. (2003). Fgfr1 regulates
patterning of the pharyngeal region. Genes Dev. 17, 141-153.
Yoshida, T., Vivatbutsiri, P., Morriss-Kay, G., Saga, Y. and Iseki, S. (2008). Cell
lineage in mammalian craniofacial mesenchyme. Mech. Dev. 125, 797-808.
Yu, L., Gu, S., Alappat, S., Song, Y., Yan, M., Zhang, X., Zhang, G., Jiang, Y.,
Zhang, Z., Zhang, Y. et al. (2005). Shox2-deficient mice exhibit a rare type of
incomplete clefting of the secondary palate. Development 132, 4397-4406.
Zhang, Z., Song, Y., Zhao, X., Zhang, X., Fermin, C. and Chen, Y. (2002). Rescue
of cleft palate in Msx1-deficient mice by transgenic Bmp4 reveals a network of
BMP and Shh signaling in the regulation of mammalian palatogenesis.
Development 129, 4135-4146.
DEVELOPMENT
4038 RESEARCH ARTICLE