Human disease-causing NOG missense mutations:
Effects on noggin secretion, dimer formation,
and bone morphogenetic protein binding
Jose Marcelino*, Christopher M. Sciortino†, Michael F. Romero†, Lynn M. Ulatowski‡, R. Tracy Ballock‡,
Aris N. Economides§, Peter M. Eimon¶, Richard M. Harland¶, and Matthew L. Warman*?
*Department of Genetics and Center for Human Genetics, and‡Department of Orthopaedics, Case Western Reserve University and University Hospitals of
Cleveland, Cleveland, OH 44106;†Department of Physiology and Biophysics, Case Western Reserve University, Cleveland, OH 44106;§Regeneron
Pharmaceuticals, Inc., Tarrytown, NY 10591; and¶Department of Molecular and Cell Biology, Division of Genetics and Development,
University of California, Berkeley, CA 94720
Communicated by Michael S. Levine, University of California, Berkeley, CA, July 17, 2001 (received for review June 1, 2001)
Secreted noggin protein regulates bone morphogenetic protein
activity during development. In mice, a complete loss of noggin
protein leads to multiple malformations including joint fusion,
normal. In humans, heterozygous NOG missense mutations have
been found in patients with two autosomal dominant disorders of
joint development, multiple synostosis syndrome (SYNS1) and a
milder disorder proximal symphalangism (SYM1). This study inves-
tigated the effect of one SYNS1 and two SYM1 disease-causing
missense mutations on the structure and function of noggin. The
SYNS1 mutation abolished, and the SYM1 mutations reduced, the
secretion of functional noggin dimers in transiently transfected
to resemble the heterozygous state, did not interfere with wild-
type noggin secretion. These data indicate that the human disease-
functional dimeric noggin. Therefore, we conclude that noggin has
both species-specific and joint-specific dosage-dependent roles
during joint formation. Surprisingly, in contrast to the COS-7 cell
studies, the SYNS1 mutant was able to form dimers in Xenopus
laevis oocytes. This finding indicates that there also exist species-
specific differences in the ability to process mutant noggin
bone morphogenetic protein (BMP) family of secreted signaling
molecules comprises a large fraction of the transforming growth
factor ? superfamily and has important roles in this process (1).
BMPs were first purified from demineralized bone matrix as
proteins capable of causing ectopic bone formation after s.c. or
i.m. administration in rodents (2); the BMP protein family has
grown to include 15 distinct members. The essential roles that
BMPs play in skeletogenesis include recruiting mesenchymal
cells into future skeletal anlagen, promoting mesenchymal cell
proliferation and differentiation into chondroblasts and osteo-
blasts, and inducing apoptosis at sites of future joints (3–5). An
individual BMP may play each of these roles in a temporally
dependent, site-specific manner. Regulation of BMP activity
occurs in several different ways including the direct inhibition of
BMP signaling by secreted antagonists.
laevis embryos, noggin can bind and inhibit BMP-4 and thereby
induce the formation of the head and other dorsal structures (7).
Noggin is posttranslationally modified and is secreted as a
disulfide-bonded homodimer (8). In addition to binding to
BMP-4, it binds to several other BMP family members including
differentiation factor-5) (7, 9, 10). Roles for noggin during
mammalian development were demonstrated by knocking out
tructures of the appendicular skeleton arise from the con-
densation and differentiation of mesenchymal cells. The
mutation in the noggin gene (Nog) appear normal, whereas mice
homozygous for Nog-null alleles display several developmental
defects, including a severely malformed skeletal system (11, 12).
Noggin-null mice have shorter bones, are missing skeletal ele-
ments, and lack multiple articulating joints (11).
We have previously shown that heterozygous missense muta-
tions in the human noggin gene (NOG) cause two similar, but
clinically distinct, skeletal dysplasias, proximal symphalangism
and SYNS1 have apical joint fusions as their principal feature
(13), but SYNS1 is distinguished from SYM1 by also having hip
and vertebral fusions (14). The absence of a phenotype in mice
that are heterozygous for noggin-null mutations contrasts with
the autosomal dominant skeletal phenotype observed in hu-
mans. Furthermore, NOG missense mutations appear capable of
causing a spectrum of joint involvement (i.e., SYM1 and
SYNS1). These observations led us to study the synthesis,
secretion, and BMP binding activity of wild-type and mutant
noggin to explore the mechanism by which human NOG mis-
sense mutations exert their effect.
Materials and Methods
Noggin and BMP Expression Vectors. The coding sequence for
human noggin, a 232-aa residue polypeptide, is contained within
a single exon (13). The noggin coding sequence from the
genomic DNA of two patients with SYM1 and one patient with
SYNS1 was amplified by using the previously described primers,
NOG1 and NOG5 (13). The SYM1-derived mutants were a
glycine to cysteine substitution at residue 189 (G189C) and a
proline to leucine substitution at residue 223 (P223L); the
SYNS1-derived mutant was a tryptophan to glycine substitution
at residue 217 (W217G) (13). The PCR amplimers were cloned
alleles were identified by DNA sequencing. Wild-type NOG
sequence and the three mutant NOG sequences were moved
from pCR2.1 into the Xenopus expression vector pT7TS (gift of
P. Krieg, Univ. of Texas, Austin) by using EcoRV and SpeI
cloning sites. Wild-type and mutant NOG sequence was moved
from pT7TS into the mammalian expression vector pcDNA3
(Invitrogen) by using HindIII and XbaI cloning sites. A wild-type
noggin construct containing three tandem-myc epitopes
(EQKLISEEDLGG) at its C-terminal end was obtained from
Regeneron Pharmaceuticals. This tagged version of noggin was
subsequently moved to pcDNA3 to maintain vector uniformity.
Abbreviations: BMP, bone morphogenetic protein; SYM1, proximal symphalangism;
SYNS1, multiple synostosis syndrome.
?To whom reprint requests should be addressed. E-mail: firstname.lastname@example.org.
The publication costs of this article were defrayed in part by page charge payment. This
article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C.
§1734 solely to indicate this fact.
September 25, 2001 ?
vol. 98 ?
no. 20 ?
A BMP-14 (also known as CDMP1 and GDF5) expression
construct in which the active domain of BMP-14 is myc-tagged
and fused to the pro-domain of dorsalin has been described (15),
as has a myc-tagged BMP-4 expression construct (16).
Transfections. COS-7 cells were cultured in growth medium
(DMEM with 10% FBS?100 units/ml penicillin?100 ?g/ml strep-
tomycin?250 ng/ml amphotericin B; GIBCO?BRL) at 37°C in an
atmosphere containing 5% CO2. Cells at ?80% confluence in
Molecular Biochemicals) and 5 ?g of mammalian expression
plasmid DNA in Opti-MEM I according to manufacturer’s
recommendations. When cotransfection of noggin wild-type?
noggin mutant or noggin?BMP was performed, 5 ?g of each
plasmid was used. Titration experiments were also performed by
cotransfecting 1 ?g of untagged noggin (wild-type and mutants)
with decreasing amounts (1.0, 0.5, and 0.1 ?g) of myc-tagged
noggin (wild type). Transfected COS-7 cells were maintained for
24 h in growth medium and then switched to 10 ml of serum-free
DMEM?plate. Transfection efficiency (?50%) was assessed by
cotransfection with a green fluorescent protein construct. After
24 h, the culture media and the cell layer were separately
Coculture experiments of wild-type or mutant noggin (1
?g)-transfected COS-7 cells and myc-tagged BMP-14 (1 ?g)-
transfected COS-7 cells were performed by using cell culture
inserts (6-well plate, Falcon). BMP-14-transfected COS-7 cells
were maintained for 24 h in growth medium and then switched
to serum-free DMEM. Twenty-four hours later, inserts contain-
ing noggin transfected COS-7 cells were added to this system.
Total conditioned media (5 ml) were collected after 24 h of
Oocyte Isolation and Injection. X. laevis were purchased from
Xenopus Express (Beverly Hills, FL). Oocytes were removed,
and collagenase dissociated (2 mg?ml) in Ca2?-free Ringer’s
solution for 35–40 min as described (17). Capped RNA was
synthesized by using the linearized pT7TS expression constructs
and the T7 mMessage mMachine kit according to manufactur-
er’s instructions (Ambion, Austin, TX). Oocytes were injected
with 50 nl of 1 ?g??l (50 ng) noggin-capped RNA or water and
maintained at 18°C in OR3 (modified Leibovitz media, 5%
Pen?Strep, pH 7.5). Three to four oocytes were incubated in 200
?l of OR3solution for 72 h after injection.
Western Blot Analysis.EachinjectedXenopusoocytewasdissolved
in 20 ?l of 2? sample loading buffer, boiled for 5 min, and then
centrifuged at 12,000 ? g for 5 min at room temperature; 30 ?l
was loaded per lane. Conditioned media from the COS-7 cells
were cleared by centrifugation at 1,800 ? g for 10 min at room
temperature. After centrifugation, equal aliquots of conditioned
media supernatant and 2? sample loading buffer were mixed
and boiled for 5 min; 20 ?l was loaded per lane. The COS-7 cell
layer was scraped and extracted with 1 ml of RIPA buffer (150
mM NaCl?1% Nonidet P-40?0.5% deoxycholate?0.1% SDS?50
mM Tris, pH 8) and centrifuged at 12,000 ? g for 10 min at room
temperature. Equal aliquots of the cell lysis supernatant and 2?
sample loading buffer were mixed and boiled for 5 min; 20 ?l was
loaded per lane.
Samples were separated by SDS?PAGE under nonreducing or
reducing (5% ?-mercaptoethanol) conditions and transferred
onto Immobilon-P (Millipore). Blots were blocked for 1 h with
5% nonfat dry milk in TBS (150 mM NaCl?20 mM Tris, pH 7.5)
and then incubated with either 50 ng?ml anti-noggin antibody
(rat anti-noggin RP57–16, Regeneron Pharmaceuticals) or a
1:250 dilution anti-myc antibody (Santa Cruz Biotechnology) in
antibody buffer (2.5% nonfat dry milk?150 mM NaCl?20 mM
Tris?0.05% Tween 20, pH 7.5) for 1 h. After three washes with
TBST (150 mM NaCl?20 mM Tris?0.05% Tween 20, pH 7.5),
blots were incubated with a 1:8,000 dilution of horseradish-
peroxidase secondary antibody (goat anti-rat, Santa Cruz Bio-
technology) in antibody buffer for 1 h and then washed again
three times with TBST. Immunoreactive proteins were detected
with the ECL Plus chemiluminescent detection system (Amer-
Coimmunoprecipitation. Conditioned media (0.5 ml) from COS-7
cells that had been transfected with the pcDNA3 constructs
(wild type or mutants) were mixed with conditioned media (0.5
ml) from cells that had been transfected with myc-tagged BMP
(BMP-4 or BMP-14) constructs. Media were incubated for 16 h
at 4°C and then immunoprecipitated by using anti-myc antibody
(1:100) and 40 ?l of protein G (Amersham Pharmacia) for 16 h
cotransfected with pcDNA3 noggin constructs (wild type or
mutants) and myc-tagged BMP constructs (BMP-4 or BMP-14)
were similarly immunoprecipitated. Immunoprecipitates were
washed three times with RIPA buffer, dissolved in 40 ?l of 1?
sample loading buffer, boiled for 5 min, briefly vortexed, and
centrifuged at 12,000 ? g for 5 min at room temperature; 10 ?l
of sample was loaded per lane. Western analysis of coimmuno-
precipitated noggin was performed as described above.
Noggin Functional Assay. Wild-type and mutant NOG sequences
were moved from pT7TS into pCS107 (18) Xenopus expression
vectors by EcoRI digest. Capped synthetic mRNAs were tran-
scribed by using linearized expression constructs and the mMes-
sage mMachine kit (Ambion); 50 pg of synthetic mRNA was
injected into the marginal zone of a single ventral blastomere of
a four-cell embryo. All injections were in volumes of about 10 nl.
Embryos were allowed to develop to tailbud stage when they
were assayed for the presence of secondary axes.
Noggin Missense Mutations Affect the Protein’s Ability to Form
Secreted Disulfide-Linked Dimers in COS-7 Cells. SDS?PAGE per-
formed under reducing conditions indicated that the transiently
transfected COS-7 cells transcribed and translated mutant nog-
gin similar to wild-type noggin, although some mutant protein
had altered posttranslational processing (Fig. 1B). Because
wild-type noggin polypeptides normally form disulfide-linked
homodimers, the polypeptides were analyzed under nonreducing
conditions (Fig. 1A). Mutant noggin polypeptides formed
disulfide-linked dimers less efficiently than wild-type noggin.
Furthermore, mutant polypeptides formed a large molecular
weight aggregate that is not found in wild-type noggin, indicative
of abnormally disulfide-bonded protein.
Because noggin is a secreted protein, wild-type and mutant
noggin expression was also analyzed by using the conditioned
media from the transfected COS-7 cells. Western blots of
conditioned media electrophoresed under nonreducing condi-
tions revealed that the SYM1 noggin mutants (P223L and
G189C) showed decreased levels of disulfide-linked dimeric
noggin compared with wild type; in fact, the G189C mutant
protein is barely able to form dimers (Fig. 2A). Some mutant
noggin also seems to be secreted in either monomeric or at least
nondisulfide bonded form (Fig. 2A). The SYNS1 noggin mutant
(W217G) did not appear secreted as either a monomer or a
dimer. As with the cell extracts, all three mutant noggin polypep-
tides were present as high molecular weight disulfide-linked
aggregates. When conditioned medium was run in reducing
conditions to assess the apparent levels of secreted noggin, the
SYNS1 mutant (W217G) showed the least amount of immuno-
detectable noggin (Fig. 2B). However, as indicated by Fig. 1B,
this decrease is not caused by altered epitope recognition of this
particular mutant protein or a decrease in protein synthesis.
www.pnas.org?cgi?doi?10.1073?pnas.201367598Marcelino et al.
These results suggest that there are differences in each missense
mutation’s ability to form disulfide-linked dimers and to be
Mutant Noggin Polypeptides Do Not Affect Wild-Type Noggin Expres-
sion. Because SYM1 and SYNS1 are associated with heterozy-
gous NOG missense mutations, it is possible that the product of
the mutant allele could interfere with the product of the
wild-type allele. To test this hypothesis, a myc-tagged, wild-type
noggin construct was cotransfected with wild-type and mutant
constructs. Western blot analysis of conditioned media from the
different cotransfectants showed that the synthesis, dimeriza-
tion, and secretion of myc-tagged, wild-type noggin were not
significantly affected by the mutant allele (Fig. 3A). Specifically,
myc-tagged wild-type noggin was not found in monomeric form
or in high molecular weight species. Furthermore, altering the
relative amount of mutant to wild-type noggin during the
cotransfection experiments did not affect the ability of wild-type
protein to be synthesized or secreted (data not shown), indicat-
ing that the synthesis of mutant polypeptide does not interfere
with the synthesis of wild-type noggin.
Both Wild-Type and Mutant Noggin Dimers Can Bind BMPs. Noggin
can bind to several BMPs, including BMP-14 (also known as
GDF5 for growth?differentiation factor 5 or CDMP1 for
cartilage-derived morphogenetic protein 1) (10). To determine
whether the secreted noggin mutants are able to bind BMPs,
conditioned media from noggin transfectants and BMP trans-
fectants were mixed to allow the interaction of noggin proteins
with myc-tagged BMP-14. Immunoprecipitation with an anti-
myc antibody directed against myc-tagged BMP-14 protein
resulted in the marked coprecipitation of wild-type noggin and
the two SYM1 mutants (Fig. 4). Coimmunoprecipitation exper-
iments using myc-tagged BMP-4 were also performed, and
similar results were obtained (data not shown). Only disulfide-
stabilized dimeric noggin was coprecipitated with BMP-14 and
BMP-4, suggesting that free noggin monomer cannot bind to
BMPs (data not shown). Because cells may coexpress both
noggin and BMPs during development, coimmunoprecipitation
of noggin mutant proteins with BMP-14 was also performed on
with noggin and BMP-14 constructs. Wild-type noggin and both
SYM1 mutant noggins were coprecipitated as disulfide-linked
dimers (Fig. 5). SYNS1-derived mutant noggin (W217G) coim-
munoprecipitated with BMP-14 as a high molecular aggregate
(Fig. 5). It is interesting that dimeric SYM1-derived mutant
protein (G189C) is strongly coimmunoprecipitated when coex-
pressed with BMP-14, despite its predominant expression as a
monomer when solely expressed (see Figs. 2 and 5). This result
suggests that BMP-14 coexpression might facilitate formation of
mutant noggin dimers.
Intracellular BMP-14 Facilitates Formation of Mutant Noggin Dimers.
BMP-14 and the different noggin constructs were coexpressed in
COS-7 cells, and the conditioned media were analyzed by
analysis of cell lysates of COS-7 cells transfected with pcDNA3 containing
wild-type sequence (WT), no insert (? control), SYM1-derived sequences
(P223L, G189C), or SYNS1-derived sequence (W217G). (A) Nonreducing
SDS?8% PAGE and immunodetection using an anti-noggin mAb. Wild-type
noggin forms disulfide-bonded dimers and a higher molecular weight species
(denoted by arrow) that is an artifact of the electrophoresis method. SYM1-
derived mutant proteins are present as monomer, disulfide-bonded dimer,
and as higher molecular weight species (denoted by*) that do not penetrate
the separating or stacking gel. SYNS1-derived mutant protein is present as
monomer and as higher molecular weight species that do not penetrate the
separating or stacking gel. (B) Reducing SDS?10% PAGE and immunodetec-
tion using anti-noggin antibody. All higher molecular weight species resolve
to monomeric noggin. All three mutant noggins are synthesized in amounts
similar to wild-type noggin. A second, slightly slower migrating band in the
mutants indicates additional posttranslational modifications that are not
present in wild-type protein.
Noggin mutant proteins synthesized by COS-7 cells. Western blot
of conditioned media from COS-7 cells transfected with pcDNA3 containing
NOG wild-type sequence (WT), SYM1-derived sequences (P223L and G189C),
or SYNS1-derived sequence (W217G). (A) Nonreducing SDS?8% PAGE and
immunodetection using anti-noggin antibody. Wild-type noggin forms disul-
fide-bonded dimers and a higher molecular weight species (denoted by
arrow) that is an artifact of the electrophoresis method. The SYM1-derived
mutant (P223L) is secreted as a dimer, but less efficiently than wild type. The
SYM1-derived mutant (G189C) is secreted predominantly as a monomer; a
small amount of disulfide-bonded dimer is also secreted. The SYNS1 mutant
proteins are also present as higher molecular weight species (denoted by*),
weight species resolve to monomeric noggin. Less mutant protein is secreted
for the mutants (P223L and W219G) when compared with wild type.
Marcelino et al.
September 25, 2001 ?
vol. 98 ?
no. 20 ?
Western blot. Cotransfection with BMP-14 leads to a reproduc-
ible increase in the secretion of dimeric noggin species for both
SYM1-derived mutants (Fig. 6). The most striking effect occurs
in the SYM1 mutant (G189C), where the increase in dimeric
noggin is accompanied by a decrease in monomeric noggin.
Coculture experiments indicated that BMP-14 acts intracellu-
larly, rather than by facilitating or stabilizing noggin dimer
formation extracellularly, as coculturing noggin-expressing cells
with BMP-14-expressing cells did not enhance noggin dimer
formation (data not shown).
Mutant Noggin Proteins Can Form Dimers in Xenopus Oocytes.TheX.
laevis oocyte system has previously been used to study wild-type
noggin protein expression (8). In this system, all three mutant
noggin proteins were able to form disulfide-stabilized dimers;
the SYNS1 mutant, which did not form dimers in COS-7 cells,
dimerized more efficiently than one of the SYM1 mutants in the
Xenopus oocyte (Fig. 7).
Noggin Mutant Proteins Are Functional in X. laevis Embryos. Having
found that SYM1 mutant noggin protein, although secreted less
efficiently than wild-type protein by COS-7 cells, still retains the
ability to bind BMP, and that SYNS1 mutant protein can form
dimers in Xenopus oocytes, we sought to determine whether
these mutant noggin dimers retain function. Fifty picograms of
mRNA encoding each mutant was tested for dorsalizing activity
by injection into a single ventral blastomere of a four-cell
Xenopus embryo. Injected embryos were allowed to grow to the
tailbud stage and were scored for secondary axes. As shown in
Fig. 8, embryos ectopically expressing wild-type human noggin
on the ventral side display a partial secondary axis as expected
for a secreted BMP antagonist (19). Interestingly, injection of
of conditioned media from COS-7 cells that had been cotransfected with
pcDNA3 containing myc-tagged, wild-type NOG sequence (WT-myc) and
pcDNA3 containing wild-type sequence (WT), SYM1-derived sequences
(P223L, G189C), or SYNS1-derived sequence (W217G). (A) Nonreducing SDS?
PAGE and immunodetection using anti-myc antibody. Myc-tagged, wild-type
noggin is present only in dimeric form. (B) Nonreducing SDS?PAGE and
immunodetection using anti-noggin antibody. Most noggin is present as a
disulfide-linked dimer, indicating that myc-tagged wild-type noggin may be
more efficiently expressed than mutant noggin. It is possible that some
heterodimeric wild-type?mutant noggin may also be present. (C) Reducing
10% SDS?PAGE and immunodetection using anti-noggin antibody. Slightly
faster migrating bands in the WT and G189C lanes probably represent un-
tagged protein, which is 36 aa residues smaller than the myc-tagged noggin.
Coexpression of mutant and wild-type noggin. Western blot analysis
after precipitation with anti-myc antibody. Conditioned media from myc-
tagged BMP-14 transfected COS-7 cells (?BMP-14) was mixed with condi-
tioned media from COS-7 cells that had been transfected with pcDNA3 con-
taining NOG wild-type sequence (WT), SYM1-derived sequences (P223L,
G189C), or SYNS1-derived sequence (W217G). As negative controls, anti-myc
immunoprecipitation was performed on the conditioned media from the
different noggin transfectants not mixed with BMP-14 conditioned media
(?BMP-14). Samples were subjected to nonreducing SDS?8% PAGE and im-
noggin. Band migrating above the arrow is because of cross reaction with the
coprecipitate with BMP-14.
Interaction of noggin with BMP-14. Western blot analysis of noggin
Conditioned media from COS-7 cells cotransfected with pcDNA3 containing
myc-tagged BMP-14 (?BMP-14) and pcDNA3 containing NOG wild-type se-
quence (WT), SYM1-derived sequences (P223L, G189C), or SYNS1-derived
sequence (W217G). As negative controls, anti-myc immunoprecipitation was
performed on the conditioned media from the noggin constructs expressed
alone (?BMP-14). Samples were subjected to nonreducing SDS?8% PAGE and
of noggin, and asterisks denote high molecular weight species of noggin.
Band migrating between*and arrow is the anti-myc antibody. Wild-type
noggin dimer and the two SYM1 mutant dimers coprecipitate with BMP14.
Monomeric noggin does not coprecipitate with BMP-14. A much larger
amount of dimeric noggin mutant (G189C) coprecipitates than might be
expected based on this mutant’s predominant secretion as a monomer when
expressed alone (see Fig. 2).
Interaction of noggin mutant proteins and coexpressed BMP-14.
www.pnas.org?cgi?doi?10.1073?pnas.201367598Marcelino et al.
the same amount of P223L noggin mutant mRNA also elicited
a secondary axis in Xenopus embryos. Similar results were
obtained for W217G and G189C mutants (data not shown).
These results indicate that the noggin mutant proteins retain
their ability to antagonize BMP signaling in an intact embryo.
Noggin can bind to BMPs and inhibit their BMP signaling.
Analysis of mouse mutants shows that this interaction is essential
to modulate the activities of BMPs during development (11). In
contrast to the Nog-null allele in mice, which is asymptomatic in
the heterozygous state but causes a perinatal lethal, recessive
phenotype when homozygous, several heterozygous NOG mis-
sense mutations have been identified in human families with the
autosomal dominant phenotypes SYM1 and SYNS1 (13). To
explore the mechanism of mutational effect of the human
missense mutations in the noggin polypeptide, we studied their
expression using transiently transfected mammalian COS-7 cells
and noggin mRNA-injected Xenopus oocytes.
oocytes to handle mutant protein. Each missense mutation was
able to form disulfide-stabilized dimers in the Xenopus system
and to dorsalize developing Xenopus embryos (Figs. 7 and 8),
suggesting that the mutant proteins retain their ability to bind to
and antagonize BMPs. Using expression in COS-7 cells to
analyze synthesis and secretion of protein, the two SYM1 mutant
proteins, but not the SYNS1 mutant protein, were able to form
for the two SYM1 mutants was consistently reduced in compar-
ison with wild type (Figs. 1 and 2). Cotransfection studies in
COS-7 cells indicated that mutant polypeptide did not interfere
with the dimerization and secretion of wild-type protein (Fig. 3).
Taken together, these results suggest that reduction in the
amount of secreted dimeric noggin accounts for the two human
analysis of conditioned media from COS-7 cells transfected with pcDNA3
containing NOG wild-type sequence (WT), SYM1-derived sequences (P223L,
nonreducing SDS?8% PAGE and immunodetected using an anti-noggin anti-
body. The SYM1 mutant proteins have increased disulfide-linked dimer for-
mation. The SYM1 mutant (G189C) has the most striking change in relative
abundance of secreted monomer and dimer when coexpressed with BMP-14.
blot analysis of Xenopus oocytes that were injected with capped RNA from
pT7TS constructs containing SYM1-derived sequences (P223L and G189C) and
SYNS1-derived sequence (W217G). Samples were subjected to nonreducing
SDS?12% PAGE and immunodetected using an anti-noggin antibody. All
noggin mutants form disulfide-bonded dimers, albeit with varying efficiency.
Monomeric noggin migrates at 32 kDa, disulfide-bonded dimeric noggin
migrates at 52 kDa, and a high molecular weight noggin aggregate (denoted
by*) does not penetrate the separating gel.
Noggin mutant proteins synthesized by Xenopus oocytes. Western
formation in Xenopus embryos. Capped RNA (50 pg) synthesized from NOG
constructs (WT or P223L) were injected into the marginal zone of the ventral
blastomere of a four-cell embryo. Secondary axis formation is observed in
embryos but not in uninjected embryos.
Noggin mutant proteins are capable of inducing secondary axis
Marcelino et al.
September 25, 2001 ?
vol. 98 ?
no. 20 ?
skeletalphenotypesandthatallelesthatproducelessproteinwill Download full-text
cause more severe phenotypes. This study also highlights the
difference between humans and mice in their dosage require-
ments for secreted growth regulators during skeletal develop-
ment, as mice that are heterozygous for Nog-null alleles have
normal joints. Why humans and mice differ in their dosage
requirement for noggin is not known. Skeletal elements in
humans are larger and take longer to form than do the compa-
rable elements in mice. Therefore, humans may require more
noggin protein to form a gradient over larger distances or to
maintain a gradient over longer duration. Alternatively, cells
that comprise the human and mouse skeletal anlages may differ
in their absolute threshold requirements for BMP signals. At
present, noggin is the only secreted inhibitor of BMP signaling
that has been associated with a human disease phenotype.
Therefore, it is not yet possible to determine whether interspe-
cific differences in dosage requirements for inhibiting BMP
support for the latter hypothesis derives from two other auto-
somal dominant human phenotypes, brachydactyly type C and
situs ambiguus, which have been attributed to heterozygous
loss-of-function mutations that affect the secreted transforming
growth factor ? superfamily members BMP-14 and Lefty-1,
respectively (20, 21); comparable mutations in the mouse or-
thologs of these genes act as recessive alleles (22, 23).
not in COS-7 cells. This observation could reflect Xenopus
oocytes having factors that facilitate efficient noggin dimeriza-
tion, or COS-7 cells having a more sophisticated system for
detecting and not secreting mutant noggin polypeptides. Al-
though we do not know the precise reason for this difference, the
observation indicates a need to be cautious when exploring the
consequences of mutations in different model systems. Even
within the same organism, diverse cell types differ in their
abilities to process mutant polypeptides. For example, osteo-
blasts obtained from patients with osteogenesis imperfecta are
less able to identify and sequester mutant collagen molecules
than are the patients’ skin fibroblasts (24, 25). A second unex-
pected finding was that the SYM1 mutants had enhanced dimer
secretion from COS-7 cells when they were coexpressed with
BMP-14. This suggests that noggin may normally interact intra-
cellularly with its binding partner and that such interactions
could modify the consequences of some missense mutations by
improving their ability to dimerize and be secreted. Because
noggin and BMPs have overlapping patterns during develop-
ment, it is intriguing to speculate that endogenous intracellular
interactions between noggin and BMPs may be a particularly
potent regulator of BMP function during development.
We thank Dr. J. Terrig Thomas for providing the BMP-14 expression
construct and Dr. Daniel Constam for providing the BMP-4 expression
construct. This work was supported by National Institutes of Health
Developmental Biology Grant HD 07104 (to J.M.), a biomedical re-
search grant from the Arthritis Foundation (to M.L.W.), and National
Institutes of Health Grants AR 43527 (to M.L.W.) and GM 49346 (to
R.M.H). M.L.W. is an Assistant Investigator with the Howard Hughes
Scientist Award in Translational Research.
1. Hogan, B. L. (1996) Curr. Opin. Genet. Dev. 6, 432–438.
2. Van de Putte, K. A. & Urist, M. R. (1965) Clin. Orthop. 43, 257–270.
3. Groeneveld, E. H. & Burger, E. H. (2000) Eur. J. Endocrinol. 142, 9–21.
4. Wozney, J. M. & Rosen, V. (1998) Clin. Orthop. 346, 26–37.
5. Zou, H. & Niswander, L. (1996) Science 272, 738–741.
6. Smith, W. C. & Harland, R. M. (1992) Cell 70, 829–840.
7. Zimmerman, L. B., De Jesus-Escobar, J. M. & Harland, R. M. (1996) Cell 86,
8. Smith, W. C., Knecht, A. K., Wu, M. & Harland, R. M. (1993) Nature (London)
9. Chang, C. & Hemmati-Brivanlou, A. (1999) Development (Cambridge, U.K.)
10. Merino, R., Macias, D., Ganan, Y., Economides, A. N., Wang, X., Wu, Q.,
Stahl, N., Sampath, K. T., Varona, P. & Hurle, J. M. (1999) Dev. Biol. 206,
11. Brunet, L. J., McMahon, J. A., McMahon, A. P. & Harland, R. M. (1998)
Science 280, 1455–1457.
12. McMahon, J. A., Takada, S., Zimmerman, L. B., Fan, C. M., Harland, R. M.
& McMahon, A. P. (1998) Genes Dev. 12, 1438–1452.
13. Gong, Y., Krakow, D., Marcelino, J., Wilkin, D., Chitayat, D., Babul-Hirji, R.,
Hudgins, L., Cremers, C. W., Cremers, F. P., Brunner, H. G., et al. (1999) Nat.
Genet. 21, 302–304.
14. Krakow, D., Reinker, K., Powell, B., Cantor, R., Priore, M. A., Garber, A.,
Lachman, R. S., Rimoin, D. L. & Cohn, D. H. (1998) Am. J. Hum. Genet. 63,
15. Thomas, J. T., Kilpatrick, M. W., Lin, K., Erlacher, L., Lembessis, P., Costa,
T., Tsipouras, P. & Luyten, F. P. (1997) Nat. Genet. 17, 58–64.
16. Constam, D. B. & Robertson, E. J. (1999) J. Cell Biol. 144, 139–149.
17. Sciortino, C. M. & Romero, M. F. (1999) Am. J. Physiol. 277, F611–F623.
18. Baker, J. C., Beddington, R. S. & Harland, R. M. (1999) Genes Dev. 13,
19. Glinka, A., Wu, W., Onichtchouk, D., Blumenstock, C. & Niehrs, C. (1997)
Nature (London) 389, 517–519.
20. Polinkovsky, A., Robin, N. H., Thomas, J. T., Irons, M., Lynn, A., Goodman,
F. R., Reardon, W., Kant, S. G., Brunner, H. G., van der Burgt, I., et al. (1997)
Nat. Genet. 17, 18–19.
21. Kosaki, K., Bassi, M. T., Kosaki, R., Lewin, M., Belmont, J., Schauer, G. &
Casey, B. (1999) Am. J. Hum. Genet. 64, 712–721.
22. Storm, E. E., Huynh, T. V., Copeland, N. G., Jenkins, N. A., Kingsley, D. M.
& Lee, S. J. (1994) Nature (London) 368, 639–643.
23. Meno, C., Shimono, A., Saijoh, Y., Yashiro, K., Mochida, K., Ohishi, S., Noji,
S., Kondoh, H. & Hamada, H. (1998) Cell 94, 287–297.
24. Mundlos, S., Chan, D., Weng, Y. M., Sillence, D. O., Cole, W. G. & Bateman,
J. F. (1996) J. Biol. Chem. 271, 21068–21074.
25. Sarafova, A. P., Choi, H., Forlino, A., Gajko, A., Cabral, W. A., Tosi, L., Reing,
C. M. & Marini, J. C. (1998) Hum. Mutat. 11, 395–403.
www.pnas.org?cgi?doi?10.1073?pnas.201367598Marcelino et al.