Download Failure of extraembryonic membrane formation and mesoderm

Proc. Natl. Acad. Sci. USA
Vol. 95, pp. 9378–9383, August 1998
Developmental Biology
Failure of egg cylinder elongation and mesoderm induction in
mouse embryos lacking the tumor suppressor smad2
(transforming growth factor bysmad4ygastrulationyextraembryonic membranes)
MICHAEL WEINSTEIN, XIAO YANG, CUILING LI, XIAOLING XU, JESSICA GOTAY,
AND
CHU-XIA DENG*
Laboratory of Biochemistry and Metabolism, National Institute of Diabetes and Digestive and Kidney Diseases, 10y9N105, 10 Center Drive, National Institutes of
Health, Bethesda, MD 20892
Communicated by Igor B. Dawid, National Institute of Child Health and Human Development, Bethesda, MD, June 12, 1998 (received for review
April 21, 1998)
phosphorylation these four SMAD proteins all form a stable
complex with SMAD4, with which they are translocated into
the nucleus (21, 22). In addition to the pathway-specific and
common SMADs there are also inhibitory SMAD proteins,
SMAD6 and SMAD7 (12, 13, 23, 24). Both smad2 and smad4
are on chromosome 18q21 in humans, and both have been
identified as tumor suppressor genes.
A great deal has been learned about the function of smad2
in Xenopus, where it has been shown to induce dorsal mesoderm (8, 11, 25). To learn more about the functions of smad2
in mammals we have disrupted it by gene targeting in mice.
Removal of the carboxyl-terminal half of the MH2 domain
results in recessive embryonic lethality at day 6.5 (E6.5) of
development because of a failure to form the extraembryonic
portion of the egg cylinder. The embryonic ectoderm is able to
proliferate, however, and fills the interior of the embryo at
later stages. At this time point they resemble embryos deficient
in the TGFb homologue NODAL (26–28).
ABSTRACT
smad genes constitute a family of nine members whose products serve as intracellular mediators of
transforming growth factor b signals. SMAD2, which is a
tumor suppressor involved in colorectal and lung cancer, has
been shown to induce dorsal mesoderm in Xenopus laevis in
response to transforming growth factor b and activins. The
smad2 gene is expressed ubiquitously during murine embryogenesis and in many adult mouse tissues. Animals that lacked
smad2 died before 8.5 days of development (E8.5). E6.5
homozygous mutants were smaller than controls, lacked the
extraembryonic portion of the egg cylinder, and appeared
strikingly similar to E6.5 smad4 mutants. This similarity was
no longer evident at E7.5, however, because the smad2 mutants
contained embryonic ectoderm within their interiors. Molecular analysis showed that smad2 mutant embryos did not
undergo gastrulation or make mesoderm. The results demonstrate that smad2 is required for egg cylinder elongation,
gastrulation, and mesoderm induction.
The transforming growth factor b (TGFb) superfamily consists of numerous related cytokines, including TGFb itself,
activins, bone morphogenic proteins (BMPs), and others (reviewed in refs. 1 and 2), which perform a variety of biological
functions. The ligands of the TGFb superfamily transmit their
signal through a number of related transmembrane seriney
threonine kinases. These receptors are heteromeric, with a
type I and a type II subunit required for signal transduction.
Upon ligand binding the type II receptor phosphorylates the
type I receptor, which then transmits the signal to downstream
targets (3–5).
The recent discovery of the highly conserved vertebrate
smad genes, smad1–9 (6–16), has advanced our understanding
of the downstream signaling cascade utilized by the type I
receptors. The products of the smad genes are phosphorylated
by the type I receptors and transmit the TGFb signal to the
nucleus, where they participate in the activation of downstream
genes (reviewed in refs. 17 and 18).
The SMAD proteins consist of a highly conserved aminoterminal MH1 domain and a carboxyl-terminal MH2 domain,
which are separated by a proline-rich linker region (6) and are
found as homotrimers in solution (19). The type I receptors
phosphorylate the SMAD proteins at a highly conserved
SS(VyM)S motif located at the carboxyl terminus of the MH2
domain (20, 21). However, SMAD1–SMAD5 are pathway
specific, in that they cannot be phosphorylated by every
receptor. SMAD1 and SMAD5 mediate signals in the BMP
pathway (6, 8, 21), whereas SMAD2 and SMAD3 convey
signals originated by TGFb and activins (7, 8, 11, 14, 20). Upon
MATERIALS AND METHODS
Targeting Vector. An 1130-bp smad2 cDNA sequence was
amplified from brain cDNA by using primers (59-GTGTGGATTGTTACCTTTGG-39) and (59-CATGAATACTACGACGGAGG-39) and used to screen a mouse genomic library
(Stratagene). A 2.5-kb EcoRIyBamHI fragment containing
the most 39 exon of smad2 was subcloned into the EcoRI site
of ploxP (29) to create pMW110. A 7-kb AspIyBamHI fragment containing upstream smad2 sequences was subcloned
into pBSKS (Stratagene). This subclone was digested with AspI
and treated with the Klenow fragment of DNA polymerase,
and the smad2 sequences were released with NotI. This
fragment was cloned into NotIyHpaI-digested pMW110 to
create psmad2neo (Fig. 1A).
Electroporation of ES Cells and Generation of Germ-Line
Chimeras. psmad2neo was digested with NotI and electroporated into TC1 ES cells as described (30). Genomic DNAs were
isolated, digested with BglII, and electrophoresed for Southern
blotting. Targeted ES cells were microinjected into C57BLy6
blastulae to generate chimeras, which were mated to NIH
Black Swiss females (Taconic Farms). Half of the agouti
progeny from these matings carried the smad2DC allele.
Genotype Analysis. Mice were genotyped either by Southern
blotting as above or by PCR. For PCR analysis the mutant
allele was amplified by using a primer from the smad2 sequences that were not deleted in the targeting (59CATGAATACTACGACGGAGG-39) and a primer from neo
(59-ATCGCCT TCTATCGCCT TCT TGACGAGT TC-39),
which amplify a 600-bp product from the mutant allele. Mutant
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.
Abbreviations: TGFb, transforming growth factor b; BMP, bone
morphogenic protein; En, day n of development; ES, embryonic stem.
*To whom reprint requests should be addressed. e-mail: ChuxiaD@
bdg10.niddk.nih.gov.
0027-8424y98y959378-6$0.00y0
PNAS is available online at www.pnas.org.
9378
Developmental Biology: Weinstein et al.
Proc. Natl. Acad. Sci. USA 95 (1998)
9379
Table 1. Genotypic and phenotypic analysis of embryos derived
from smad2DCy1 crosses
No. with phenotype
Age
Normal
E6.5
17 (17)
E7.5
40 (34)
E8.5
30 (30)
E9.5
15 (15)
Total 102
P21
50
No. with genotype
Abnormal
Resorbtions*
1y1
1y2
2y2
3 (3)
5 (4)
5 (5)
0
13
0
1
5
8
3
17
—
6
8
4
7
25
20
11
26
26
8
71
30
3
4
5
0
12
0
Numbers inside parentheses are the number of embryos genotyped.
P21, postnatal day 21 (weaning).
*Resorbtions were not subjected to genotypic analysis.
FIG. 1. (A) Disruption of smad2 through homologous recombination in embryonic stem (ES) cells. The targeting vector psmad2neo
deletes the carboxyl-terminal portion of smad2. The top line is the
targeting vector, the middle the genomic locus before recombination,
and the bottom the locus after recombination. 59 is on the left. The size
and position of exons are not drawn to scale. The arrow represents the
direction of neo transcription, the dashed lines the portion deleted, and
the short black lines the 39 flanking and 59 internal probes used to
detect homologous recombination. (B) Southern blot of ES DNAs.
The 39 flanking probe recognizes a 6-kb BglII fragment from the
wild-type allele; this fragment is reduced to 4.3 kb in the targeted allele
because of the presence of a BglII site on the neo gene. Correct
targeting was confirmed with a HindIIIyEcoRV digest, which releases
a 9-kb band from the wild type and a 5-kb band from the targeted
allele. The 59 internal probe was used to further confirm correct
targeting (data not shown). Both correctly targeted clones are shown.
Hi, HindIII; No, NotI; Bg, BglII; Ba, BamHI; RI, EcoRI; and RV,
EcoRV.
smad2 carboxyl-terminal domain is therefore expected to
create a null allele, which we refer to as smad2DC.
The psmad2neo vector was electroporated into TC1 ES (35),
and 2 of the 179 G418yFIAU-resistant clones analyzed were
found to be correctly targeted (Fig. 1B) [FIAU is 1-(29-deoxy29f luoro-b-D-arabinofuranosyl)-5-iodouracil]. Both clones
were injected into blastocysts and both resulted in germ-line
transmission of the smad2DC allele. Southern blots and PCR
analysis showed that the mutation was passed on to agouti
offspring of the chimeras (data not shown).
The smad2DC Mutation Is a Recessive Embryonic Lethal.
Mice heterozygous for the smad2DC mutation are viable,
healthy, and fertile. They are indistinguishable from their
wild-type siblings in growth rates and litter sizes. They have not
shown any tendency to develop tumors during a 9-month study
period, although an increased rate of cancer could require
more time to become evident.
To address the effect of a loss of smad2 function on murine
development, smad2DC/1 animals were bred inter se to yield
offspring homozygous for the disrupted allele. However, no
homozygotes were found in the 50 offspring genotyped at
weaning (Table 1), suggesting that the mutation resulted in
recessive embryonic lethality.
alleles were also identified by primers that amplified an
internal segment of the neo gene. For the wild-type allele, the
smad2 primer from above was used in conjunction with
primers from the region deleted in the smad2DC mutation.
These included (59-CTCCTTGATGGATGAACTTC-39),
which amplifies a 150-bp product, and (59-GGACCAGACTCACTAGTTCA-39), which amplifies a 300-bp product.
Histological Analysis and in Situ Hybridization. Histology
and in situ hybridization were carried out by using standard
procedures. The probes used included T (30), lim1 (31), fgf8
(32), bmp4 (a kind gift of C. Chang, Vanderbilt Univ., Nashville, TN), and H19 (33).
RESULTS
Disruption of smad2 by Gene Targeting. The targeting
vector psmad2neo (Fig. 1 A) contains 9.5 kb of genomic smad2
sequences with a 2.5-kb deletion, into which we have placed a
PGKneo cassette (34). This deletion removes the carboxylterminal 86 amino acids from the SMAD2 protein, which itself
contains 487 amino acids. Homologous recombination between the targeting vector and the endogenous locus will result
in loss of the phosphorylation site and approximately half of
the MH2 domain. Phosphorylation of SMAD2 by the type I
TGFb receptors has been shown to be required for nuclear
localization and signal transduction activity (11, 20), and
phosphorylation-deficient mutants of smad2 have been found
associated with colorectal tumors (11). The deletion of the
FIG. 2. Morphological analysis of smad2DC/DC embryos (arrows)
and their littermate controls. (A) E6.5 embryos. Arrowhead points to
the boundary between extraembryonic and embryonic portions of the
control embryo. The mutant embryo is round without such a boundary. (B) E7.5 embryos. Mutant embryos remain small and do not
contain any normal structures such as headfold (hf) and primitive
streak (ps). The ectoplacental cone (epc) is normal in the mutants. (C
and D) E8.5 mutant (C) and control (D) embryos. Note that the
mutant embryos exhibit some variation in size. However, even the
largest mutant embryo (right in C) is much smaller than its littermate
control (D). (Bar, 20 mm for A and B and 40 mm for C and D.)
9380
Developmental Biology: Weinstein et al.
The timing of this lethality was determined by examining
embryos at several stages of development, as shown in Table
1. Abnormal or resorbed embryos were found between E6.5
and E8.5, and PCR analysis showed all were homozygous for
the deleted smad2 sequences (data not shown). This suggests
that the deletion of the smad2 carboxyl-terminal domain
caused abnormal development and embryonic death.
E6.5 smad2DC/DC embryos were much smaller than their
sibling controls (Fig. 2A) with a pronounced reduction in the
extraembryonic portion of the egg cylinders. Indeed they bore
a superficial resemblance to embryos lacking smad4 (29, 36).
By E7.5 normal embryos had developed readily recognizable structures, such as the ectoplacental cone, primitive
streak, and headfold (Fig. 2B). smad2DC/DC embryos either
were resorbed (Table 1) or did not form a distinguishable
headfold or primitive streak. They did exhibit an increase in
size over E6.5 smad2DC/DC embryos, although they were still
diminutive compared with their normal siblings. Interestingly,
the ectoplacental cone appeared normal in E7.5 smad2DC/DC
embryos (Fig. 2B), as it did in embryos lacking smad4 (29, 36).
By E8.5 many of the smad2DC/DC embryos were either
resorbed or in the process of resorption (Table 1). The
Proc. Natl. Acad. Sci. USA 95 (1998)
remainder were poorly organized and extremely small (Fig.
2C). Some of the smad2DC/DC mutants formed small embryos
within larger membranous sacs (embryo on the right in Fig.
2C), whereas the wild-type and smad2DC/1 embryos have
formed a visible embryonic axis, a defined head, somites, and
other embryonic structures (Fig. 2D and not shown). These
results clearly demonstrate that smad2 is required for early
embryonic development, and its loss results in early postimplantation lethality.
Intact decidua of E6.5–E8.5 embryos were subjected to
histological sectioning to better understand the defect present
in the smad2DC/DC mutants. At E6.5 normal embryos undergo
a process of rapid cell division and elongation to form the egg
cylinder (Fig. 3A). A clear boundary was seen between the
embryonic and extraembryonic portions (arrow in Fig. 3A),
and we often detected mesoderm involuting at this stage.
Cuboidal cells of the visceral endoderm could be seen encompassing the entire structure, although they flattened out at the
distal end of the egg cylinder (Fig. 3A). In contrast, the E6.5
smad2DC/DC embryos lacked extraembryonic ectoderm and
extraembryonic endoderm, and the ectoplacental cone was
FIG. 3. Histological sections of embryos generated from crosses between smad2DC/1 mice. (A and B) Sagittal section of E6.5 control (A) and
mutant (B) embryos. Arrows point to the boundary between extraembryonic and embryonic portions of the embryo. epc, ectoplacental cone; ec,
embryonic ectoderm; ve, visceral endoderm; xec, extraembryonic ectoderm; xee, extraembryonic endoderm. (C and D) E7.5 control (C) and mutant
(D) embryos. The embryonic ectoderm of the mutant embryo grew significantly, whereas no normal structures were visible. al, allantois; am,
amnion; ch, chorion. The arrow in D points to an infolding of embryonic ectoderm. (E and F). E8.5 control (E) and mutant (F) embryos. In mutant
embryos, the presumptive embryonic ectoderm continues to grow without forming any structures seen in normal embryos such as heart (ht) and
neural tube (nt). (Bar, 200 mm for A and B, 260 mm for C, 130 mm for D, 970 mm for E, and 110 mm for F.)
Developmental Biology: Weinstein et al.
found directly adjacent to the epiblast (Fig. 3B). At this stage
smad2DC/DC and smad4ex8/ex8 embryos bore a striking similarity
(29, 36), suggesting that SMAD2 and SMAD4 are required to
function together for egg cylinder elongation. In contrast to
what was found in embryos lacking smad4 (36), smad2DC/DC
embryos appear to have a well developed visceral endoderm
surrounding the epiblast, although the cuboidal cell architecture is maintained around the circumference of the embryo.
This result suggests that other SMAD proteins may synergize
with SMAD4 to direct the development of this lineage.
At E7.5 normal embryos had a well defined mesodermal
layer. The allantois, chorion, and amnion were clearly visible
(Fig. 3C). Those smad2DC/DC embryos that were not resorbed
at this stage either lacked morphologically distinguishable
chorion, amnion, and allantois (Fig. 3D) or exhibited a diminished extaembryonic portion (not shown). The visceral
endoderm and ectoderm were visible, but a mesodermal cell
layer could not be seen in the smad2DC/DC embryos (Fig. 3D).
The smad2DC/DC embryos were enlarged compared with their
counterparts at E6.5, however, and in sharp contrast to embryos lacking smad4, smad2DC/DC embryos exhibited infolding
of the embryonic ectoderm (arrow in Fig. 3D) that in most
embryos appeared as aggregates of cells filling the proamniotic
cavity (not shown). In this regard embryos resemble those
deficient in NODAL, a TGFb homologue that itself is necessary for proper gastrulation (26–28). Reichert’s membrane
and parietal endoderm also appeared normal in the
smad2DC/DC embryos (not shown), despite the other defects
seen. E8.5 animals had many structures recognizable in sections, such as the heart and the neural tube (Fig. 3E). However,
E8.5 smad2DC/DC embryos lacked any recognizable structure or
Proc. Natl. Acad. Sci. USA 95 (1998)
9381
organization and closely resembled E7.5 smad2DC/DC embryos
(Fig. 3F). Endodermal and ectodermal layers are evident, cells
fill the embryonic interior, and no mesoderm can be seen.
Reichert’s membrane and the parietal endoderm still appeared normal in a number of cases, however. We therefore
believe that the smad2DC/DC animals arrest at E7.5 without
undergoing gastrulation.
smad2DC/DC Embryos Exhibited a Defect in the Extraembryonic Portion of the Egg Cylinder. Both the morphological
and histological analyses suggested that the E6.5 smad2DC/DC
embryos had a defect in the extraembryonic portion of the egg
cylinder. We therefore characterized the extraembryonic lineages by examining expression of the H19 gene, which marks
the extraembryonic endoderm and ectoderm, the ectoplacental cone, and the trophoblastic giant cells. A clear boundary
can be seen between the epiblast and extraembryonic portion
in normal E6.5 embryos (Fig. 4 A and C), and the extraembryonic endoderm and the extraembryonic ectoderm are well
organized. In contrast the smad2DC/DC embryos exhibited
abundant staining of H19 in the ectoplacental cone, but not in
any extraembryonic ectoderm or extraembryonic endoderm
(Fig. 4 B and D). The visceral endoderm was clearly of uniform
thickness around the mutant embryo (arrowhead in Fig. 4D),
unlike the normal embryos, in which the visceral endoderm
flattened out at the distal tip of the egg cylinder (arrowhead
in Fig. 4C). This observation may reflect a developmental
delay, as the visceral endoderm is of uniform thickness at E5.5,
or it could result from the failure of egg cylinder elongation.
At E7.5 the smad2DC/DC embryos exhibited H19 staining in the
ectoplacental cone and a small extraembryonic membrane,
which itself was lacking in many mutants (not shown). The
FIG. 4. In situ analysis of smad2DC/DC embryos at E6.5. A, B, E, and F are bright-field views, and the four others are dark-field views. (A–D)
H19 expression in control (A and C) and mutant (B and D) embryos. H19 marks the ectoplacental cone (epc), extraembryonic endoderm (xee)
and ectoderm (xec), and visceral endoderm (ve). The H19 expression domain in the E6.5 mutant is smaller than that of the control, and the
extraembryonic ectoderm and extraembryonic endoderm appear to be missing. Arrows in A–D point to the boundary between extraembryonic and
embryonic portions, and arrowheads in C and D point out the visceral endoderm, which is flattened at the distal tip of the control embryo (C),
but remains thickened in the mutant (D). T expression is shown in normal (E and G) and mutant (F and H) embryos. T labels involuting mesodermal
cells (arrow in G), which are not in the smad2DC/DC mutants. (Bar, 258 mm for A–D and 180 mm for E–H.)
9382
Developmental Biology: Weinstein et al.
Proc. Natl. Acad. Sci. USA 95 (1998)
against E7.5 and E8.5 mutants, which similarly lacked T
staining (n 5 5, data not shown). In addition, we performed
whole mount in situ hybridization against both early and late
mesodermal markers (T and lim1, respectively). E7.0 and E7.5
smad2DC/DC embryos lacked staining for T (n 5 6, Fig. 5A), and
E8.5 embryos were not labeled with T or lim1 (data not
shown). We also failed to detect expression of bmp4, a marker
of extraembryonic mesoderm, in the E7.0 and E7.5 smad2DC/DC
embryos (n 5 6, Fig. 5B). Similarly, fgf8, which is found in the
primitive streak, was not seen in the mutant embryos at E7.5
(n 5 3, Fig. 5C).
DISCUSSION
FIG. 5. smad2DC/DC mutants do not make mesoderm. Whole mount
in situ hybridization against markers of embryonic lineages. (A) T in E7.5
embryos. The primitive streak (ps) can be found in the normal embryo
(Left) but not in the mutant embryo (Right). epc, ectoplacental cone.
(B) bmp4 in E7.5 embryos. The arrow points to the allantois of the
normal embryo (Left), whereas the chorion and amnion are less visible.
The three mutant embryos shown on the Right fail to exhibit staining
for this marker. (C) fgf8 in E7.5 embryos. The primitive streak
(arrowhead) is stained in the normal embryo (Left). Note the lack of
staining in the two mutant embryos on the Right. (Bar 5 458 mm for
A and C and 523 mm for B.)
failure of egg cylinder elongation was also seen in embryos that
lacked smad4; however, the defect in the smad2DC/DC embryos
was considerably less severe, as the visceral endoderm appears
less severely affected, and the ectoderm is capable of further
development.
smad2DC/DC Embryos Fail to Form Mesoderm. We used in
situ hybridization to examine mesodermal induction in the
smad2DC/DC mutants. Litters from smad2DC/1 crosses at E6.5
were sectioned serially and labeled with the widely used
mesodermal marker T. Normal embryos exhibited involuting
mesoderm at E6.5 (Fig. 4 E and G), whereas E6.5 smad2DC/DC
embryos (n 5 6) showed no detectable T staining (Fig. 4 F and
H). We also performed serial section in situ hybridizations
smad2 Functions in Egg Cylinder Elongation and Mesoderm
Induction. SMAD2 is necessary for egg cylinder elongation,
and in its absence the extraembryonic portion of the egg
cylinder is not formed. Embryos that lack smad2 fail to
undergo normal gastrulation, and they become arrested without inducing any mesoderm. Our present level of analysis is
insufficient to determine which of these defects is primary.
Failure of mesodermal induction could result from the absence
of the extraembryonic endoderm and extraembryonic ectoderm; conversely, the failure of egg cylinder elongation may
result from an absence of mesoderm. It is also possible that the
two phenotypes result independently.
smad2 mutants have been recently observed with defects
considerably different from those communicated here (37).
Specifically, the embryos examined by Waldrip et al. (37)
survived 1 day longer than the smad2DC/DC mutants, were able
to form relatively normal extraembryonic membranes, and
induced mesoderm, as determined by expression of T and
bmp4. Interestingly, Waldrip et al. demonstrated an essential
function of smad2 in the extraembryonic membranes for
establishment of the anterioposterior embryonic axis.
We believe that the disruption introduced in Waldrip et al.
(37) may not have resulted in a complete loss of function.
Although neither our study nor that of Waldrip et al. has
demonstrated the generation of a null mutation, the targeting
reported here deleted the carboxyl terminus, which is known
to be essential for SMAD2 activity (20, 21). We have also
shown that homozygosity for this deletion results in embryonic
lethality earlier than that seen by Waldrip et al. (37).
The mutation introduced by Waldrip et al. (37) abolished the
translation start site; however, translation could have started
from a downstream methionine codon in SMAD2, creating a
product truncated in the amino-terminal domain. This truncation would actually be expected to increase the activity of the
protein, as a similar protein exhibited an increased ability to
induce mesoderm, morphogenetic cell movements, and axis
duplication in Xenopus (25), probably because of the loss of an
inhibitory domain in the amino terminus (38). The mutation
examined by Waldrip et al. (37) may have left sufficient smad2
activity to allow for formation of the extraembryonic membranes, but not enough to pattern the embryo. Production of
SMAD2 protein in the mutants was not analyzed by Waldrip
et al.
The smad2 mutants reported here have shown no evidence
of normal egg cylinder formation, and we have found no
expression of either T or bmp4 at any embryonic stage.
Moreover, deletions of either the amino-terminal or the
carboxyl-terminal domains of smad2 generated in another
laboratory resulted in a phenotype similar to that shown here
(E. Li, personal communication).
smad2 and smad4 Function Cooperatively in Egg Cylinder
Elongation. We believe that SMAD2 and SMAD4 cooperate
to convey signals necessary for egg cylinder elongation, as
these two mutants both display this unusual phenotype. However the smad2DC/DC embryos are less affected than the smad4
mutants generated in our laboratory (29), as proliferating cell
Developmental Biology: Weinstein et al.
nuclear antigen (PCNA) staining and blastocyst culture experiments indicate they do not suffer the severe reduction in
cell proliferation experienced by smad4 mutants (data not
shown).
smad2 Is Necessary for Mesoderm Induction. None of the
markers used in this study, including T, lim1, bmp4, and fgf8,
detected mesoderm in the smad2DC/DC embryos. These observations indicate that smad2 is essential for mesoderm induction, correlating well with earlier studies showing that SMAD2
was able to induce dorsal mesoderm in Xenopus (8, 11, 25).
Data from our laboratory suggest that smad2DC/DC ES cells can
form mesodermal derivatives (not shown), similar to cells that
lack smad4 (36). SMAD2 may be needed to generate or receive
inductive signals, or the failure of mesodermal induction may
be a nonspecific consequence of the abnormal embryonic
architecture.
smad2 May Mediate Signals from Several TGFb Family
Members. E7.5 smad2DC/DC embryos do resemble nodal mutants, in that they both accumulate ectoderm within the
embryonic interior due to folding of the ectoderm (26). nodal
mutants are far less affected than smad2DC/DC embryos however, as the former have an overproliferation of embryonic
ectoderm and trophectoderm cells, and they are able to make
some mesodermal cells (27). The smad2DC/DC mutation may
abrogate nodal signals, but there are other TGFb signals that
are terminated as well.
Several genes within the TGFb signal transduction pathway
function in the formation of the extraembryonic membranes.
The smad4 gene has been alluded to earlier, but in addition
mice lacking either the type I BMP receptor (BmpR1) or the
type I activin receptor (ActRIB) also fail to make normal
extraembryonic membranes or mesoderm (39, 40). It is unlikely that smad2 is involved in the BMP pathway, as it has been
shown both biochemically and functionally to be specifically
activated by the receptors for TGFb and activin, and not by the
BMP receptors (8, 11, 22). SMAD2 is expected to transmit the
signal generated by ActRIB, however, and it is noteworthy that
mutants lacking ActRIB do not form an organized extraembryonic epithelium, do not induce mesoderm, and exhibit
marked similarities to smad2DC/DC mutants (40).
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
Kingsley, D. M. (1994) Genes Dev. 8, 133–146.
Massague, J. (1996) Cell 85, 947–950.
Wrana, J. L., Attisano, L., Carcamo, J., Zentella, A., Doody, J.,
Laiho, M., Wang, X. F. & Massague, J. (1992) Cell 71, 1003–1014.
Letsou, A., Arora, K., Wrana, J. L., Simin, K., Twombly, V.,
Jamal, J., Staehling-Hampton, K., Hoffmann, F. M., Gelbart,
W. M., Massague, J., et al. (1995) Cell 80, 899–908.
Ruberte, E., Marty, T., Nellen, D., Affolter, M. & Basler, K.
(1995) Cell 80, 889–897.
Liu, F., Hata, A., Baker, J. C., Doody, J., Carcamo, J., Harland,
R. M. & Massague, J. (1996) Nature (London) 381, 620–623.
Zhang, Y., Feng, X., We, R. & Derynck, R. (1996) Nature
(London) 383, 168–172.
Graff, J. M., Bansal, A. & Melton, D. A. (1996) Cell 85, 479–487.
Hoodless, P. A., Haerry, T., Abdollah, S., Stapleton, M.,
O’Connor, M. B., Attisano, L. & Wrana, J. L. (1996) Cell 85,
489–500.
Hahn, S. A., Schutte, M., Hoque, A. T., Moskaluk, C. A., da
Costa, L. T., Rozenblum, E., Weinstein, C. L., Fischer, A., Yeo,
C. J., Hruban, R. H. & Kern, S. E. (1996) Science 271, 350–353.
Eppert, K., Scherer, S. W., Ozcelik, H., Pirone, R., Hoodless, P.,
Kim, H., Tsui, L. C., Bapat, B., Gallinger, S., Andrulis, I. L., et
al. (1996) Cell 86, 543–552.
Proc. Natl. Acad. Sci. USA 95 (1998)
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
9383
Imamura, T., Takase, M., Nishihara, A., Oeda, E., Hanai, J.,
Kawabata, M. & Miyazono, K. (1997) Nature (London) 389,
622–626.
Hayashi, H., Abdollah, S., Qiu, Y., Cai, J., Xu, Y. Y., Grinnell,
B. W., Richardson, M. A., Topper, J. N., Gimbrone, M. A., Jr.,
Wrana, J. L. & Falb, D. (1997) Cell 89, 1165–1173.
Nakao, A., Roijer, E., Imamura, T., Souchelnytskyi, S., Stenman,
G., Heldin, C. H. & ten Dijke, P. (1997) J. Biol. Chem. 272,
2896–2900.
Chen, Y., Bhushan, A. & Vale, W. (1997) Proc. Natl. Acad. Sci.
USA 94, 12938–12943.
Watanabe, T. K., Suzuki, M., Omori, Y., Hishigaki, H., Horie, M.,
Kanemoto, N., Fujiwara, T., Nakamura, Y. & Takahashi, E.
(1997) Genomics 42, 446–451.
Heldin, C. H., Miyazono, K. & ten Dijke, P. (1997) Nature
(London) 390, 465–471.
Baker, J. C. & Harland, R. M. (1997) Curr. Opin. Genet. Dev. 7,
467–473.
Shi, Y., Hata, A., Lo, R. S., Massague, J. & Pavletich, N. P. (1997)
Nature (London) 388, 87–93.
Macias-Silva, M., Abdollah, S., Hoodless, P. A., Pirone, R.,
Attisano, L. & Wrana, J. L. (1996) Cell 87, 1215–1224.
Kretzschmar, M., Liu, F., Hata, A., Doody, J. & Massague, J.
(1997) Genes Dev. 11, 984–995.
Lagna, G., Hata, A., Hemmati-Brivanlou, A. & Massague, J.
(1996) Nature (London) 383, 832–836.
Hata, A., Lagna, G., Massague, J. & Hemmati-Brivanlou, A.
(1998) Genes Dev. 12, 186–197.
Nakao, A., Afrakhte, M., Moren, A., Nakayama, T., Christian,
J. L., Heuchel, R., Itoh, S., Kawabata, M., Heldin, N. E., Heldin,
C. H. & ten Dijke, P. (1997) Nature (London) 389, 631–635.
Baker, J. C. & Harland, R. M. (1996) Genes Dev. 10, 1880–1889.
Iannaconne, P. M., Zhou, X., Khoka, M., Boucher, D. & Kuehn,
M. R. (1992) Dev. Dyn. 194, 198–208.
Conlon, F. L., Lyons, K. M., Takaesu, N., Barth, K., Kispert, A.,
Herrmann, B. & Robertson, E. J. (1994) Development (Cambridge, U.K.) 120, 1919–1928.
Varlet, I., Collignon, J. & Robertson, E. J. (1997) Development
(Cambridge, U.K.) 124, 1033–1044.
Yang, X., Li, C., Xu, X. & Deng, C. (1998) Proc. Natl. Acad. Sci.
USA 95, 3667–3672.
Deng, C., Wynshaw-Boris, A., Shen, M. M., Daugherty, C.,
Ornitz, D. M. & Leder, P. (1994) Genes Dev. 8, 3045–3057.
Barnes, J. D., Crosby, J. L., Jones, C. M., Wright, C. V. E. &
Hogan, B. L. M. (1994) Dev. Biol. 161, 168–178.
Heikinheimo, M., Lawshe, A., Shackleford, G. M., Wilson, D. B.
& MacArthur, C. A. (1994) Mech. Dev. 48, 129–138.
Poirier, F., Chan, C. T., Timmons, P. M., Robertson, E. J., Evans,
M. J. & Rigby, P. W. (1991) Development (Cambridge, U.K.) 113,
1105–1114.
Tybulewicz, V. L., Crawford, C. E., Jackson, P. K., Bronson, R. T.
& Mulligan, R. C. (1991) Cell 65, 1153–1163.
Deng, C., Wynshaw-Boris, A., Zhou, F., Kuo, A. & Leder, P.
(1996) Cell 84, 911–921.
Sirard, C., de la Pompa, J. L., Elia, A., Itie, A., Mirtsos, C.,
Cheung, A., Hahn, S., Wakeham, A., Schwartz, L., Kern, S. E., et
al. (1998) Genes Dev. 12, 107–119.
Waldrip, W. R., Bikoff, E. K., Hoodless, P. A., Wrana, J. L. &
Robertson, E. J. (1998) Cell 92, 797–808.
Hata, A., Lo, R. S., Wotton, D., Lagna, G. & Massague, J. (1997)
Nature (London) 388, 82–87.
Mishina, Y., Suzuki, A., Ueno, N. & Behringer, R. R. (1995)
Genes Dev. 9, 3027–3037.
Gu, Z., Nomura, M., Simpson, B. B., Lei, H., Feijen, A., van den
Eijnden-van Raaij, J., Donahoe, P. K. & Li, E. (1998) Genes Dev.
12, 844–857.