August 15, 2001;
Endoderm specification and differentiation in Xenopus embryos.
It is known from work with amniote embryos that regional specification of the gut
requires cell-cell signalling between the mesoderm
and the endoderm
. In recent years, much of the interest in Xenopus endoderm
development has focused on events that occur before gastrulation and this work has led to a different model whereby regional specification of the endoderm
is autonomous. In this paper, we examine the specification and differentiation of the endoderm
in Xenopus using neurula
-bud-stage embryos and we show that the current hypothesis of stable autonomous regional specification is not correct. When the endoderm
is isolated alone from neurula
and tail bud
stages, it remains fully viable but will not express markers of regional specification or differentiation. If mesoderm
is present, regional markers are expressed. If recombinations are made between mesoderm
, then the endodermal markers expressed have the regional character of the mesoderm
. Previous results with vegetal explants had shown that endodermal differentiation occurs cell-autonomously, in the absence of mesoderm
. We have repeated these experiments and have found that the explants do in fact show some expression of mesoderm
markers associated with lateral
plate derivatives. We believe that the formation of mesoderm
cells by the vegetal explants accounts for the apparent autonomous development of the endoderm
. Since the fate map of the Xenopus gut
shows that the mesoderm
of each level do not come together until tail bud
stages, we conclude that stable regional specification of the endoderm
must occur quite late, and as a result of inductive signals from the mesoderm
[+] show captions
FIG. 1. Model for the development of the digestive tract occurs in three stages. (1) Formation. Establishment of endodermal cell fate occurs
early in embryogenesis and is autonomous. This observation comes from recent work in Xenopus showing that localized transcription
factors will activate the transcription of Edd and that loss of maternal VegT leads to a complete loss of endoderm. (2) Regional specification.
Once cell fate is established, the mesoderm then acts to pattern the endoderm. Results from mouse, chick, frog, and fly embryos have
shown that specification of the endoderm depends on the character of the overlying mesoderm. Since different mesoderm can respecify the
endoderm, patterning in the endoderm is labile until the third stage. (3) Differentiation. By the last stage, cell fate has become determined
irreversibly. This is characterized by the synthesis of mRNAs specific for a particular function. (Adapted from Horb, 2000.)
FIG. 2. Schematic diagram illustrating how the explants were made. (A) Neurula stage 15 embryo. The bold line indicates where the initial
dissection was made to remove the neural plate, cement gland region, and blastopore. This isolated fragment is called whole endoderm plus
mesoderm (WEM). The dashed line indicates where the WEM was bisected to generate the anterior (AEM) and posterior (PEM) explants. (B)
Tail bud stage 25 embryo. As before, the bold line indicates where the head, spinal cord, and tail regions were dissected away. (C) Stage 15
whole endoderm (WE) explant. The WE explants elongate much more than the other explants. Notice, too, that there are no recognizable
organ structures. (D) Stage 15 anterior endoderm (AE) explant. Unlike the WE explants, the AE explants remain as round balls and do not
elongate. (E) Stage 15 posterior endoderm (PE) explant. These explants elongate much more than the AE explants.
IG. 3. Expression pattern analysis of endodermal markers in stage 15 explants by whole-mount in situ hybridization. (A–E) Xlhbox8 expression.
Notice that expression of Xlhbox8 is only seen in WEM and AEM explants, but not in PEM or WE explants. In isolated whole guts at stage 42,
Xlhbox8 is expressed in the pancreas and the stomach. (F–J) IFABP expression. IFABP is expressed in WEM, AEM, and PEM explants, but not in
WE explants. Also, IFABP is expressed in both AEM and PEM explants, as predicted by the fate map. In isolated whole guts, IFABP is expressed
from the base of the stomach through the middle of the intestine. No IFABP expression is detected in the posterior intestine. (K–O) Xcad2
expression. Xcad2 is also expressed in WEM, AEM, and PEM explants. In contrast to IFABP, however, Xcad2 is expressed more highly in PEM
explants than in AEM explants. In isolated whole guts, Xcad2 is expressed from the beginning of the small intestine through to the proctodeum.
(P–T) Edd expression. Edd is expressed in all four explants. It is not expressed throughout the AEM explants, but is expressed throughout the PEM
explants. In contrast to the other markers, Edd is expressed throughout the WE explants. In isolated whole guts at this stage, expression has become
downregulated in the pancreas or stomach.
FIG. 4. RT-PCR analysis of stage 15 explants cultured until stage
42. cDNA was generated from five explants and used to analyze the
expression of several endodermal differentiation markers by PCR.
None of the markers tested, including Xlhbox8, LFABP, Xhex,
IFABP, and Xcad2, are expressed in the isolated whole endoderm.
Xlhbox8 and LFABP are only expressed in the AEM explants, but
not, as expected, in the PEM explants. Xhex, IFABP, and Xcad2 are
expressed in both AEM and PEM explants. In contrast to the other
markers, Edd expression is increased in the WE explants compared
to the others. This agrees with our assertion that the WE explants
have not undergone regional specification, as Edd is normally lost
from the pancreas and stomach.
FIG. 5. Mesodermal marker expression in neurula stage 15 explants. (A) XNkx-2.5, FoxF1, XTbx5, and xFOG are not detected in the whole
endoderm explants. All of these markers are, however, expressed in the explants with mesoderm to various degrees. Notice that XNkx-2.5
is expressed most strongly in the AEM explants. This is in agreement with its endogenous expression pattern in the mesoderm surrounding
the duodenum. (B) Neither cardiac actin, a-T3 globin, nor Xtwist is expressed in the whole endoderm explants. Again, expression of each
of these markers is seen in the explants with the mesoderm to varying degrees. (C) To confirm that the markers used to detect the presence
of mesoderm are indeed expressed in the mesoderm, we examined their expression in explants of isolated anterior or posterior mesoderm
from stage-25 embryos.
FIG. 6. Temporal expression of endodermal differentiation markers
in whole embryos. Whole embryos were collected at various
stages from 10.5 to 40. In agreement with previous results, a low
level of Xlhbox8 expression is detectable at stages 20 and 25, but
there is a substantial increase in expression beginning at stage 30.
Other markers are first expressed between stages 25 and 35.
FIG. 7. RT-PCR analysis of vegetal explants. The explants were isolated at late blastula stage 9 and grown in vitro until stage 42. (A)
Endodermal marker expression. Xlhbox8 and IFABP, but not Xcad2, are expressed. This suggests that posterior endoderm does not become
specified in these vegetal explants. (B) Mesodermal marker expression. All four of the mesodermal markers tested here show expression in
the vegetal explants: XNkx-2.5, FoxF1, XTbx5, and xFOG. (C) Expression of previously used mesodermal markers cardiac actin, globin, and
Xtwist is absent in the vegetal explants. This confirms that our explants were dissected properly. (D, E) In situ hybridization for FoxF1 (D)
and nonexpressed control gene (E) showing scattered positive cells expressing FoxF1 in the vegetal explant. (F) Mesodermal marker
expression in mixer-injected animal caps. Mixer-injected animal caps express both XTbx5 and xFOG, but not FoxF1. Uninjected animal
caps do not express any of these three markers.
FIG. 8. RT-PCR expression of endodermal markers in mesoderm–endoderm recombinants. (A) Xlhbox8 is only expressed in the PEAM
recombinants, and barely in the AEPM recombinants. Conversely, Xcad2 is only expressed in the AEPM recombinants. IFABP is, as
expected, expressed in both recombinants. The regional specificity is characteristic of the mesoderm and not the endoderm. (B) To confirm
that our mesodermal explants were themselves free from contaminating endoderm, we examined the expression of Edd in anterior
endoderm (AE), anterior mesoderm (AM), posterior endoderm (PE), and posterior mesoderm (PM) explants. No expression of Edd is seen in
the mesodermal explants, while abundant expression of Edd is seen in the endodermal explants.