Supplementary
Fig. 1 Differential expression pattern of COUP-TFII using knockin lacZ reporter as a
marker in the vasculature.
(a-d). Using COUP-TFII/lacZ knock-in mice, we examined the expression pattern of
COUP-TFII in the vascular tissue at E10.5 (a and b) and E11.5 (c and d). COUP-TFII
is expressed in the endothelial cells of the anterior cardinal vein (ACV) (a, arrows),
umbilical vein (UV) (c, arrows) and vitelline vein (VV) (d, arrows), but not in the
endothelial cells of dorsal aorta (DA) (a, arrows), internal carotid artery (ICA) (b,
arrows), and umbilical artery (UA) (c, arrows). However, COUP-TFII is highly
expressed in the smooth muscle cell layer surrounding the DA (a), ICA (b) and UA
(c).
Fig. 2 Model of cell autonomous versus non-cell autonomous functions.
To generate wild type and COUP-TFII null ES cells containing the ROSA26 allele,
COUP-TFII+/- mice were first crossed with the ROSA26 line1 to generate COUPTFII+/-; ROSA26/+ compound heterozygote. The compound heterozygote was then
crossed to COUP-TFII+/- mice to obtain blastocysts from dated pregnant females. The
stable ES cell lines with COUP-TFII+/+; ROSA26/+ and COUP-TFII-/-; ROSA26/+
genotypes were generated from the isolated blastocysts. To analyze the cellautonomous verse non cell-autonomous function of COUP-TFII, the ES cells were
then microinjected into the wild type blastocyst (shown in pink) to generate chimeric
embryos2, 3. The cells derived from the ES cells can be identified by X-Gal staining
(shown in blue). The mutant cells will not be able to contribute to the tissues where
they are required cell-autonomously. In contrast, if mutant cells play a non cell-
autonomous function in the tissues, they will be able to contribute as well as the wild
type cells.
Fig. 3 Endothelial-specific knockout of COUP-TFII by Tie2-Cre.
(a) Generation of floxed COUP-TFII and COUP-TFII/LacZ knock-in mice. Using
homologous recombination strategy, the targeting vector was inserted into genomic
COUP-TFII locus. Upon Cre- mediated recombination, ES cells with floxed COUPTFII and LacZ knock-in alleles were generated and used to create mice. Floxed
COUP-TFII mice can be used for tissue specific knockout using tissue specific Cre
lines. Upon Cre-mediated recombination between these two loxP sites, the expression
of lacZ, containing the nuclear localization signal, will be turned on under the control
of COUP-TFII promoter, permitting the determination of the recombination event. In
addition, the LacZ knock-in mice will facilitate the monitoring of the cell specific
expression pattern of COUP-TFII during development.
(b-d) Whole embryo X-Gal staining showed the Cre-mediated recombination by
crossing Tie2-Cre with either R26R reporter mice or floxed COUP-TFII mice at E9.5.
X-Gal staining was apparent in all endothelium of reporter mice subsequent to Tie2Cre crossing with R26R reporter mice (c), but not the control littermate (b). Only the
venous endothelium was marked by X-Gal staining in case of Tie2-Cre/+; COUPTFIIflox/+ embryos (d). It is notable that the staining present in dorsal aorta and
brachial arch arteries of the Tie2-Cre/+; R26R/+ reporter embryo (c) was missing in
Tie2-Cre/+; COUP-TFIIflox/+ embryos as indicated by arrows (d).
(e-h) Cre- mediated recombination in Tie2-Cre/+; R26R/+ and Tie2-Cre/+; COUPTFIIflox/+ embryos at E9.5. At cardinal vein level, the endothelial-specific excision
visualized by X-Gal staining was observed in all endothelial cells in Tie2-Cre/+;
R26R/+ reporter embryos (e), while Cre-mediated recombination was only seen in
venous endothelium in Tie2-Cre/+; COUP-TFIIflox/+ embryos (f). Similarly, X-Gal
positively stained cells were seen at both arterial and venous endothelium in Tie2Cre/+; R26R/+ reporter embryos (g), whereas the X-Gal positive signals were absent
from arterial endothelium in Tie2-Cre/+; COUP-TFIIflox/+ embryos (h) at umbilical
arteries (UA) and veins (UV) region. Since the inserted LacZ gene containing a
nuclear localization signal in floxed COUP-TFII allele, the appearance of the nuclear
X-Gal staining pattern is different from cytoplasmic staining pattern of R26R reporter
line. A, aorta; V, vein.
Fig. 4 The endothelial-specific COUP-TFII null mutants exhibited a variety of
vascular defects. (a and b) The endothelial-specific COUP-TFII null mutants die in
utero by E12. Mutant embryos at E11.5 show widespread hemorrhage (b) compared
with the littermate (a). (c and d) The major histopathological features of mutants who
survived at E11.5 show thin and well-dilated vessels. Transverse sections of the E11.5
embryos were stained by hematoxylin and eosin. The dilated vessels at the posterior
cardinal vessel level were observed in the mutant embryo (d) but not in the control
embryo (c). Also noted is the cell layers that separated the artery and vein (indicated
by arrows) are greatly reduced in the mutant (d) as compared to the littermate control
(c). A, aorta; V, vein.
Fig. 5 The hematopoietic-specific markers in hematopoietic cell clusters of the
mutant.
The identities of these hematopoietic cell clusters from mutant veins were further
confirmed using anti-CD45 (a, b) and anti-c-kit (CD117) (c) antibodies in
immunostaining. CD45 antigen is a tyrosine phosphatase, which is presented on the
hematopoietic cells. Cells arised from the head vein (a) and umbilical vein (b) of the
mutant are round in appearance and express CD45 (brown), indicated that they are
hematopoietic cells. As c-Kit had been shown to express on both hematopoietic stem
cells (HSC) and endothelial cells (EC)4. The positive signal (brown) was observed in
these two compartments in the mutant umbilical vein (c).
Supplemental References
1. Friedrich, G. & Soriano, P. Promoter traps in embryonic stem cells: a genetic
screen to identify and mutate developmental genes in mice. Genes Dev. 5, 15131523 (1991).
2. Bradley, A., Evans, M., Kaufman, M. H. & Robertson, E. Formation of germ-line
chimaeras from embryo-derived teratocarcinoma cell lines. Nature 309, 255-256
(1984).
3. Nagy, A. et al. Embryonic stem cells alone are able to support fetal development in
the mouse. Development 110, 815-821 (1990).
4. Bernex, F. et al. Spatial and temporal patterns of c-kit-expressing cells in WlacZ/+
and WlacZ/WlacZ mouse embryos. Development 122, 3023-3033 (1996).