First published online 5 January 2006
doi: 10.1242/dev.02220
Development 133, 569-578 (2006)
Published by The Company of Biologists 2006
GLI3-dependent transcriptional repression of Gli1, Gli2 and kidney patterning genes disrupts renal morphogenesis
Ming Chang Hu1,2,
Rong Mo1,
Sita Bhella1,
Christopher W. Wilson3,
Pao-Tien Chuang3,
Chi-chung Hui1,4 and
Norman D. Rosenblum1,2,5,6,7,*
1 Program in Developmental Biology, Hospital for Sick Children, University of
Toronto, Toronto, Canada.
2 Division of Nephrology, The Hospital for Sick Children, University of Toronto,
Toronto, Canada.
3 Cardiovascular Research Institute, University of California, San Francisco,
CA, USA.
4 Department of Medical and Molecular Genetics, University of Toronto, Toronto,
Canada.
5 Department of Paediatrics, University of Toronto, Toronto, Canada.
6 Department of Physiology, University of Toronto, Toronto, Canada.
7 Department of Laboratory Medicine and Pathobiology, University of Toronto,
Toronto, Canada.
View larger version (90K):
[in a new window]
|
Fig. 1. Deficient SHH-SMO signaling disrupts kidney development.
(A,C,D) Gross anatomical features of kidney development in wild-type
and Shh–/– mice at E18.5. In contrast to
wild-type mice (A), Shh–/– mice exhibited
either absence of both kidneys (C) or the presence of a single ectopic kidney
located in the pelvis (D). (B,E) Renal histological phenotype in E18.5
mice. In contrast to the organized appearance of glomeruli and tubules in the
renal cortex of wild-type mice (B), the single kidney formed in  50% of
Shh–/– mice (E) was characterized by a paucity
of glomeruli and the presence of dilated tubules. Scale bar: 100 µm.
(F-I) Effect of cyclopamine on renal development. Treatment of mice
with cyclopamine starting at E9.5 for four consecutive days blocked renal
development. In contrast to kidneys from mice treated with drug vehicle alone
(F,G), kidneys in cyclopamine-treated mice (H,I) demonstrated a marked
decrease in ureteric bud branches and epithelial metanephric derivatives.
Scale bar: 100 µm. (J,K) Ureteric bud branching in embryonic kidneys
isolated at E11.5 from wild-type mice. Kidneys from the same mouse were
cultured as pairs in the presence of drug vehicle (J, 100% ethanol in culture
medium) or cyclopamine (K) for 4 days. Ureteric bud branches were identified
with Dolichos Biflorus Agglutinin. In contrast to vehicle, cyclopamine
decreased the number of ureteric bud branches formed. Scale bar: 100 µm. G,
gonad; K, kidney.
|
|
View larger version (97K):
[in a new window]
|
Fig. 2. Effects of Shh deficiency or SMO inhibition on GLI protein
expression. (A-L) Immunohistochemistry of 4 µm tissue
sections generated from E14.5 kidney tissue using specific anti-GLI
antibodies. The substrate reaction generated a red color. Tissue was
counterstained with Hematoxylin (blue). In tissue generated from wild-type
mice (A,E,I), GLI1, GLI2 and GLI3 were each detected in metanephric-derived
epithelial structures and ureteric bud branches. Shh deficiency
decreased GLI1 and GLI2 expression compared with wild type (D,H versus A,E).
By contrast, total GLI3 expression in kidney was not decreased in
Shh–/– mice (L). Antibody specificity was
demonstrated by lack of staining in corresponding Gli-deficient mice
(B,C,F,G,J,K). Scale bar: 100 µm. (M) Western analysis of E14.5
kidney tissue lysates generated from wild-type and
Shh–/– mice. Shh deficiency decreased
GLI1 and GLI2. Although GLI3 activator (190 kDa) was decreased, GLI3 repressor
(89 kDa) was unaffected compared with wild type. The ratio of GLI3 activator
to GLI3 repressor in renal tissue was markedly decreased from 3.25 in wild
type to 0.23 in Shh–/– mice. (N)
Western analysis of tissue lysates generated from cultured embryonic kidneys
isolated from E11.5 wild-type mice and treated for 4 days with culture medium
alone or with culture medium supplemented with drug vehicle (100% ethanol),
SHH-N or cyclopamine. In the presence of culture medium, embryonic kidneys
expressed GLI1, GLI2 and GLI3 activator (190 kDa) in excess of GLI3 repressor
(89 kDa) (GLI3 activator: GLI3 repressor=3.68). Treatment with SHN-N increased
GLI1, GLI2 and the relative expression of GLI3 activator versus GLI3 repressor
(GLI3 activator:GLI3 repressor=11.9). Treatment with drug vehicle had no
significant effect on GLI protein expression. By contrast, cyclopamine
decreased GLI1 and GLI2 and the relative expression of GLI3 activator versus
GLI3 repressor (GLI3 activator:GLI3 repressor=1.43).
|
|
View larger version (68K):
[in a new window]
|
Fig. 3. Shh and Gli3 interact to control Pax2 and
Sall1 expression during kidney development. (A-E)
Pax2 mRNA expression in E11.5 mouse embryos. Pax2 mRNA was
identified by whole-mount in situ hybridization using digoxigenin-labeled
probes. In wild-type and Gli3–/– embryos,
Pax2 mRNA is expressed in the Wolffian duct (long arrow) and
metanephros (short arrow). Pax2 mRNA was barely detectable in the
Wolffian duct in 50% of Shh–/– embryos
(compare B with A). Pax2 mRNA was rescued to wild-type levels in all
Shh–/–;Gli3–/– embryos
examined. (F-J) Kidney histology in E14.5 embryos. Tissue sections (4
µm) were stained with Hematoxylin. Wild-type (F),
Gli2–/– (H) and
Gli3–/– (I) mice exhibit two normally
positioned kidneys. A single, ectopic, dysplastic kidney was observed in
Shh deficient mice (G). By contrast, kidney number and histology was
rescued in
Shh–/–;Gli3–/– mice
(J). Scale bar: 200 µm. (K-O) Pax2 mRNA expression in E14.5
mouse kidneys. Pax2 mRNA was detected using a digoxigenin-labeled
probe and in situ hybridization in 4 µm tissue sections. In wild-type (K),
Gli2–/– (M) and
Gli3–/– (N) kidneys, Pax2 mRNA was
expressed in ureteric bud branches and metanephric-derived epithelial
structures. In Shh–/– mice (L), Pax2
mRNA expression was markedly diminished but was rescued in
Shh–/–;Gli3–/– mice
(O). Scale bar: 100 µm. (P-T) Sall1 mRNA expression in
E14.5 mouse kidneys. Sall1 mRNA was detected using a
digoxigenin-labeled probe and in situ hybridization in 4 µm tissue
sections. In wild-type (P), Gli2–/– (R), and
Gli3–/– (S) kidneys, Sall1 mRNA was
expressed in metanephric-derived epithelial structures. In
Shh–/– mice (Q), Sall1 mRNA
expression was markedly diminished but was rescued in
Shh–/–;Gli3–/– mice
(T). Scale bar: 100 µm.
|
|
View larger version (48K):
[in a new window]
|
Fig. 4. Shh and Gli3 interact to control cell proliferation
and expression of cell cycle regulatory proteins. (A-D) In situ
analysis of BrdU incorporation at E14.5. Pregnant mice were injected with BrdU
and sacrificed 4 hours later. BrdU incorporation was detected by an in situ
BrdU incorporation assay. Red-stained nuclei are BrdU positive. Tissues were
counterstained with Hematoxylin (blue). In wild-type (A),
Gli3–/– (C) and
Shh–/–;Gli3–/– (D)
mice, BrdU incorporation was observed in metanephric mesenchyme cells,
mesenchyme-derived epithelial structures and ureteric bud tips. In
Shh–/– mice (B), BrdU incorporation was
markedly decreased. Scale bar: 100 µm. (E) Quantitative analysis of
BrdU incorporation at E14.5. BrdU incorporation is expressed as a fraction of
the total number of cells in the ureteric bud (white bars) and its branches or
in the metanephric mesenchyme (black bars). Shh deficiency decreased
BrdU incorporation in both ureteric bud and mesenchyme-derived epithelial
structures significantly compared with wild type. Removal of Gli3 in
Shh–/– mice rescued BrdU incorporation to
levels not significantly different than those observed in wild type. Removal
of Gli3 increased BrdU incorporation in ureteric bud derived cells to
a level greater than that observed in wild type. (F) Western analysis
of E14.5 kidney tissue lysates. Expression of cyclin D1 and MYCN was markedly
decreased in Shh–/– compared with wild type
and was rescued to wild-type levels in
Shh–/–;Gli3–/– mice. By
contrast, expression of cyclin D2 and MYC was not affected by Shh
deficiency.
|
|
View larger version (37K):
[in a new window]
|
Fig. 5. cyclopamine changes binding of GLI protein species to 5' flanking
regions in Shh target genes. (A) Identification of GLI
consensus binding sequences in the 5' flanking region of mouse Pax2,
Sall1, cyclin D1 and Mycn. Nucleotides represented in upper case
are exact matches to those identified previously in GLI consensus binding
sequences (see text). Arrows indicate promoter segments amplified by PCR
during chromatin immunoprecipitation (see below). (B) PCR products
identified by agarose electrophoresis after chromatin immunoprecipitation
using E11.5 wild-type kidney tissue cultured for 4 days in the presence of
drug vehicle or cyclopamine. In vehicle-treated embryonic kidneys, GLI1 and/or
GLI2 bound 5' flanking regions containing GLI consensus binding regions
in Pax2, Sall1, cyclin D1 and Mycn. No binding of GLI3 with
these regions was detectable. Cyclopamine decreased GLI1 and GLI2 binding most
markedly in Pax2 and Sall1, and induced binding of GLI3 to
each promoter region. Neg, DNA amplified after immunoprecipitation with
non-immune serum; Pos, DNA amplified using genomic DNA and specific primers;
Input, DNA amplified from cross-linked DNA before immunoprecipitation.
|
|
View larger version (52K):
[in a new window]
|
Fig. 6. GLI3 controls GLI1 and GLI2 expression via transcriptional
mechanisms. (A) Western analysis of E14.5 kidney tissue lysates
using specific anti-GLI antibodies. In Shh–/–
mice, expression of GLI1 and GLI2 was decreased compared with wild type and
Gli3–/–. In
Shh–/–;Gli3–/– mice,
GLI1 and GLI2 expression was rescued to wild-type levels. (B) Agarose
gel electrophoresis of products generated by RT-PCR using RNA isolated from
E14.5 embryonic kidney. Expression of Gli1 and Gli2 mRNA was
decreased in Shh–/– compared with wild type
and Gli3–/– and was rescued in
Shh–/–;Gli3–/– mice.
(C) Identification of GLI consensus binding sequences in the 5'
flanking region of mouse Gli1 and Gli2. Nucleotides
represented in upper case are exact matches to those identified previously in
GLI consensus binding sequences (see text). Arrows indicate promoter segments
amplified by PCR during chromatin immunoprecipitation (see below). (D)
PCR products identified by agarose electrophoresis after chromatin
immunoprecipitation using E11.5 wild-type kidney tissue cultured for 4 days in
the presence of drug vehicle or cyclopamine. In vehicle-treated embryonic
kidneys, GLI1 and/or GLI2 bound 5' flanking regions containing GLI
consensus binding regions in Gli1 and Gli2. No binding of
GLI3 with this region in Gli1 was detectable. Cyclopamine decreased
GLI1 and GLI2 binding and induced binding of GLI3 to each promoter region.
Neg, DNA amplified after immunoprecipitation with non-immune serum; Pos, DNA
amplified using genomic DNA and specific primers; Input, DNA amplified from
cross-linked DNA before immunoprecipitation.
|
|
View larger version (29K):
[in a new window]
|
Fig. 7. Gli3 is required for SMO-dependent control of GLI expression
and kidney development. (A) Western analysis of kidney tissue
lysates using GLI-specific antibodies. Embryonic kidneys were harvested from
wild-type or Gli3–/– mice at E11.5 and
cultured in the presence of drug vehicle or cyclopamine for 4 days. Treatment
of kidneys isolated from wild-type mice with cyclopamine markedly decreased
GLI1 and GLI2. This inhibitory effect was totally abrogated in
Gli3-deficient mice. (B) Ureteric bud branching in cultured
kidney explants. Treatment of kidneys isolated from E11.5 wild-type mice with
cyclopamine decreased the number of ureteric bud branches identified by
Dolichos Biflorus Agglutinin. This inhibitory effect was abrogated in
Gli3-deficient mice. (C) Model of GLI3 activity in a state of
SMO inhibition. Decreased SMO activity generated by SHH deficiency or
treatment with cyclopamine increases formation of GLI3 repressor. GLI3
repressor controls expression of Gli1 and Gli2 as well as
SHH target genes that control renal morphogenesis (Pax2 and
Sall1) and cell proliferation (cyclin D1 and Mycn).
|
|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati 
Twitter What's this?
© The Company of Biologists Ltd 2006
