Download Expression of a truncated Sall1 transcriptional

Human Molecular Genetics, 2003, Vol. 12, No. 17
DOI: 10.1093/hmg/ddg233
2221–2227
Expression of a truncated Sall1 transcriptional
repressor is responsible for Townes–Brocks
syndrome birth defects
Susan McLeskey Kiefer1, Kevin K. Ohlemiller2, Jing Yang1, Bradley W. McDill1,
Jürgen Kohlhase3 and Michael Rauchman1,*
1
Renal Division, School of Medicine, Washington University, St Louis and 2Central Institute for the Deaf,
St Louis, MO 63110, USA and 3Institut für Humangenetik der Universität Göttingen, Gosslerstrasse 12d,
D-37073 Göttingen, Germany
Received May 19, 2003; Revised June 27, 2003; Accepted July 5, 2003
Townes–Brocks syndrome (TBS, OMIM #107480) is an autosomal dominant disorder that causes multiple
birth defects including renal, ear, anal and limb malformations. Mutations in SALL1 have been postulated to
cause TBS by haploinsufficiency; however, a mouse model carrying a sall1-null allele does not mimic the
human syndrome. Since the mutations that cause TBS could express a truncated SALL1 protein containing
the domain necessary for transcriptional repression but lacking the complete DNA binding domain, we
hypothesized that TBS is due to dominant-negative or gain-of-function activity of a mutant protein. To test
this hypothesis, we have created a mutant allele, sall1-DZn2-10, that produces a truncated protein and
recapitulates the abnormalities found in human TBS. Heterozygous mice mimic TBS patients by displaying
high-frequency sensorineural hearing loss, renal cystic hypoplasia and wrist bone abnormalities.
Homozygous sall1-DZn2-10 mutant mice exhibit more severe defects than sall1-null mice including complete
renal agenesis, exencephaly, limb and anal deformities. We demonstrate that truncated Sall1 mediates
interaction with all Sall family members and could interfere with the normal function of all Sall proteins.
These data support a model for the pathogenesis of TBS in which expression of a truncated SALL1 protein
causes abnormal development of multiple organs.
INTRODUCTION
The role of transcriptional repression as an essential counterpart to gene activation has been recently recognized for many
processes including embryonic development. The zinc finger
transcription factor, Sall1, has been shown to repress gene
expression (1,2) and play a key role in organogenesis (3). The
Drosophila sall1 orthologues are critical for proper development of the fly eye, wing, trachea and antenna (4–7), and
mutations in human SALL1 result in the developmental defects
of the Townes–Brocks syndrome (TBS) (8). TBS patients
exhibit malformed ears and hearing loss, imperforate anus,
limb defects, and absent or hypoplastic kidneys. It has been
postulated that TBS is due to haploinsufficiency of SALL1
(8,9); however, a null-allele of mouse sall1 does not mimic
the human syndrome (3), suggesting the possibility of an
alternative genetic mechanism. In support of this idea, mutations that cause TBS could result in production of a truncated
protein encoding the N-terminus of SALL1, the region that
we have shown mediates transcriptional repression (1).
We thus postulated that expression of a truncated SALL1
protein is responsible for the developmental defects observed in
TBS. In this report we describe a sall1 mutant allele, sall1DZn2-10, designed to mimic a mutation shown to cause TBS.
This mutant expresses a truncated Sall1 protein and displays
a striking recapitulation of TBS phenotypes. We further
demonstrate that truncated Sall1 protein can mediate biochemical interactions with all wild-type Sall family members
and thereby may interfere with their normal function. Together,
these data provide genetic evidence supporting a dominantnegative or gain-of-function mechanism in the pathogenesis
of TBS.
*To whom correspondence should be addressed at: Department of Medicine and Biochemistry and Molecular Biology, Saint Louis University and
Veterans Affairs Medical Center, Renal Division 657/111B-JC, 915 N Grand Avenue, St Louis, MO 63106, USA. Tel: þ1 3142896485; Email:
[email protected]
Human Molecular Genetics, Vol. 12, No. 17 # Oxford University Press 2003; all rights reserved
2222
Human Molecular Genetics, 2003, Vol. 12, No. 17
RESULTS
Sall1 repression domain correlates with TBS mutations
The mutations that cause TBS are clustered in the N-terminal
third of the SALL1 coding region and could produce truncated
proteins encoding only the N-terminus of SALL1, a domain
that mediates transcriptional repression (1). Strikingly, the most
N-terminal mutation that has been shown to cause TBS (9),
419delC (Fig. 1A), would encode a truncated protein consisting
of the first 140 amino acids of human SALL1 that includes the
130 amino acid domain that is both necessary and sufficient for
transcriptional repression by mouse Sall1 (Fig. 1B). When
expressed as a fusion protein with a heterologous DNA binding
domain (GAL4DB), the first 130 amino acids of mouse Sall1
and a longer 435 amino acid truncated protein were capable of
repressing transcription over 100-fold from a GAL4DBresponsive luciferase reporter. Analogous constructs expressed
as GST-fusion proteins interacted with a previously characterized (1) repression complex containing histone deacetylases
(HDACs) (Fig. 1C). Removal of the first 130 amino acids
abrogated both repression and interaction with the complex and
mutation of the N-terminal C2HC zinc finger did not affect
either. Because TBS mutations correlated with this important
functional domain, we postulated that the pathogenesis of TBS
is not due to haploinsufficiency of SALL1 and sought to test
whether the expression of a Sall1 truncated protein could
recapitulate TBS abnormalities in mice.
Generation of sall1-DZn2-10 mice expressing truncated
Sall1 protein
To test this hypothesis, we generated mice carrying a targeted
mutant allele (Fig. 2A), sall1-DZn2-10, to mimic one mutation
shown to cause TBS (Fig. 1, 1277–1278delGA). After
confirming each genotype by Southern blot (Fig. 2B), we
analyzed embryonic lysates for the presence of Sall1 wild-type
and truncated proteins (Fig. 2C). Western blotting with an
N-terminal antibody revealed that a truncated protein of the
expected size (Sall1-DZn2-10) was robustly expressed from
the mutant allele (Fig. 2C) in all embryonic tissues affected in
TBS patients (data not shown). Concomitant loss of expression
of the wild-type Sall1 protein confirmed disruption of the
sall1 locus by our targeting strategy. Thus, the sall1-DZn2-10
allele produces a truncated protein that has the potential to
exert dominant-negative or gain-of-function effects during
embryogenesis.
Nishinakumura et al. (3) have shown that sall1-null mice
(sall1/) are born at the expected Mendelian frequency and
die soon after birth from renal failure. In contrast, no live
homozygous sall1-DZn2-10 pups (sall1D/D) are born. The
distribution of genotypes at embryonic day 14 to 16 (E14–
E16) [60 sall1þ/þ (31%), 114 sall1þ/D (58%), 21 sall1D/D
(11%)] showed that half of the sall1D/D embryos die during
gestation prior to E14. Those that survived until E14 exhibited
a variety of defects including renal agenesis (100%, Fig. 3B),
exencephaly (38%, Fig. 3D), limb abnormalities (57%)
including syndactlyly (Fig. 3F) or missing digits (Fig. 4G)
and imperforate anus (>50%, Fig. 4C). Removal of Neo by
crossing to b-actin cre transgenic mice had no effect on the
phenotypes observed (data not shown). Because sall1-DZn2-10
mice exhibit more severe and multi-organ defects not seen in
sall1-null mice, we next sought to investigate in more detail
sall1-DZn2-10 heterozygous and homozygous mice for abnormalities that mimic TBS.
sall1-DZn2-10 mice mimic TBS developmental defects
Similar to 90% of TBS patients who are diagnosed with hearing loss (9), sall1þ/D mice were found to have a high frequency
hearing loss when compared to wild-type littermates (Fig. 4A).
Differences in thresholds for auditory evoked brain-stem
responses were consistent with a high frequency sensorineural
hearing deficit and a few individual mice also exhibited a mild
conductive deficit (data not shown). This pattern of hearing loss
is strikingly comparable to TBS patients who have been
documented to have higher threshold responses at highfrequency with a small variable conductive component (10).
sall1-DZn2-10 homozygous mice exhibit defects in formation
of their anus. Over 50% of homozygous embryos exhibit the
most serious TBS ano-rectal defect, imperforate anus (Fig. 4B
and C). While wild-type embryos exhibited an evident external
anal opening, the rectum of the mutant appeared to fuse with
the urogenital system. Histologic sections (insets 1–3) of
dissected bladder (b), urethra (u) and rectum (r) revealed that
the mutant ano-rectal region fused with the urethra to form a
common lumen for the urogenital and digestive systems. This
type of ano-rectal malformation shares characteristics with
hedgehog and Gli mutants (11). Since the Drosophila and
Medaka sal genes have been shown to be activated by
hedgehog signalling, overlapping anal phenotypes may reflect
action of these molecules in a common signalling pathway
(5,12). Heterozygous sall1-DZn2-10 mice exhibit a lethal, low
penetrance (5%) gastrointestinal phenotype manifested as
bowel obstruction at 3–6 weeks of age (data not shown).
This abnormality may be the counterpart of a suspected bowel
motility disorder manifested by severe constipation of some
TBS patients (J. Kohlhase, unpublished data).
The third birth defect that is most commonly observed in
TBS patients is limb malformations. Skeletal staining revealed
the absence of the triquetral bone in the wrist in all sall1þ/D
adult limbs (Fig. 4D and E). The triquetrum, normally present
on the dorsal side of the wrist (open arrowhead), was
completely absent in all sall1þ/D mice. In its absence, the
underlying pisiform (solid arrowhead) was fully visible from
the dorsal aspect. In addition, the hamate (asterisk) did not
adopt its characteristic triangular appearance to extend to the
distal side of the 5th metacarpal. These abnormalities of the
carpal bones are identical to those described in the original
patient reports by Townes and Brocks (13). Additional skeletal
abnormalities described for TBS patients were also seen in
sall1D/D mutant embryos where 12 out of 21 (57%) exhibited
gross limb abnormalities, including an absent first digit
(Fig. 4G) or syndactyly of digits 2 and 3 (Fig. 3F) affecting
one or more limbs. Skeletal staining revealed more severe
wrist defects including the absence of multiple wrist bones,
metacarpal hypoplasia and fusion (Fig. 4G, digits 4 and 5).
Analogous defects described for the forelimbs were also
found in the hindlimbs of sall1þ/D and sall1D/D mice (data
not shown).
Human Molecular Genetics, 2003, Vol. 12, No. 17
2223
Figure 1. Truncated proteins produced by TBS mutations would include the Sall1 transcriptional repression and HDAC interaction domains. (A) Schematic of the
mouse Sall1 N-terminus with protein domains noted [C2HC-type zinc finger (oval), threonine-rich (substituted by a serine-rich domain in human SALL1), glutamine-rich, and serine-rich]. The length of the truncated protein produced by TBS mutations is indicated [a, 419delC (9); b, 817delG (27); c, 840delC (28);
d, 826C > T (9); e, 1115C > A (8); f, 1146delT (9); g, 1200–1206del7 (9); h, 1277–1278delGA (9)]. (B) The first 130 amino acids domain of mouse Sall1 is
necessary and sufficient for repression. GAL4DB-Sall1-N-terminal constructs were expressed in cells with a reporter in which luciferase gene transcription is under
the control of five GAL4DB binding sites. Normalized luciferase activity of each construct was divided by the activity of cells transfected with GAL4DB alone to
obtain fold repression. Values are plotted as the mean SD of triplicate transfection from two independent experiments. (C) The same 130 amino acid domain is
necessary and sufficient for interaction with HDAC1 and 2. The Sall1-N constructs used for the repression assays were constructed as GST-fusions and overexpressed in cells. Glutathione pulldown experiments demonstrated an interaction with endogenous HDAC1 and 2 for only those contructs that exhibit repression
activity.
Approximately 60% of TBS patients with documented
SALL1 mutations have been reported to have kidney defects,
most commonly renal hypoplasia with or without cysts (9).
These renal abnormalities are recapitulated in the sall1-DZn2-10
mouse because at least 25% of sall1þ/D mice were found to
have hypoplastic, cystic kidneys (Fig. 4H and I). Numerous
small and, on occasion, large cysts covered the kidney surface
of some sall1þ/D adult mice and were never observed in wildtype littermates. These macroscopic cysts were confirmed by
histology that revealed dilated renal tubules and mimicked
documented renal abnormalities of TBS patients (14). We also
observed isolated renal hypoplasia without cystic changes in
both adult and embryonic sall1þ/D mice indicating that
improper kidney development led to reduced kidney size (data
not shown).
Truncated Sall1 interacts with all Sall family members
Comparison of the multi-organ defects of sall1-DZn2-10 mice
and the exclusively renal phenotype of sall1-null mice suggests
that the truncated Sall1 protein acts in a dominant-negative
or gain-of-function fashion. Since Sall family members are
co-expressed in many developing tissues and are thought to
be functionally redundant (4,15), it has been suggested that the
truncated Sall1 protein may interfere with the function of
wild-type Sall proteins through direct interaction (1,16). In
support of this mechanism, we found that the N-terminal
protein expressed by sall1-DZn2-10 mice can bind specifically
with all four Sall proteins, but not with an unrelated zinc finger
transcriptional repressor, ZEB1 (Fig. 5). Full-length Sall1 also
interacts with all wild-type Sall family proteins, suggesting
that homo- and heteromeric complexes are an integral part of
normal Sall function. Sall1 and Sall2 exhibit a genetic
interaction in mutant mice (J. Kohlhase, unpublished data),
substantiating the in vivo relevance of this functional interaction. These data support a model in which the Sall1 truncated
protein could interfere with any wild-type Sall protein in a
tissue where they are co-expressed.
DISCUSSION
Molecular genetic studies of congenital anomaly syndromes
have shown that many are due to haploinsufficiency of
transcription factors (17). For example, Renal Coloboma
syndrome and Aniridia have been shown genetically to arise
from haploinsufficiency of the paired box transcription factors
PAX2 and PAX6 (18,19). SALL1 mutations have been suggested
to cause TBS by the same paradigm, although a sall1-null
mouse does not mimic the human syndrome. Recent in vitro
studies have raised the possibility that expression of a truncated
protein rather than haploinsufficiency causes TBS (1,16), and
in this report we provide in vivo genetic evidence to confirm
this hypothesis. Because the phenotypes of sall1-DZn2-10 mice
are more severe than sall1-null mice and recapitulate the multiorgan abnormalities that define TBS, we conclude that this
syndrome is caused by expression of a truncated Sall1 protein
exerting severe effects on developing organs.
Truncated Sall1 protein could act either in a dominantnegative or gain-of-function fashion. We show that truncated
Sall1 protein produced by the sall1-DZn2-10 mutant allele can
2224
Human Molecular Genetics, 2003, Vol. 12, No. 17
Figure 2. sall1 gene disruption. (A) Schematic of the sall1 locus, the targeting vector and the sall1-DZn2-10 mutant allele (X, Xho1; H, HindIII). Zinc finger
domains (white ovals) and clustered TBS mutations that could result in a truncated protein are indicated (bracket). The targeting vector introduced a stop codon
(stop), LoxP sites (gray triangles) and a neomycin cassette (Neo) after this region. The probe and predicted HindIII restriction fragments for Southern analysis are
shown. (B) Southern analysis of embryos from a sall1þ/D intercross by HindIII digestion of embryo tail DNA. (C) Western analysis of embryo limb lysates show
loss of the wild-type protein and production of a truncated protein (Sall1-DZn2-10) from the mutant allele.
interact with all Sall proteins and therefore has the potential to
exert deleterious effects on all Sall functions in the cell
(Fig. 5A). The glutamine-rich domain that has been shown to
be critical for Sall1–Sall3 interaction (16) is present in the
protein expressed by sall1-DZn2-10 mice, but would be absent
in the shortest protein that could be expressed in a TBS patient
(9). Preliminary studies suggest that this shorter protein can
also interact with other Sall family members, albeit at much
lower affinity (data not shown), but it remains to be elucidated
if this level of interaction is physiologically relevant.
Interaction of truncated Sall1 could prevent other Sall proteins
from compensating for loss of wild-type Sall1 or interfere with
their function in tissues where Sall proteins are co-expressed
but subserve unique roles. An analogous paradigm where
expression of a truncated protein disrupts development has
been demonstrated for the zinc finger transcription factors
Ikaros and Gli3 (20–22). Recently, it has been shown that
Okihiro syndrome is also caused by truncating mutations in the
SALL4 gene (23,24), implying a related mechanism.
Interestingly, the phenotypic spectrum of TBS and Okihiro
overlap, thus, we suggest that expression of a truncated SALL4
protein may exert similar activity to cause some of the same
developmental defects in both syndromes.
Previously, we have demonstrated that the N-terminus of
Sall1 mediates transcriptional repression by recruitment of an
HDAC-containing complex resembling NuRD (1). Here, we
have further mapped this domain and show that the first 130
amino acids of Sall1 are necessary and sufficient to mediate
repression (Fig. 1A). Based on the human SALL1 mutations
reported to date (9), the shortest protein that could be expressed
in TBS patients corresponds to this minimal repression domain.
Therefore, we hypothesize that the truncated protein could be
affecting repression of target genes. In sall1-DZn2-10 mice, the
truncated protein is present in both the nucleus and cytoplasm
(data not shown), however it lacks the zinc fingers thought
to interact with DNA. The interaction of truncated Sall1
with full-length Sall proteins could provide a mechanism by
which the truncated protein is brought to DNA to directly
alter gene regulation. Alternatively, heterodimerization of
truncated Sall1 with wild-type Sall family members could
prevent normal Sall regulation of target genes by affecting
cellular localization (16) or DNA binding. Finally, the ability
of truncated Sall1 to interact with an HDAC-containing
co-repressor complex (1) suggests the possibility of more
widespread gain-of-function effects on gene expression by
sequestering common co-factors, such as HDACs, that are
utilized by many transcription factors.
Analysis of sall1-DZn2-10 mice provides insight into the
function of sall genes in organogenesis and may identify TBS
defects that may be under-reported or unrecognized clinically.
Human Molecular Genetics, 2003, Vol. 12, No. 17
2225
Figure 3. Phenotypes of sall1-DZn2-10 homozygous embryos. (A and B) Renal
agenesis. Dissection of the bladder (b), gonads (g), kidneys (k) and adrenals (a)
from E16 sall1þ/þ (A) and sall1D/D (B) embryos shows that 100% (n ¼ 21) of
sall1D/D embryos lack kidneys. (C and D) Exencephaly. Thirty-eight percent of
sall1D/D (D) E14–E16 embryos (n ¼ 8) develop a neural tube closure defect that
is not present in sall1þ/þ (C) or sall1þ/ littermates. (E and F) Syndactyly.
Forelimbs from E15 embryos show a normal appearance of the digits for
sall1D/þ (E) limbs and fusion of digits 2 and 3 (arrowheads) for a sall1D/D
(F) littermate. Fifty-seven percent of limbs from sall1D/D embryos (n ¼ 12) have
a gross limb defect such as syndactyly or oligodactyly.
For instance, bowel obstruction observed in some sall1þ/D
mice (data not shown) suggests an intestinal motility
disorder that is manifested as severe chronic constipation in
some TBS patients (J. Kohlhase, unpublished data) and
as Hirschsprung’s disease in Okihiro’s syndrome (25). The
severe limb and CNS defects observed in sall1-DZn2-10
homozygous mutants highlight a previously unrecognized
essential role of sall genes in formation of the carpel/tarsal
elements and in neural tube closure. Consistent with our results
implicating sall1 in neural tube defects, TBS patients have been
recently identified with type I Arnold–Chiari malformation
(J. Kohlhase, unpublished data). In addition to recapitulating
the defects seen in TBS patients, anomalies seen in sall1DZn2-10 mice such as hearing loss, imperforate anus and
hypoplastic kidneys, represent major human congenital
malformations that occur sporadically. The sall1 mutant mouse
described here is an ideal animal model to investigate the
function of sall genes in these important developmental
processes.
In summary, these studies establish sall1-DZn2-10 mice as a
strikingly faithful mouse model of TBS and provide new
insights into the pathogenesis of this disorder. Further study of
this mouse mutant will uncover additional roles of sall genes as
critical developmental regulators, give insights into their
mechanism of action individually and in combination, and
will serve as a powerful tool to identify genes acting in a
common pathway to control vertebrate organogenesis.
Figure 4. Phenotypes of sall1-DZn2-10 mice mimic TBS abnormalities. (A)
Hearing defects. Audiometric thresholds in decibel levels of sall1þ/D mice
(open square) are higher (P < 0.001, 2-way ANOVA) than wild-type littermates
(filled square) and suggest a high frequency sensorineural defect. (B and C)
Anal defects. Normal appearance of sall1þ/þ and fusion of the urethra (u) with
the rectum (r) for sall1D/D embryos (arrowheads). H&E stained sections (insets
1–3) confirm a recto-urethral fistula in the mutant. (D and E) Limb defects in
sall1þ/D mice. Skeletal analysis of left forelimbs (dorsal view) reveal an absent
triquetrum (open arrowhead, missing in E) and a misshapen hamate (*) in
sall1þ/D mice. (F and G) Limb defects of sall1D/D mice. Skeletal staining of
E16 limbs of wild type (F) and sall1D/D (G) embryos show that the mutant
embryo exhibits several absent wrist bones, fusion of the 4th and 5th metacarpal bones and an absent thumb. (H and I) Kidney defects. Some sall1þ/D
kidneys were cystic and hypoplastic and show dilated renal tubules.
MATERIALS AND METHODS
Plasmid construction
Sall1 N-terminus (amino acids 2–435, Sall1-DZn2-10), mutated
Zn (amino acids 2–435 with the C2HC zinc finger mutated to
TSHC), N-(130–435), N-(1–130) and full-length Sall1 were
2226
Human Molecular Genetics, 2003, Vol. 12, No. 17
Disruption of sall1
Figure 5. Specific interaction of Sall1 with all Sall proteins. GST-fusion
proteins representing the N-terminal Sall1 protein expressed in sall1-DZn2–10
mice (GST-Sall1-DZn2-10) and full-length Sall1 (GST-Sall1) were co-expressed
in COS-1 cells with epitope-tagged Sall proteins or an unrelated zinc finger
transcriptional repressor ZEB1. Pulldown experiments revealed that GSTSall1-DZn2–10 and GST-Sall1 were capable of binding all four Sall proteins
but not ZEB1. GST alone did not bind.
constructed as GAL4DB- and GST-fusion proteins. Sall3 and
Sall4 were cloned based on published sequence information
and Sall2 (T. Benjamin, Harvard) and ZEB1 (D. Dean,
Washington University) were generously donated. The Sall2
sequence was modified to encode the exon 1 splice variant and
all Sall sequences were modified with a C-terminal epitope tag.
Repression assays
Briefly, cells were transiently transfected with GAL4DB-Sall1
fusion plasmid, a GAL4DB-responsive luciferase reporter and
a plasmid containing constitutively expressed beta galactosidase to normalize for transfection efficiency. After 48 h, the
cells were assayed for luciferase activity exactly as described
previously (1). The normalized luciferase activity of each
GAL4DB-Sall1 construct was compared to the activity of cells
transfected with GAL4DB alone to calculate fold repression.
Each construct was transfected in triplicate wells and two
independent assays were averaged.
129/SvJ-derived RW-4 embryonic stem cells were electroporated with a vector designed to substitute a stop codon and a
floxed neomycin cassette for nucleotides 1300–3443 of the
mouse sall1 cDNA within exon 2 of the sall1 locus. One
homologous recombinant clone was used to establish chimeric
founders that were mated to outbred mice to generate mice that
carried the sall1-DZn2-10 allele in a mixed 129-SvJ/ICR
outbred background. Wild-type and mutant alleles were
detected by PCR genotyping and Southern blotting by
established methods. Embryonic limb protein extracts were
prepared using RIPA buffer containing inhibitors (1 mg/ml
leupeptin, 2 mg/ml antipain, 10 mg/ml benzamidine, 1 mg/ml
chymostatin, 1 mg/ml pepstatin, 24 mg/ml Pefabloc, 20 mM NaF
and 2 mM sodium molybdate) and western blotted with an
antibody directed against the N-terminus of Sall1 (1). All
experiments were performed under the auspices of the
Institutional Animal Studies Committee.
Hearing tests
Auditory evoked brain-stem responses were recorded in 5week-old F1 sall1þ/D mice anesthetized with IP injection of
80 mg/kg ketamine, 15 mg/kg xylazine. Platinum needle
electrodes were inserted subcutaneously just behind the right
ear (active), at the vertex (reference) and in the back (ground)
and used to record bioelectric potentials in response to 5 ms
tonebursts emanating from a speaker located 7 cm lateral to the
right ear, concentric with the external auditory meatus. Stimuli
were presented freefield and calibrated using a microphone
placed where the pinna would normally be. Responses were
filtered (0.1–10 kHz), digitized at 30 kHz, and averaged over
1000 stimulus presentations (20/s). The minimum sound
pressure level required for a response (short-latency negative
wave) was determined at 5, 10, 20, 28.3 and 40 kHz, using a
5 dB minimum step size. Seven sall1þ/D and seven sall1þ/þ
mice were tested.
Morphology analysis
Tissues were photographed using a Nikon SMZ800 dissecting
microscope. Skeletal staining was performed as described (26).
Adult kidney and embryo anal regions were paraffin embedded,
sectioned and analyzed by hematosylin/eosin staining.
Interaction assays
For HDAC interaction assays, the same constructs assayed for
repression as GAL4DB fusion proteins were constructed as
GST-Sall1 fusion proteins and overexpressed in COS-1 cells.
After 48–72 h, cells were lysed, precipitated with glutathione
sepharose and analyzed for the presence of endogenous HDAC1
and 2 by western blot using commercially available antibodies
as described previously (1). For Sall–Sall interaction studies,
GST, GST-Sall1-N (Sall1-DZn2-10) and GST-Sall1 were overexpressed with either epitope-tagged Sall1, Sall2, Sall3 or
Sall4 in COS-1 cells. Lysates were analyzed for the presence of
overexpressed epitope-tagged Sall proteins by western blot
as above.
ACKNOWLEDGEMENTS
We thank the Washington University Cancer Center Embryonic
Stem Cell core, Mouse Genetics core, and the Molecular
Biology and Pharmacology Morphology core for invaluable
assistance, and Raphael Kopan, Sandy McLeskey, David
Ornitz, Heiko Peters, YiPing Chen and Jim Kiefer for critical
reading of the manuscript. The authors wish to thank the
following funding agencies for support: NIH Training Grant
5T32DK07126 (S.M.K.), Deutsche Forschungsgemeinschaft
KO 1850/3-2 (J.K.) and Edward Mallinckrodt Foundation
(K.K.O.).
Human Molecular Genetics, 2003, Vol. 12, No. 17
REFERENCES
1. Kiefer, S.M., McDill, B.W., Yang, J. and Rauchman, M. (2002) Murine
Sall1 represses transcription by recruiting a histone deacetylase complex.
J. Biol. Chem., 277, 14869–14876.
2. Netzer, C., Rieger, L., Brero, A., Zhang, C.D., Hinzke, M., Kohlhase, J. and
Bohlander, S.K. (2001) SALL1, the gene mutated in Townes–Brocks
syndrome, encodes a transcriptional repressor which interacts with
TRF1/PIN2 and localizes to pericentromeric heterochromatin. Hum. Mol.
Genet., 10, 3017–3024.
3. Nishinakamura, R., Matsumoto, Y., Nakao, K., Nakamura, K., Sato, A.,
Copeland, N.G., Gilbert, D.J., Jenkins, N.A., Scully, S., Lacey, D.L. et al.
(2001) Murine homolog of SALL1 is essential for ureteric bud
invasion in kidney development. Development, 128, 3105–3115.
4. Mollereau, B., Dominguez, M., Webel, R., Colley, N.J., Keung, B.,
de Celis, J.F. and Desplan, C. (2001) Two-step process for photoreceptor
formation in Drosophila. Nature, 412, 911–913.
5. de Celis, J.F., Barrio, R. and Kafatos, F.C. (1996) A gene complex acting
downstream of dpp in Drosophila wing morphogenesis. Nature, 381,
421–424.
6. Franch-Marro, X. and Casanova, J. (2002) Spalt-induced specification of
distinct dorsal and ventral domains is required for Drosophila tracheal
patterning. Dev. Biol., 250, 374–382.
7. Dong, P.D., Dicks, J.S. and Panganiban, G. (2002) Distal-less and
homothorax regulate multiple targets to pattern the Drosophila antenna.
Development, 129, 1967–1974.
8. Kohlhase, J., Wischermann, A., Reichenbach, H., Froster, U. and Engel,W.
(1998) Mutations in the SALL1 putative transcription factor gene cause
Townes–Brocks syndrome. Nat. Genet., 18, 81–83.
9. Kohlhase, J., Taschner, P.E., Burfeind, P., Pasche, B., Newman, B.,
Blanck, C., Breuning, M.H., ten Kate, L.P., Maaswinkel-Mooy, P.,
Mitulla, B. et al. (1999) Molecular analysis of SALL1 mutations in
Townes–Brocks syndrome. Am. J. Hum. Genet., 64, 435–445.
10. Rossmiller, D.R. and Pasic, T.R. (1994) Hearing loss in Townes–Brocks
syndrome. Otolaryngol. Head Neck Surg., 111, 175–180.
11. Mo, R., Kim, J.H., Zhang, J., Chiang, C., Hui, C.C. and Kim, P.C. (2001)
Anorectal malformations caused by defects in sonic hedgehog signaling.
Am. J. Pathol., 159, 765–774.
12. Koster, R., Stick, R., Loosli, F. and Wittbrodt, J. (1997) Medaka spalt acts
as a target gene of hedgehog signaling. Development, 124, 3147–3156.
13. Townes, P.L. and Brocks, E.R. (1972) Hereditary syndrome of imperforate
anus with hand, foot, and ear anomalies. J. Pediatr., 81, 321–326.
14. Barakat, A.Y., Butler, M.G., Salter, J.E. and Fogo, A. (1988) Townes–
Brocks Syndrome: report of three additional patients with previously
undescribed renal and cardiac abnormalities. Dysmorphol. Clin. Genet.,
2, 104–108.
2227
15. Milan, M., Perez, L. and Cohen, S.M. (2002) Short-range cell interactions
and cell survival in the Drosophila wing. Dev. Cell, 2, 797–805.
16. Sweetman, D., Smith, T., Farrell, E.R., Chantry, A. and Munsterberg, A.
(2002) The conserved glutamine rich region of chick csal1 and csal3
mediates protein interactions with other spalt family members—
implications for Townes–Brocks syndrome. J. Biol. Chem., 278,
6560–6566.
17. Seidman, J.G. and Seidman, C. (2002) Transcription factor
haploinsufficiency: when half a loaf is not enough. J. Clin. Invest.,
109, 451–455.
18. Eccles, M.R., He, S., Legge, M., Kumar, R., Fox, J., Zhou, C., French, M.
and Tsai, R.W. (2002) PAX genes in development and disease: the
role of PAX2 in urogenital tract development. Int. J. Dev. Biol., 46,
535–544.
19. Glaser, T., Cai, J., Epstein, J., Walton, D.S., Jepeal, L. and Maas, R. (1985)
In Wiggs, J.R. (ed.), Molecular Genetics of Eye Disease. John Wiley and
Sons, New York, pp. 51–82.
20. Sun, L., Liu, A. and Georgopoulos, K. (1996) Zinc finger-mediated
protein interactions modulate Ikaros activity, a molecular control of
lymphocyte development. EMBO J., 15, 5358–5369.
21. Georgopoulos, K., Bigby, M., Wang, J.H., Molnar, A., Wu, P., Winandy, S.,
and Sharpe, A. (1994) The Ikaros gene is required for the development
of all lymphoid lineages. Cell, 79, 143–156.
22. Bose, J., Grotewold, L. and Ruther, U. (2002) Pallister-Hall syndrome
phenotype in mice mutant for Gli3. Hum. Mol. Genet., 11, 1129–1135.
23. Al Baradie, R., Yamada, K., St Hilaire, C., Chan, W.M., Andrews, C.,
McIntosh, N., Nakano, M., Martonyi, E.J., Raymond, W.R., Okumura, S.
et al. (2002) Duane Radial Ray Syndrome (Okihiro Syndrome) Maps to
20q13 and Results from Mutations in SALL4, a New Member of the SAL
Family. Am. J. Hum. Genet, 71, 1195–1199.
24. Kohlhase, J., Heinrich, M., Schubert, L., Liebers, M., Kispert, A., Laccone, F.,
Turnpenny, P., Winter, R.M. and Reardon, W. (2002) Okihiro syndrome is
caused by SALL4 mutations. Hum. Mol. Genet, 11, 2979–2987.
25. Okihiro, M.M., Tasaki, T., Nakano, K.K. and Bennett, B.K. (1977) Duane
syndrome and congenital upper-limb anomalies. A familial occurrence.
Arch. Neurol., 34, 174–179.
26. Peters, H., Neubuser, A., Kratochwil, K. and Balling, R. (1998)
Pax9-deficient mice lack pharyngeal pouch derivatives and teeth and
exhibit craniofacial and limb abnormalities. Genes Dev., 12, 2735–2747.
27. Salerno, A., Kohlhase, J. and Kaplan, B.S. (2000) Townes–Brocks
syndrome and renal dysplasia: a novel mutation in the SALL1 gene.
Pediatr. Nephrol., 14, 25–28.
28. Blanck, C., Kohlhase, J., Engels, S., Burfeind, P., Engel, W., Bottani, A.,
Patel, M.S., Kroes, H.Y. and Cobben, J.M. (2000) Three novel SALL1
mutations extend the mutational spectrum in Townes–Brocks syndrome
[letter]. J. Med. Genet., 37, 303–307.