This was explained by the compensatory function of Six1 considering the similar expression pattern of both genes and the func-tional similarity in activating their target gene myo-genin.
Trang 1proteins, Six1 and Six4
Zen-ichi Ando, Shigeru Sato, Keiko Ikeda and Kiyoshi Kawakami
Division of Biology, Center for Molecular Medicine, Jichi Medical School, Tochigi, Japan
The Six homeobox gene is characterized by the
con-served Six domain (SD) and homeodomain (HD),
both of which are required for specific DNA binding
[1,2] The prototype of this gene family is Drosophila
sine oculis, which is essential for compound eye
for-mation [3,4] Six members (Six1–Six6) of the Six
family gene have been identified in mouse and
human [1] Six3 and Six6 are essential for forebrain
formation and eye development [5–11], whereas Six5
is involved in cataractogenesis and spermatogenesis
[12–14]
Among the Six family genes, Six1 and Six4 show
a remarkably similar expression pattern [1,15,16] Both
Six1 and Six4 bind to the MEF3 site in the myogenin
promoter and positively regulate the activity of the
promoter in conjunction with their coactivator Eya
proteins [17,18] Based on these observations, Six1 and
Six4 are thought to be functionally similar in vivo Six4–⁄ – mice showed little anomaly in embryogenesis including skeletal muscles [16] This was explained by the compensatory function of Six1 considering the similar expression pattern of both genes and the func-tional similarity in activating their target gene myo-genin In contrast, Six1–⁄ – mice showed anomalies in the development of various organs such as the inner ear, nose, thymus, kidney and skeletal muscle [19–23] These results indicate that Six4 does not compensate for the functional loss of Six1, whereas Six1 compen-sates for the function of Six4, suggesting a distinct function of Six1 and Six4 in vivo To understand the basis for the difference between Six1 and Six4, we screened possible target genes by using an approach similar to that applied in our previous study for searching the targets of Six5 protein [24] We identified
Keywords
dorsal root ganglia; Six1; Six4; Slc12a2;
transcriptional targets
Correspondence
K Kawakami, Division of Biology, Center for
Molecular Medicine, Jichi Medical School,
Yakushiji, Minamikawachi, Tochigi,
329-0498, Japan
Fax: +81 285 44 5476
Tel: +81 285 58 7312
E-mail: kkawakam@jichi.ac.jp
(Received 13 December 2004, revised
15 March 2005, accepted 11 April 2005)
doi:10.1111/j.1742-4658.2005.04716.x
Six genes are homologs of Drosophila sine oculis and encode transcription factors that are characterized by a conserved Six domain and homeo-domain Of the six family members (Six1–Six6) in mice, Six1 and Six4 show similar expression patterns during embryogenesis Six1–⁄ – mice show defective formation of various organs such as inner ear, nose, skeletal mus-cle, kidney and thymus, whereas Six4–⁄ –mice show little anomaly in organo-genesis To understand the molecular basis for the differential function of Six1and Six4 in vivo, we screened target genes of Six1 and Six4 and found that Six1 and Six4 differentially regulated a set of target genes Gel-retarda-tion assays indicated that the promoter region of one of the targets, sodium– potassium–chloride cotransporter 1(Slc12a2), contains multiple Six1-binding sites and one common binding site of Six1 and Six4, suggesting that the DNA-binding specificity of Six1 is distinct from that of Six4 This underlies the differential regulation of common target genes by Six1 and Six4 Fur-thermore, in situ hybridization demonstrated that the expression of Slc12a2 was reduced in the developing dorsal root ganglia of Six1– ⁄ –
⁄ Six4– ⁄ –mice, suggesting that Six1 and Six4 regulate Slc12a2 in vivo
Abbreviations
Atp1a1, sodium–potassium ATPase alpha 1 subunit; Clcn5, chloride channel 5; Coll2a1, procollagen type 2 alpha 1; Figf, c-fos induced growth factor; Gas1, growth arrest-specific 1; GST, glutathione S-transferase; HD, homeodomain; MCK, muscle creatine kinase; SD, Six domain; Slc12a2, sodium–potassium–chloride cotransporter 1; Trex, transcriptional regulatory element X.
Trang 2three putative categories of targets, Six1-specific
targets, Six4-specific targets and common targets We
then focused our analysis on one of the typical
com-mon target genes, sodium-potassium-chloride
cotrans-porter 1 (Slc12a2) to understand the molecular basis
for the distinct functions of Six1 and Six4
Results
Identification of potential downstream target
genes for Six1 and Six4
Studies of Six1–⁄ – mice revealed that Six1 is required
for inner ear, nose, thymus, skeletal muscle and kidney
formation [19–23] Six4 is also reported to be
expressed in these regions [16] To explore the target
genes of Six1 and Six4 in the physiological context of
development, we took advantage of using the cell line
mK4 isolated from the kidneys of transgenic mice that
express SV40 T-antigen under the control of Hoxa11
promoter The mK4 cells are thought to represent
embryonic metanephric mesenchyme that undergoes
epithelial conversion and expresses genes typical of late
mesenchyme such as E-cadherin, Wnt4, Pax2, Lim1,
Pax8 and Bmp7 in addition to Six1, Six4, Eya2 and
Eya3 [25] (data not shown) By overexpressing
tran-scriptionally constitutive active fusion proteins VP16–
Six1 and VP16–Six4, the direct target genes are
expec-ted to be activaexpec-ted through VP16–Six1 or VP16–Six4,
respectively, tethered to target gene promoters in the
mK4 cells As a control, fusion proteins VP16–
Six1W171R and VP16–Six4W263R, which are
defect-ive in DNA binding (data not shown), were
overex-pressed The mK4 cells were infected with recombinant
adenovirus expression vectors expressing VP16–Six
fusion proteins and cultured for 24 h, then used as the
source of poly(A)+ RNA to prepare hybridization
probes Expression profiling was performed by
hybridi-zation of an oligo microarray containing 20 371 mouse
cDNAs Scatter plot analysis of the microarray data
showed that most of the data points fell along the
diagonal, indicating that most of the genes were
equally expressed in the two samples (supplementary
Fig S1) The genes that showed more than 1.5-fold
higher level of expression in cells infected with VP16–
Six1 adenovirus compared with cells infected with
VP16–Six1W171R were considered potential Six1
tar-get genes, whereas 1.5-fold higher level of expression
in cells infected with VP16–Six4 compared with VP16–
Six4W263R were considered potential Six4 target
genes A total of 363 Six1 target genes and 149 Six4
targets genes were identified Of these, 63 were
com-mon target genes The data were deposited in GEO
database GSE2043 (a complete list of these genes appears in Table S1 and a partial list is shown in Table 1) Because only a single microarray was used for each condition, genes with a relatively low level of expression were excluded Examples of target genes included the cyclin-dependent kinase inhibitor 1C (Cdkn1c), a cell-cycle regulator expressed in the kidney and cochlea, which controls the number of podocytes and glomerular size in the kidney [26] Another exam-ple is the sodium–potassium–chloride cotransporter 1 (Slc12a2) that plays an important role in dorsal root ganglia function, spermatogenesis and inner ear forma-tion, and mutation of this gene causes impairment of hearing [27–30] Other examples include the sodium– potassium ATPase alpha 1 subunit (Atp1a1), which was originally identified as the target gene of Six4 [2,31]; c-fos induced growth factor (Figf), which shows remark-ably similar patterns of Six1 and Six4 expression [32]; growth arrest-specific 1 (Gas1), which encodes a sonic hedgehog inhibitor [33,34] and whose mutation causes cerebellar and eye defects [33]; chloride channel 5 (Clcn5), which is mainly expressed in the kidney, and mutation of this gene causes kidney stone disease [35,36]; and procollagen type 2 alpha 1 (Col2a1), whose missense mutation causes spondyloepiphyseal dyspla-sia, hearing loss and retinoschisis [37]
To confirm the above microarray results, we per-formed semiquantitative RT-PCR using RNA pre-pared from recombinant virus-infected mK4 cells As summarized in Table 2, the results of RT-PCR con-firmed > 1.5-fold difference in the expression levels of
13 of 14 genes examined The result seems to suggest that our strategy using microarray as an initial screen-ing to search for potential Six1- and Six4-target genes was largely successful
Differential regulation of potential target genes
by Six1 and Six4
To verify that the above identified putative target genes are regulated by Six1 or Six4, we performed transient transfection assays and examined the effects
of Six1 and Six4 on the promoter activity We used luciferase reporter constructs harboring the Gas1 pro-moter region of )3408 to +19 (Gas1)3408Luc), the Clcn5 promoter region of )1325 to +2529 (Clcn5) 1325Luc) and the Slc12a2 promoter region of )1938
to +149 (Slc12a2)1938Luc) Gas1 promoter showed similar extent of activation by both Six1 and Six4 in
a dose-dependent manner (Fig 1A), although it was listed as a potential Six1-specific target gene Clcn5 pro-moter showed a moderate repression by Six1 and a strong repression by Six4 in a dose-dependent manner
Trang 3Table 1 A partial list of potential target genes regulated by Six1 and ⁄ or Six4 in mk4 cells Expression profiling was performed using the Mouse Development Oligo Microarray containing 20 371 mouse cDNAs (Agilent Technologies, Palo Alto, CA) Total RNA was prepared as described previously [24] from mK4 cells infected with recombinant viruses To prepare fluorescence-labeled cDNA probes, total RNA (20 lg) was reverse transcribed using an oligo(dT) primer in the presence of aminoallyl dUTP and single-stranded cDNAs were coupled with Cy3 (VP16–Six1wt and VP16–Six4wt infected samples) or Cy5 (VP16–Six1W171R and VP16–Six4W263R samples) dyes Hybridization to the microarray was carried out at 65 C for 17 h according to the instructions provided by the manufacturer The arrays were washed, dried and scanned using ScanArray 5000 (GSI Lumonics Inc., ON, Canada) Cy3 and Cy5 intensities for each spot on the array were determined by QUANTARRAY software (GSI Lumonics Inc.) The raw data were processed and the Cy3 to Cy5 ratios were calculated as follows: (a) subtrac-tion of the fluorescence intensity of negative control spots as background from the intensity of each of the Cy3 and Cy5 spots, (b) normaliza-tion of the entire data set using the global normalizanormaliza-tion method, (c) eliminanormaliza-tion of spots with high background intensity for either dye, (d) determination of the Cy3 to Cy5 ratios The microarray data were deposited in the GEO database under accession number GSE2043 Among the genes with a Cy3 ⁄ Cy5 ratio of > 1.5, those with a relatively high level of expression (normalized signal value > 500 in both dyes) were considered to be potential Six1- and Six4-target genes Data are from a single microarray experiment and no dye-swap experiment was car-ried out.
Symbol
VP16–Six1 a
wt ⁄ W171R
VP16–Six4 b
wt ⁄ W263R
Kidney c function
Six1–Six4 common target genes
Six1-specific target genes
Six4-specific target genes
a Ratio of expression level in VP16–Six1wt-expressing mK4 cells to expression level in VP16–Six1W171R-expressing cells b Ratio of expres-sion level in VP16–Six4wt-expressing mK4 cells to expresexpres-sion level in VP16–Six4W263R-expressing cells.cGenes involved in kidney devel-opment or function.
Trang 4(Fig 1B) In our screening, Clcn5 was listed as a poten-tial Six1-specific target and it is reasonable that even the genes naturally repressed by Six1 were activated by VP16–Six1 in our screening condition Slc12a2 promo-ter was strongly activated by Six1, but weakly activated
by Six4 in a dose-dependent manner (Fig 1C) The apparent discrepancy in the activation⁄ repression pro-files of Six1 and Six4 based on the results of transient transfection assays and microarray analyses may be explained by the involvement of other transcription fac-tors in cells used in transfection assays VP16-fusion
Table 2 RT-PCR analysis of potential Six1- and Six4-target genes
identified by microarray Total RNA (10 or 100 ng) prepared from
mk4 cells infected with adenoviruses overexpressing VP16–Six1 or
VP16–Six4 fusion proteins was subjected to RT-PCR Aliquots of
PCR products were removed from the thermal cycler at multiple
cycle numbers, separated on a 5% acrylamide gel, stained with the
fluorescent dye, scanned and quantitated Linearity of PCR
amplifi-cation was maintained over several cycles, and the amount of PCR
products at 16–28 cycles, depending on the gene, was selected for
comparison Three independent RT-PCR reactions were set up
from the same RNA samples Data are shown as mean ± SEM.
We could not find 1.5-fold difference of expression in Lampt4A.
Symbol
Six1wt ⁄ Six1W171R a
Mean (± SEM)
Six4wt ⁄ Six4W263R b
Mean (± SEM) Cyclesc Six1–Six4 common targets
Six1-specific target genes
Six4-specific target genes
a
Ratio of expression level in VP16–Six1wt-expressing mk4 cells to
expression level in VP16–Six1W171R-expressing cells b Ratio of
expression level in VP16–Six4wt-expressing mk4 cells to
expres-sion level in VP16–Six4W263R-expressing cells.c The number of
PCR cycles at which the relative amount of PCR products were
determined.
10
8
6
4
2
0
1.2 1.0 0.8 0.6 0.4 0.2 0
C
35 30 25 20 15 10 5 0
D
25
20
15
10
5
0
Gas1-3408Luc Clcn5-1325Luc
Slc12a2-1938Luc pGL3MG-185
E
Six1
Slc12a2 pGL3MG
Gas1 Clcn5
Six4
Fig 1 Differential regulations of the potential target genes by Six1
and Six4 Transient transfection assays with 175 ng of the indicated
promoter-luciferase reporter construct were carried out as
des-cribed in Experimental procedures Luciferase activity was
normal-ized to the protein content and expressed relative to the value in
the presence of 75 ng pFLAG-CMV2 (white bars), which was set at
1 As effectors, increasing amounts (25 and 75 ng) of pfSix1 (black
bars) and pfSix4 (gray bars) were cotransfected into COS7 cells.
Data are mean ± SEM of three independent experiments (each
per-formed in duplicate) (A) Trans-activation of Gas1 promoter by Six1
and Six4 (B) Regulation of Clcn5 promoter by Six1 and Six4 (C)
activation of Slc12a2 promoter by Six1 and Six4 (D)
Trans-activation of myogenin promoter by Six1 and Six4 (E)
Gel-retarda-tion assays of Flag–Six1 and Flag–Six4 in nuclear extracts (800 ng
protein for Six1 and 860 ng protein for Six4) from transiently
trans-fected cells using 5 fmol of C3 oligonucleotide probe Arrows
indi-cate the positions of specific retarded complexes and arrowhead
indicates the position of nonspecific complex.
Trang 5protein may bypass the effects of other transcription
factors that may affect the promoter activity in cells
The control myogenin promoter was moderately
activa-ted by Six1 and strongly by Six4 (Fig 1D) To ensure
that the differential regulation by Six1 and Six4 is not
due to the altered expression levels of cotransfected
Six1 and Six4, we performed gel-retardation analysis
of nuclear extracts from cotransfected cells Similar
amounts of gel-retarded complexes of Flag–Six1 and
Flag–Six4 (Fig 1E) were detected among each set of
nuclear extracts from the transfected cells except in the
case of Flag–Six1 in the presence of the myogenin
reporter These findings indicate that Six1 and Six4
differentially regulate their target genes
Identification of Six1- and Six4-responsive
elements in the Slc12a2 promoter
We focused our analysis on Slc12a2 promoter to
deter-mine the molecular basis of the differential regulation
by Six1 and Six4 To localize Six1 and Six4 responsive
elements in the Slc12a2 promoter, we prepared a set
of deletion constructs harboring the Slc12a2 promoter
regions )1938 to +149 (Slc12a2)1938Luc), )820 to
+149 (Slc12a2–820Luc) and )97 to +149 (Slc12a2–
97Luc) Activation by Six1 and Six4 was observed in
the series of deletion constructs and similar activation
level was still observed in the shortest construct,
Slc12a2–97Luc (Fig 2A)
To explore the possibility that Six1- and
Six4-bind-ing sites lie within )97 to +149, we next performed
gel retardation assays using probes of NarI–HinfI
frag-ment ()97 to +77, Fig 2B, probe A), Cfr10I–MspI
fragment (+2 to +103, Fig 2B, probe B) and HinfI–
SacI fragment (+75 to +149, Fig 2B, probe C) The
expressed glutathione S-transferase (GST) fusion
pro-tein of Six1 (GST–Six1) bound to probes A, B and C,
whereas GST fusion protein of Six4 (GST–Six4) bound
to probe C, but not to probes A and B (Fig 2C)
These results suggest the existence of at least two
Six1-specific binding sites, and only one binding site
com-mon to Six1 and Six4
To analyse the location of each binding site, we
syn-thesized the double-stranded competitor
oligonucleo-tides (oligo 1 to oligo 9) that covered various portions
of the promoter region (Fig 3A) The formation of a
complex by Six1 was strongly competed by oligos 2, 3
and 6 for probe A, by oligos 6 and 7 for probe B and
by oligos 6, 7 and 9 for probe C (Fig 3B) In contrast,
the formation of a complex by Six4 was competed only
by oligo 9 for probe C (Fig 3C)
To precisely localize the binding element for Six1
and Six4, we generated 4-bp substitution mutations
in various regions of oligo 3, oligo 6 and oligo 9 (Table 3) and each mutated oligo was added to the gel retardation assay mixture to examine the competition
to the binding of Six1 and probes A, B and C Oligo
C
probe A B C A B C
GST-Six1 GST-Six4
1 2 3 4 5 6
complex
free probe
free probe free probe
Relative luciferase activity
A
Slc12a2-1938Luc
Slc12a2-820Luc
Slc12a2-97Luc
pGL3-basic
probe A probe B probe C
B
Fig 2 Identification of Six1 and Six4 responsive regions in the Slc12a2 promoter (A) Effects of Six1 or Six4 on the Slc12a2 pro-moter in COS7 cells In these studies, 175 ng of luciferase reporter constructs containing a series of deletions of Slc12a2 promoter fragments were cotransfected with increasing amounts (25 and
75 ng) of pfSix1(black bars) and pfSix4 (gray bars) Luciferase activ-ity was normalized to the protein content and expressed relative to the value in the presence of 75 ng pFLAG-CMV2 (white bar), which was set at 1 (B) Schematic representation of the Slc12a2 promoter with the position of a major transcription start site (+1) Solid bars
at the bottom show positions of probes used in gel retardation assays (C) Binding of Six1 and Six4 to the promoter DNA frag-ments of Slc12a2 Gel-retardation assays were performed using bacterially expressed 10 ng of GST–Six1 and 30 ng of GST–Six4 proteins with each 5 fmol probe indicated in (B) Lanes 1–3 and lanes 4–6 were from separated lanes in the same gel Free probes are indicated by the arrows and shifted complexes are indicated by the bracket.
Trang 63mut()13 ⁄ )10) showed the most reduced competition among the mutated oligos examined (Fig 4A, lane 7) The reduction of Six1 binding was confirmed by com-paring the binding of Six1 to the oligo 3wt probe and the oligo 3mut()13 ⁄ )10) probe The amount of the retarded complex composed of Six1 and oligo 3mut()13 ⁄ )10) was approximately threefold less com-pared with oligo 3wt (Fig 4B, lanes 2 and 4) As for probe B, all substitution mutations of oligo 6 showed similar competition compared with oligo 6wt (data not shown), suggesting the presence of multiple binding sites for Six1 in oligo 6 We also tested deletion oligos and found that oligo 6 with 10-bp deletion in +65 to +74 [oligo 6del(+65⁄ +74)] slightly reduced the com-petition compared with oligo 6wt (Fig 4A, lanes 11 and 12) The double mutation oligo 6 containing del(+65⁄ +74) and substitution mutation in +89 to +92 [del(+65⁄ +74) ⁄ mut(+89 ⁄ +92)] showed further reduced competition against the probe B–Six1 protein complex formation (Fig 4A, lane 13) The reduction
in the binding was confirmed by comparing the bind-ing of Six1 to the oligo 6wt probe and oligo 6del(+65⁄ +74) ⁄ mut(+89 ⁄ +92) probe and the bind-ing was approximately threefold weaker in the latter compared with the former (Fig 4B, lanes 6 and 10) The oligo 9mut(+135⁄ +138) showed the most reduced competition among the oligos examined against the probe C–Six1 protein complex formation (Fig 4A, lane 19) The reduction in the binding was confirmed by comparing the binding of Six1 to the
oli-go 9wt probe and the olioli-go 9mut(+135⁄ +138) probe and the Six1 binding to the latter was approximately threefold less than the former (Fig 4B, lanes 12 and 14) Likewise, the oligo 9mut(+135⁄ +138) showed the most reduced competition to the probe C–Six4 complex formation (Fig 4C, lane 6) The binding of Six4 to the oligo 9wt probe was compared with the oligo 9mut(+135⁄ +138) probe and in this case, the reduction of binding was 10-fold (Fig 4D, lanes 2 and 4)
In summary, the three Six1-specific binding sites reside around )13 to )10, +65 to +74 and +89 to +92 and the single common binding site resides around position +135 to +138 in the promoter region
of Slc12a2 (Fig 4E)
Activation of the Slc12a2 promoter
by Six1 and Six4
To confirm that Six1 activates through the above-identified binding sites, we introduced substitution mutations used in the gel-retardation assays into the luciferase reporter constructs We prepared the following
1 2 3 4 5 6 7
1 2 3 4 5 6
8 9 101112
4 5 6 7
1314151617
6 7 8 9
probe
competitor
probe
competitor
1 2 3 4 5 6 7 8 9 10
C
1 2 3 4 5 6 7 8 9
probe A
probe B
probe C
oligo 1 -99 -62
A
B
C
binding
binding
1.0 0.5 0.3 0.2 0.5 0.4 0.1 1.0 0.5 0.5 0.1 0.2 1.0 0.3 0.3 0.6 0.2
1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.1 1.2 0.6
Fig 3 Mapping of Six1 and Six4 binding elements in the Slc12a2
promoter (A) Positions of probes A, B and C are indicated in the
upper panel Positions of competitors (oligos 1–9) are indicated in
the lower panel (B) Gel retardation assays with GST–Six1 in the
presence (lanes 2–7, 9–12 and 14–17) or absence (lanes 1, 8 and
13) of competitors; 500-fold molar excess competitor
oligonucleo-tides were added to the reaction The protein amount of GST–Six1
was 16 ng for probe A (lanes 1–7), 15 ng for probe B (lanes 8–12)
and 13 ng for probe C (lanes 13–17) The intensities of the retarded
bands relative to that in the absence of competitors, which was
set at 1 (lanes 1, 8 and 13), are shown at the bottom (C)
Gel-retar-dation assays with 30 ng of GST–Six4 in the presence (lanes 2–10)
or absence (lane 1) of competitors; 500-fold molar excess
compet-itor oligonucleotides were added The intensities of the retarded
bands relative to that in the absence of competitors, which was
set at 1(lane 1), are shown at the bottom Free probes are indicated
by the arrow and shifted complexes by the bracket in (B) and (C).
Trang 7mutation reporters, Slc12a2–97mut()13 ⁄ )10)Luc,
Slc12a2–97del(+65⁄ +74) ⁄ mut(+89 ⁄ +92)Luc, and the
combination of the two, Slc12a2–97mut()13 ⁄ )10) ⁄
del(+65⁄ +74) ⁄ mut(+89 ⁄ +92)Luc as Six1-binding
muta-tions We also prepared Slc12a2–97mut(+135⁄ +138)Luc,
which abolished Six4 binding We performed reporter
gene assays and analysed the effects of Six1 (Fig 5A)
and Six4 (Fig 5B) Cotransfection of pfSix1 showed
activation of Slc12a2–97Luc in a dose-dependent
man-ner to an 11-fold increase in the level of luciferase
activity The activation level was decreased to
approxi-mately sixfold in Slc12a2mut()13 ⁄ )10)Luc, whereas
Slc12a2del(+65⁄ +74) ⁄ mut(+89 ⁄ +92)Luc showed
com-parable luciferase activity to the wild-type reporter
construct, Slc12a2–97Luc In contrast, the
combina-tion of these two mutacombina-tions, Slc12a2mut()13 ⁄ )10) ⁄ del(+65⁄ +74) ⁄ mut(+89 ⁄ +92)Luc, showed only 1.3-fold activation by Six1 (Fig 5A) These results clearly indicate that the three Six1-binding sites located at around )13 ⁄ )10, +65⁄ +74 and +89⁄ +92 are the responsible element as a whole for the activation by Six1 As for the effects of Six4, cotransfection of pfSix4 showed 2.6- to 4.6-fold acti-vation in the wild-type reporter construct Slc12a2– 97Luc in a dose-dependent manner By contrast, the luciferase activity of Slc12a2mut(+135⁄ +138) was enhanced to 1.6- to 2.3-fold by Six4 (Fig 5B) These results indicate that Six4 activates the Slc12a2 promoter through its binding site around +135⁄ +138
Table 3 Double-stranded oligonucleotides used in Fig 4 Nucleotide substitutions in the mutation oligonucleotides are represented in bold.
Trang 8Slc12a2 is regulated by Six1 and Six4 in vivo
To confirm that Six1 and Six4 regulate the expression of
Slc12a2 in vivo, we examined the expression of Slc12a2
in the developing embryos of wild-type, Six1–⁄ – and
Six1–⁄ –⁄ Six4– ⁄ – mice The expression of Slc12a2 was
observed in various adult organs such as the central
ner-vous system, dorsal root ganglia and renal cortex [38]
Because Six1–⁄ –and Six1–⁄ –⁄ Six4– ⁄ –
mice die soon after birth and they show developmental defects in various
organs [23] (unpublished observation), we compared the
expression pattern of Slc12a2 in these embryos The expression level of Slc12a2 was too low to allow precise comparison of its expression level in the nephrogenic cord Therefore, we analysed the expression of the gene
by in situ hybridization in the dorsal root ganglia, where Six1 and Six4 were abundantly expressed and some developmental abnormalities were observed in Six1–⁄ –⁄ Six4–⁄ – mice (K Ikeda & K Kawakami, unpublished observation) The antisense probe detected expressions
of Slc12a2 in the dorsal root ganglia in E18.5 fetus where significant expression was reported [38] (Fig 6C),
1 2 3 4 5 6 7 8 1 2 3 4
probe C
9wt 9m
oligo 9wt oligo 9m
-13/-10 +65/+74 +89/+92 +135/+138
: Six1-specific binding site : Six1/Six4 common binding site
1.0 0.3 0.3 0.3 0.6 0.8 0.3 0.3 1.0 0.1
oligo 9wt oligo 9m
1.0 0.3
5 6 7 8 9 10
oligo 6wt oligo 6del(+65/+74) oligo 6del(+65/+74)
1.0 0.6 0.3
1 2 3 4
oligo 3wt oligo 3m
1.0 0.3
1415161718192021 probe C
9wt 9m
1.0 0.4 0.3 0.4 0.5 0.8 0.3 0.4
10111213 probe B
6wt 6del(+65/+74) 6del(+65/+74) /m
1.0 0.1 0.2 0.3
1 2 3 4 5 6 7 8 9
probe A
3wt 3m
1.0 0.2 0.2 0.1 0.2 0.1 0.3 0.2 0.2
11121314
Trang 9whereas the control sense probe gave only background
signals (Fig 6D) The expression level of Slc12a2 was
apparently lower in Six1–⁄ –⁄ Six4– ⁄ –
embryo at E16.5 compared with the wild-type, whereas that in Six1– ⁄ –
was similar to the wild-type (Fig 6H–J) We observed
similar reductions of Slc12a2 expression in the dorsal
root ganglia of Six1–⁄ –⁄ Six4–⁄ – embryos at E17.5 and
E18.5 (data not shown) In the choroid plexus, where no
Six1 and Six4 were expressed, the expression level of
Slc12a2 was similar in each genotype (Fig 6N–P)
These results suggest that the low expression level of
Slc12a2 in Six1–⁄ –⁄ Six4– ⁄ – embryos is due to the
absence of Six1 and Six4 in the dorsal root ganglia
and indicate that Slc12a2 is upregulated by Six1 and
Six4 in the developing dorsal root ganglia of wild-type
embryo
To confirm that Six1 and Six4 regulate the other putative target genes in vivo, we analysed the expres-sion of Figf and Col2a1 in E10.5 embryos of wild-type and Six1– ⁄ –
⁄ Six4– ⁄ –by RT-PCR The expression levels
in Six1– ⁄ –
⁄ Six4– ⁄ – embryo were significantly reduced (to 36.4% for Figf and to 58.1% for Col2a1 compared with the wild-type), suggesting that these genes are also regulated by Six1 and Six4 in vivo (Fig 7)
Discussion Screening of putative target genes of Six1 and Six4 and effectiveness of the screening method
To understand the function of transcription factors
in organ development, it is essential to identify direct
Fig 4 Identification of Six1- and Six4-binding sites in the Slc12a2 promoter by gel retardation assays The competitor oligonucleotides har-boring mutations are shown in Table 3 (A) Competition assays were performed with GST–Six1 protein in the absence (lanes 1, 10 and 14)
or presence of 500-fold molar excess of various competitor oligonucleotides (lanes 2–9, 11–13 and 15–21) Probes A, B and C indicated in Fig 2 were used The protein amount of GST–Six1 was 20 ng for probe A (lanes 1–9), 15 ng for probe B (lanes 10–13), and 16 ng for probe
C (lanes 14–21) The intensities of the retarded bands relative to that in the absence of competitors for each probe, which was set at 1 (lanes 1, 10 and 14), are shown at the bottom Shifted complexes were least competed by oligo 3mut( )13 ⁄ )10) (lane 7), oligo 6-del(+65 ⁄ +74) ⁄ mut(+89 ⁄ +92) (lane 13), and oligo 9mut(+135 ⁄ +138) (lane 19) (B) Comparison of binding of GST–Six1 to the wild-type
(oli-go 3wt, lane 2; oli(oli-go 6wt, lane 6; oli(oli-go 9wt, lane 12) and to the mutated probes [oli(oli-go 3mut( )13 ⁄ )10), lane 4; oligo 6del(+65 ⁄ +74) and oligo 6del(+65 ⁄ +74) ⁄ mut(+89 ⁄ +92), lanes 8 and 10, respectively; oligo 9mut(+135 ⁄ +138), lane 14] Sixty-seven nanograms of GST–Six1 was used in the reactions Binding of GST–Six1 to oligo 3mut( )13 ⁄ )10) was 30% compared with that of oligo 3wt (lanes 2 and 4) Binding
of GST–Six1 to oligo 6del(+65 ⁄ +74) and oligo 6del(+65 ⁄ +74) ⁄ mut(+89 ⁄ +92) was 60 and 30%, respectively, compared with that of
oli-go 6wt (lanes 6, 8 and 10) Binding of GST–Six1 to olioli-go 9mut(+135 ⁄ +138) was 30% compared with that of oligo 9wt (lanes 12 and 14) Lanes 5–6 and lanes 7–10, lane 11 and lanes 12–14 are from separate lanes of one gel (C) Competition assays were performed with GST– Six4 protein in the absence (lane 1) or presence of 500-fold molar excess of various competitor oligonucleotides (lanes 2–8) Probe C indica-ted in Fig 2 was used Thirty nanograms of GST–Six4 was used in the reaction The intensities of the retarded bands relative to that in the absence of competitors for the probe, which was set at 1(lane 1), are shown at the bottom Shifted complexes were most weakly com-peted by the oligo 9mut(+135 ⁄ +138) (lane 6) (D) Comparison of binding of GST–Six4 to the wild-type (oligo 9wt, lane 2) and mutated probes [oligo 9mut(+135 ⁄ +138), lane 4] Thirty nanograms of GST–Six4 was used in the reactions Binding of GST–Six4 to oligo 9mut(+135 ⁄ +138) was 10% compared with that of oligo 9wt (lanes 2 and 4) (E) Schematic representation of binding elements of Six1 and Six4 in the Slc12a2 promoter Six1-specific binding elements reside in the region of )13 ⁄ )10, +65 ⁄ +74 and +89 ⁄ +92 (black ovals) and a bipartite Six1-and Six4-binding element in the region of +135 ⁄ +138 (gray box).
Fold activation
Fold activation
Slc12a2-97Luc
Slc12a2-97mut(-13/-10)Luc
Slc12a2-97del(+65/+74)
Slc12a2-97mut(-13/-10)
/mut(+89/+92)Luc
/del(+65/+74)/mut(+89/+92)Luc
Slc12a2-97Luc
Slc12a2-97mut(+135/+138)Luc
A
B
Fig 5 Transactivation of the Slc12a2 promoter by Six1 and Six4 (A, B) Slc12a2– 97Luc and four types of mutation reporter constructs (shown in the right-hand panel,
70 ng each) were cotransfected with increa-sing amounts (25 and 75 ng) of pfSix1 or pfSix4 in COS7 cells Luciferase activity was normalized to protein content and expressed relative to the value in the presence of
75 ng pFLAG-CMV2 (white bar), which was set at 1 Data are mean ± SEM of a typical result of three independent experiments (each performed in triplicate).
Trang 10target genes and recognize the gene cascade as well as
regulatory mechanisms involved in proper cell growth,
differentiation, cell movement and functional
matura-tion of the organ In this analysis, we took advantage
of a model cell line that reflects a certain developmen-tal stage of the kidney in order to identify the genes
N
Dorsal root
ganglion
Choroid
plexsus
Wild type
E
O
P
sp drg
sp drg
v
Antisense
D C
Dorsal root
ganglion
Six1-/- Six1-/-/Six4
-/-Sense
Fig 6 In situ hybridization of Slc12a2 in the dorsal root ganglia of mouse embryos (A–D) Transverse sections (16 lm) were stained with hematoxylin (A, B) and the adjacent sections hybridized with Slc12a2 antisense (C) or sense (D) probe Specific hybridization signals were detected in the dorsal root ganglia (encircled by the dotted line) of E18.5 embryo in (C) but not in (D) (E–J) Transverse sections (16 lm) from E16.5 wild-type (E, H), Six1–⁄ –(F, I) and Six1–⁄ –⁄ Six4 – ⁄ –
(G, J) embryos were stained with hematoxylin (E–G) and the adjacent sections hybridized with Slc12a2 antisense probes (H–J) The expression levels of Slc12a2 in the developing dorsal root ganglia of the Six1 – ⁄ –
embryos were similar to those in the wild-type (H, I), but markedly reduced in the Six1 – ⁄ –
⁄ Six4 – ⁄ –
embryos (J) (K–P) As controls, the signals
on the choroid plexus from E16.5 wild-type (K, N), Six1–⁄ –(L O) and Six1–⁄ –⁄ Six4 – ⁄ –
(M, P) embryos stained with hematoxylin (K–M) and with antisense probe (N–P) are shown Similar expression levels of Slc12a2 were observed in all genotypes (N–P) Scale bar: 100 lm.