Keywords microarray analysis; statistical analysis of transcription factor motifs; transcription factor binding site; Wnt pathway; Wnt target gene Correspondence O.. We compared the numb
Trang 1insights into Wnt target promoter regulation
Slava Ziegler1, Sonja Ro¨hrs1, Lara Tickenbrock2, Tarik Mo¨ro¨y3, Ludger Klein-Hitpass3,
Ingrid R Vetter1and Oliver Mu¨ller1
1 Max-Planck-Institut fu¨r molekulare Physiologie, Dortmund, Germany
2 Interdisziplina¨res Zentrum fu¨r Klinische Forschung, Universita¨tsklinikum Mu¨nster, Mu¨nster, Germany
3 Institut fu¨r Zellbiologie (Tumorforschung), Universita¨tsklinikum, Essen, Germany
The Wnt signalling pathway plays an important role in
embryonic development, cell differentiation and cell
proliferation [1] The biological effects of the Wnt
pathway are mediated via the regulation of direct
(pri-mary) and indirect (secondary) target genes (for a list
of target genes see http://www.stanford.edu/rnusse/
pathways/targets.html) [2,3] The transcription
activa-tion by TCF⁄ Lef and b-catenin is a crucial step in the direct target gene activation by the pathway Thus the activation of a Wnt target gene by TCF via a TCF⁄ Lef binding motif within the promoter region indicates the direct regulation Direct Wnt target genes are the cell cycle kinase activator cyclin D1 and the transcription factors MYC and PEA3 [4–6]
Keywords
microarray analysis; statistical analysis of
transcription factor motifs; transcription
factor binding site; Wnt pathway; Wnt
target gene
Correspondence
O Mu¨ller, Max-Planck-Institut fu¨r
molekulare Physiologie, Otto-Hahn-Straße
11, 44227 Dortmund, Germany
Fax: +49 231 133 2199
Tel: +49 231 133 2158
E-mail: oliver.mueller@mpi-dortmund.
mpg.de
(Received 15 December 2004, revised 19
January 2005, accepted 24 January 2005)
doi:10.1111/j.1742-4658.2005.04581.x
The Wnt pathway controls biological processes via the regulation of target gene expression The expression of direct Wnt target genes, e.g cyclin D1 and MYC, is activated by the transcription factor TCF, which binds to specific sequence motifs in the promoter Indirect target genes are regulated via transcription regulators, which are targets of the Wnt pathway As an example, MYC regulates the MYC interacting zinc finger protein-1 (MIZ-1), which is able to inhibit the expression of the indirect target p21WAF1
We intended to identify new Wnt target genes and to get a deeper insight into the regulatory mechanisms of Wnt target gene expression For this we analyzed the differential expression pattern of Wnt-1 activated cells by microarray analysis We identified 43 sequences including eight expressed sequence tags (ESTs), which showed increased transcript levels, and 104 sequences including 19 ESTs with decreased RNA levels Northern blot and real-time quantitative PCR analysis of the differential expression levels
of 15 genes confirmed the differential expression trends of eight candidate genes When the Wnt pathway was regulated at the lower level of glycogen synthase kinase-3b (GSK-3b) or adenomatous polyposis coli (APC), we detected discrepant expression trends We compared the number of binding sites of transcription factors in the genomic regions of all candidate target genes with the number of sites in control genes We found that the genomic regions of the down-regulated genes include an increased number of puta-tive MIZ-1 binding sites Our study introduces several new Wnt target genes and provides indications that the specific gene expression pattern depends on the type of the activation trigger or the level of interference with the Wnt pathway Furthermore, our data indicate that a high propor-tion of Wnt target genes are regulated by indirect mechanisms
Abbreviations
APC, adenomatous polyposis coli; EST, expressed sequence tag; GEO, gene expression omnibus; GSK-3b, glycogen synthase kinase-3b; MIZ-1, MYC interacting zinc finger protein-1; qPCR, real-time quantitative PCR; RT-PCR, reverse transcription PCR; TSS, transcription start site.
Trang 2While the regulatory mechanism of direct target gene
expression is well established, there are only few data
available about the regulation of indirect Wnt targets A
prominent indirect negative target is the cell cycle kinase
inhibitor p21WAF1, which is induced in cells expressing
dominant negative TCF [7] There are indications that
p21WAF1 expression is regulated by the relative levels
of the MYC interacting zinc finger protein-1 (MIZ-1)
and the direct Wnt target MYC [8] In the absence of
MYC, MIZ-1 activates p21WAF1 expression, while in
the presence of high MYC levels, MIZ-1 together
with MYC functions as a repressor of p21WAF1
tran-scription In epithelial cells of the intestine, the MYC⁄
MIZ-1⁄ p21WAF1 cascade acts as a switch between
dif-ferentiation and proliferation [7] There are several other
known genes, which show a decreased expression level
in Wnt-activated cells Examples are targets of the AP-1
family of transcription factors, which show decreased
expression in Wnt-activated cells [9]
Many microarray screenings have been performed
with the aim of identifying new target genes of the Wnt
pathway [9–12] Until now, only a limited number of
candidate target genes have been confirmed and further
characterized by experimental methods Recently, a new
statistical approach was introduced to analyse the
distri-bution of TCF sites in promoters of genes, which have
been identified as potential candidates of b-catenin⁄ TCF
signalling by microarray screening [11] We expanded
this strategy to gain general insights into the regulation
of promoters of potential target genes of the Wnt
path-way, which we identified by microarray screening
Results
Evaluation of cell models
At the beginning we evaluated our cell models First
we proved the presence of the Wnt-1 transcript in the
C57MG⁄ Wnt-1 cells (Fig 1A) A clear RT-PCR
pro-duct of the expected size indicated Wnt-1 expression in
C57MG⁄ Wnt-1 cells of the same passage as the cells,
which we used for further analysis Second, we showed
that the incubation of C57MG cells with lithium or
with Wnt-3a containing conditioned medium led to an
increase of the relative b-catenin level (Fig 1B,C) We
conclude that the Wnt pathway in C57MG cells can
be activated by lithium and by conditioned medium
from L Wnt-3a cells
Microarray analysis
We compared the gene expression patterns of
C57MG⁄ Wnt-1 cells and C57MG ⁄ PLNCx cells by
DNA microarray analysis As we used only one array per condition, this analysis served only as a very first screening for the collection of gene data with potentially differential expression The differential expression as well as the function of the genes as Wnt pathway target genes had to be evaluated in further experiments The raw data of the analysis with the list of all analysed genes and the expression factors were deposited in the GEO (Gene Expression Omnibus) database (http:// www.ncbi.nlm.nih.gov/projects/geo) under the GEO accession numbers GSM32828 and GSM32829 When using the described criteria for the definition of candi-date target genes, we were able to identify 147 candicandi-date targets of the Wnt pathway: 43 up-regulated, among them 35 known genes and eight ESTs We found 104 down-regulated sequences representing 85 known genes and 19 ESTs (supplementary tables S2, S3) Nineteen of
A
B
C
Fig 1 Evaluation of cell models (A) Agarose gel electrophoresis of RT-PCR products using template RNA purified from C57MG ⁄ PLNCx cells (P) or from C57MG ⁄ Wnt-1 cells (W) The PCR product of the length of the Wnt-1 transcript is marked The lengths of the exem-plary marker fragments are indicated in base pairs (bp) M, Marker (B) Western blot to analyse the relative b-catenin levels in C57MG cells after incubation with 20 m M KCl (K + ) or 20 m M LiCl (Li + ) The size of a marker fragment is shown on the left in kilodaltons (kD) (C) Western-blot to analyse the relative b-catenin levels in C57MG cells after incubation with the conditioned medium from non-trans-fected L-M(TK-) cells (L-CM) and from Wnt-3a transnon-trans-fected L-M(TK-) cells (L-Wnt-3a-CM) The size of a marker fragment is shown on the left in kilodaltons (kDa).
Trang 3these candidate genes have already been described as
target genes in Wnt-activated murine cells or in APC
inactivated mouse model systems (supplementary tables
S2, S3) Of these, 12 genes were described with
expres-sion trends opposite to the trends that we found
Evaluation of gene expression in C57MG/Wnt-1
cells
Several methods were used to evaluate microarray data
and to identify and to confirm target genes of the Wnt
pathway All biological experiments were performed at
least twice Representative results of the northern blot
and western blot experiments are shown To evaluate the
microarray data the expression levels of 15 selected
candi-date genes were compared in C57MG⁄ Wnt-1 cells and in
C57MG⁄ PLNCx cells by northern blot: the up-regulated
candidates Lcn2, Fos, Gnb1, Fabp5, Slc7a2, Osf2, Sdc4,
Ccl2, Pedf and Fas and the down-regulated candidates
Kit1, Lgals8, Gbp3, Mrp⁄ Plf and Ramp3 While the
dif-ferential expression of the genes Lcn2, Fos, Fabp5, Slc7a2,
Ccl2, Pedf, Fas, Mrp⁄ Plf and Ramp3 could be confirmed
(Fig 2), the genes Gnb1, Osf2, Sdc4, Kit1, Lgals8 and
Gbp3 showed similar expression levels in Wnt-1
trans-fected cells compared to the levels in control cells (not shown) Next we asked whether the differential expres-sion trends depend on the type of the trigger and the method of Wnt pathway activation For this, we tested the transcript levels after activation of the Wnt pathway
by lithium or by Wnt-3a We are aware that lithium influ-ences several pathways and inhibits other enzymes in addition to GSK-3b (e.g Na⁄ K ATPase, adenylcyclase, enzymes of the prostaglandins E1 synthesis and inositol-1-phosphatase) Nevertheless lithium incubation is a common way to activate the Wnt pathway [32–34] The candidates Fos, Slc7a2 and Ramp3 showed expression trends in the presence of lithium, which correlated to the corresponding trends in C57MG⁄ Wnt-1 cells, whereas Lcn2, Fabp5, Ccl2, Pedf, Fas and Mrp⁄ Plf showed trends contrary to the corresponding trends (Fig 3) Next we analysed the transcript levels after activation of the Wnt pathway by Wnt-3a containing medium (L-Wnt-3a-CM) [24] We found that L-Wnt-3a-CM incubation leads to
an increased relative expression rates of the genes Fos, Fabp5, Slc7a2, Ccl2, Pedf and Fas, and to decreased rates of the genes Mrp⁄ Plf and Ramp3 (Fig 4) These experimental results are consistent with the microarray results and indicate that the activation level in the
Fig 2 Northern blot analysis of expression
of candidate target genes in C57MG ⁄ PLNCx (P) and C57MG ⁄ Wnt-1 (W) cells The rele-vant bands are indicated and marked by the abbreviation codes of the corresponding genes The ethidium-bromide stained gels with the 18S and 28S rRNA species are shown for demonstrating equal loading.
Trang 4pathway and the type of the activation trigger are crucial
for the target gene regulation
To see whether the differential gene expression can
also be found on the translation level, we analyzed the
relative amounts of the extracellular proteins Lcn2, Ccl2, Pedf and Mrp⁄ Plf in the culture medium and of Ramp3 in the cellular lysates of Wnt-1 transfected and
of lithium stimulated C57MG cells (Fig 5) The west-ern blots confirmed the findings of the northwest-ern blot analysis with one exception The level of Mrp⁄ Plf was increased in Wnt transfected cells, whereas the RNA level was decreased in these cells as found by micro-array and by northern blot We decided to analyze more closely the Mrp⁄ Plf protein levels in Wnt-activa-ted cells For this we tesWnt-activa-ted the relative amount of intracellular Mrp⁄ Plf in lithium activated cells (Fig 5I) and detected a higher protein level as compared to nonactivated cells; this correlates to the high RNA level in these cells and to the high level of the extracel-lular protein (Figs 3 and 5G,H)
There are 4 Mrp⁄ Plf proteins, Plf1, Plf2, Mrp3 and Mrp4, which are highly homologous on the sequence level but differ in their patterns of glycosylation The corresponding mRNAs are transcribed from overlap-ping genomic regions [27] Using RT-PCR and restric-tion enzyme digesrestric-tion we found that the RNA, which
we purified from PLNCx or Wnt-1 transfected C57MG cells or from cells treated with K+ or Li+ contains the transcripts of Plf1 and Mrp4, rather than
Fig 3 Northern blot analysis of expression
of candidate target genes in C57MG cells
after treatment with 20 m M LiCl The
relevant bands are indicated The
ethidium-bromide stained gels with the 18S and 28S
rRNA species are shown to demonstrate
equal loading.
Fig 4 Analysis of transcript levels by real-time quantitative PCR
(qPCR) The relative expression rates of the indicated genes in
L-Wnt-3a-CM treated C57MG cells are shown (grey bars) in relation
to the expression rates in L-CM treated cells (white bars) Results
are the means of two independent analyses, which were
per-formed in triplicates.
Trang 5these of the two other isoforms (Fig 6) These results
correlate to the protein bands of 38 kDa (Plf1) and
27 kDa (Mrp4) [27], which were detected in the
west-ern blot analysis (Fig 5)
Analysis of promoter regulation by Wnt
signalling
We screened databases for the promoters of the 15
genes that we had further analyzed by northern blot
and found the promoter regions of Lcn2 [35], Fos [36],
Gnb1 (AB066210), Fabp5 [37], Sdc4 (D89572), Pedf
(AF017050), Kit1 [38] and Mrp⁄ Plf [39] (Table 1)
Pro-moter sequences of the genes Lcn2, Fos, Fabp5, Pedf
and Kit1 were used in luciferase reporter genes to test
the regulatory activities of the promoters in cells with
activated Wnt pathway When we compared the
activit-ies of some of these reporter constructs in RK13 and
SW480 cells and the TCF⁄ Lef responsive topflash motif
[29] we confirmed our results from the RNA and
pro-tein expression analysis (Table 2) (Fig 7) Lithium
acti-vation of the Wnt pathway regulated the Lcn2, Fos and
Pedf promoters as expected from the transcript levels
in the northern blot experiments Pathway inactivation
by APC transfection in SW480 cells supported the
assumption that the trend of expression regulation
depends on the level within the Wnt pathway, on which
the trigger interferes with the pathway The activation
or inactivation on the GSK-3b level by lithium or on a
further down-stream level by APC transfection had
similar effects on the promoters Lcn2 and Pedf These
effects are opposite to the effects caused by activation
of the pathway on the Wnt level by Wnt-1 transfection
Classification of analyzed genes
We screened the known promoter sequences for binding sites of the transcription factors TCF, MYC, AP-1, PEA3 and MIZ-1 (Table 1) Based on the results of this analysis we tried to define the identified genes as direct
or indirect target genes When analysing other path-ways, this discrimination could be made experimentally The inhibition of protein translation (e.g by cyclohexi-mide) led to the selective repression of indirect targets When analysing the Wnt pathway, this approach is not readily applicable because signal transduction depends
on the ongoing protein synthesis of b-catenin [10] The Lcn2 promoter does not comprise any binding site for the transcription factors TCF, MYC, AP-1 or PEA3, which are assumed to activate direct or indirect Wnt target genes suggesting that Lcn2 expression is activated via motifs outside the reported promoter region Indeed, six TCF⁄ Lef binding sites were found
at positions outside the reported Lcn2 promoter in the genomic region from position )3000 to +100 bp: )2644, )1875, )1809, )1377, )926, )827 The promot-ers of the genes Gnb1, Fabp5, Sdc4 and Pedf comprise binding sites for transcription activating, and the genes Kit1 and Mrp⁄ Plf include binding sites for expression repressing factors (Table 1)
Screening of regulatory regions of differentially expressed genes and statistical evaluation
We estimated the overall relevance of the microarray analysis using statistical analysis [11] First we screened public databases to get the genomic sequence of the
L K
I
Fig 5 Western blot analysis of protein levels in culture medium or in lysates from C57MG ⁄ PLNCx (P), C57MG ⁄ Wnt-1 (W), or C57MG cells treated with either 20 m M KCl (K + ) or with 20 m M LiCl (Li + ) (A–H) The extracellular proteins Lcn2, Ccl2, Pedf and Mrp ⁄ Plf were analyzed in the concentrated culture medium I–L, Mrp ⁄ Plf was additionally and Ramp3 was exclusively analyzed in the cellular lysates The expected bands of the proteins are indicated The sizes of exemplary marker proteins on the corresponding stained gels are shown in kilodaltons (kDa) on the right sides of the gels.
Trang 6differentially expressed genes Unknown genes (ESTs
in supplementary tables S2, S3) were not included
We were able to identify the genomic regions from
)3000 to +100 bp relative to the transcription start
of 119 of the 120 candidate genes All retrieved
seq-uences are available upon request We looked for the
presence of transcription factor motifs in these regions
using the public signal scan service (http://bimas
dcrt.nih.gov/molbio/signal) and in-house software
Analyzing the genomic regions of the up-regulated candidates, we included the motifs of four transcrip-tion factors, which are important for the direct or indi-rect activation of gene expression in Wnt-activated cells These are TCF, Myc and PEA3 [5,6] and AP-1 [31] (supplementary table S3, Figs 2 and 3) In the analysis of the down-regulated candidates we included the putative motifs of the transcription factor MIZ-1, which is assumed to be involved in the regulation of negative Wnt targets [7] In this analysis we again included the AP-1 promoter element because others identified Fos as a gene that is down-regulated in Wnt-activated cells [9]
Results of the analysis are summarized in Tables 3–5 (For detailed results with all identified positions of the screened transcription factors, see sup-plementary tables S4 and S5.) All genes showing decreased transcription levels include MIZ-1 sites Remarkably, the total number of MIZ-1c sites is signi-ficantly increased in the region )1000 to )501 bp of target genes with decreased expression, when compared
to the number of sites in the corresponding region in control genes (Table 6) The number of the transcrip-tion factors TCF⁄ Lef, MYC, AP-1 and PEA3 showed
no significant accumulation in the genes with increased
or decreased expression (Tables 4 and 6)
Fig 6 Analysis of the proliferin gene transcript (A) Schematic
pres-entation of the four possible fragments obtained by RT-PCR of the
transcripts of Mrp ⁄ Plf isoforms [27] The highly homologous
frag-ments differ in the numbers and positions of Fnu4H1 ⁄ BstX1
restriction sites All four isoforms comprise BstX1 and Fnu4H1
sites at positions 34 and 598 bp Plf1 and Mrp3 show BstX1 sites
at position 352, and Plf2 and Mrp3 show Fnu4H1 sites at
posi-tion 323 In the horizontal grey bars the lengths of the expected
restriction fragments are given in base pairs (B) Agarose gel
electrophoresis of single digestions of RT-PCR products with the
indicated restriction enzymes RNA purified from C57MG ⁄ PLNCx
cells was used as template (C) Agarose gel electrophoresis of the
Fnu4H1 ⁄ BstX1 double digestions of RT-PCR products RNA purified
from C57MG ⁄ PLNCx (P), from C57MG ⁄ Wnt-1 (W) cells, or from
C57MG cells treated either with KCl (K + ) or LiCl (Li + ) was used as
template The sizes of exemplary marker fragments are indicated in
base pairs (bp) M, marker lane.
Table 1 Number of transcription factor binding sites in promoters The promoters of the up-regulated genes were analysed for binding sites of TCF, MYC, AP-1 and PEA3 The promoters of the down-regulated genes were analysed for binding sites of AP-1 and MIZ-1.
Gene Promoter length
Promoter accession ID
Binding sites of promoters of up-regulated genes
Number of transcription factor binding sites
Binding sites of promoters of down-regulated genes Number of transcription factor binding sites AP-1 MIZ-1a MIZ-1b MIZ-1c
Trang 7Over recent years, a vast number of candidate Wnt
target genes have been identified by microarray
analysis [9–12] In the majority of these studies, one or
a few of the identified target genes were characterized
In this study, we followed more general aims First we
intended to identify and to confirm novel targets of
the Wnt pathway We identified 147 candidate targets,
while screening the differential gene expression pattern
of Wnt-1 transfected murine cells by microarray
analy-sis When judging these data principal constraints have
to be made First, the limited number of arrays might
have led to the identification of candidate genes, which
are no true target genes Second, the Wnt specific
expression pattern in Wnt-1 transfected cells might be
masked by unspecific side-effects of the stable
transfec-tion with a viral vector and the constitutively active
Wnt pathway Both limitations can be smoothened by
the experimental evaluation of the microarray data
using alternative methods and experimental models
We analyzed the differential expression levels of
exem-plary candidate genes by northern and western blots
and by qPCR The differential expression trend of the
9 genes Lcn2, Fos, Fabp5, Slc7a2, Ccl2, Pedf, Fas,
Mrp⁄ Plf and Ramp3 could be confirmed by northern
and western blot of Wnt-1 transfected cells These
genes have been described as important for apoptosis,
cell proliferation and angiogenesis in several other
studies The qPCR analysis confirmed the expression
trend of eight of these genes in Wnt-3a stimulated cells The analysis of the levels of selected proteins cor-related with the northern blots with one exception The Mrp⁄ Plf protein level in Wnt transfected cells is increased, though the RNA level is decreased in these cells We speculate that Mrp⁄ Plf is regulated by post-transcriptional mechanisms, which lead to the relative accumulation in Wnt transfected cells or to the relative protein degradation in control cells After lithium acti-vation, the genes Fos, Slc7a2 and Ramp3 showed the same regulation trend as in Wnt-1 transfected cells The genes Lcn2, Fabp5, Ccl2, Pedf, Fas and Mrp⁄ Plf showed different regulation trends in lithium treated cells and in Wnt-1 transfected cells We presume that the effects induced by lithium are due to its inhibitory effects on GSK-3b Though we can not exclude that other lithium regulated enzymes have influences on Wnt target gene expression There are two possible explanations for the discrepancies between the gene expression pattern in lithium activated cells and the pattern in Wnt-1 transfected cells First, the type of trigger and the activation level in pathway might play
a role in the regulation of gene expression The acti-vation at an upper level of the pathway by the Wnt factor might have effects that are different to those caused by activation at the GSK-3b level by lithium This difference might be caused by activation or
by inhibitory feedback loops, which are present at specific levels of the pathway Second, the detected discrepancies might be due to different methods of
Table 2 Summary of experimental results Increased levels or activation (+), decreased levels or inhibition (–) or no alteration (0) are indicated.
Gene
Regulation of gene at the following levels using methods appropriate each:
C57MG ⁄ Wnt-1 C57MG ⁄ Wnt-1 C57MG + Li C57MG + Wnt-3a C57MG⁄ Wnt-1 C57MG + Li RK13 + Li SW480 APC
Trang 8interference with the cellular metabolism The effects
caused by short-term stimuli like lithium or by the
Wnt-3a protein might differ from the effects caused by
long-term and constitutive stimulation of the Wnt
pathway like it is caused by stable transfection of
Wnt-1 This difference might be due to: promoters,
activation loops or regulatory feedback loops that are
activated only after a certain time in cells with a
con-stitutively active Wnt pathway Our results supported
the first possibility The effects caused by the
activa-tion of the Wnt pathway by transfecactiva-tion with Wnt-1
are similar to the effects caused by the extracellular
Wnt-3a protein The activities of the Lcn2, Fos and
Pedf promoters in lithium incubated cells correlate to
the detected expression trends of the corresponding
genes in lithium activated C57MG cells In SW480
cells, the pathway is constitutively active due to the
lack of functional APC When the Wnt pathway was
inhibited by transfection with APCwt the activities of
the Lcn2 and Pedf promoters increased This result
correlates to the results that were obtained when the
pathway was activated by lithium; but it is opposite to the result obtained with Wnt-1 transfected cells We conclude from these findings that the activating trigger and the level of activation within the pathway are cru-cial for the selection of a target gene and the trend of the promoter regulation
For our second aim we intended to draw general conclusions about the significance of the microarray data As the number of genes that can be analyzed experimentally by northern or western blot is limited,
we used a bioinformational approach [11] We ana-lyzed the genomic up-stream regions of all candidate targets for the presence of binding motifs for transcrip-tion factors, which play a role in the regulatranscrip-tion of Wnt target genes Generally, the highest number of relevant transcription factor binding motifs are located
in the region )500 bp to the transcription start site (TSS) [40] At the beginning, we screened only this region for transcription factor binding sites and did not find any significant accumulation of sites (Tables 3 and 4) To make sure that our analysis comprised all relevant binding motifs we decided to screen the genomic region )3000 to +100 bp relative to the TSS
We did not find a significant accumulation of TCF⁄ Lef binding sites Others have found an increased number
of TCF⁄ Lef sites in the region )1000 to )501 bp and
in the 5¢ untranslated region of positive Wnt targets in human tumours [11] We ascribe this discrepancy to the different experimental models and the different pools of candidate target genes identified by microar-ray analysis Analyzing the genomic regions, we found that the number of putative MIZ-1 binding sites is sig-nificantly increased in the region )1000 to )501 bp of down-regulated genes as compared to the same region
in control genes This outcome is a strong indication for the negative regulation of Wnt target genes via the MYC⁄ MIZ-1 cascade We are still aware that the MIZ-1 binding motifs, which we used in our screen, are yet to be confirmed by experimental data
In summary, we introduced and confirmed the novel Wnt pathway target genes Lcn2, Fos, Fabp5, Slc7a2, Ccl2, Pedf, Fas, Mrp⁄ Plf and Ramp3 The genes Lcn2, Fabp5, Ccl2, Pedf, Fas and Mrp⁄ Plf show differential expression trends, which depend on the trigger type and on the level of interference with the Wnt pathway
We conclude that the type of the trigger and the regu-lation level within the Wnt pathway decide about the gene expression pattern and thus about the kind of the specific biological answer In the second part of our study we found that the number of MIZ-1 binding sites is statistically high in the genomic regions neigh-bouring the genes, which show decreased expression levels in Wnt-activated cells
Fig 7 Exemplary results of the reporter gene assays (A) The bars
represent the relative activities of the topflash motif or the Pedf
promoter in untreated (grey) or Li+treated RK13 cells (black) (B)
The bars represent the relative activities of the Topflash motif or
the Pedf promoter in pCMV transfected (black) or APCwt
trans-fected SW480 cells (grey) Average results from three experiments
are shown Vertical lines indicate SD.
Trang 9Experimental procedures
Cell culture
Murine mammary C57MG cells can be transformed by the
retroviral infection with oncogenic Wnt-1 [13–15] The
result-ing C57MG⁄ Wnt-1 cells have been used in many studies to
characterize the Wnt pathway and to identify novel Wnt
target genes [16–20] The culture of C57MG⁄ Wnt-1 and of
C57MG⁄ PLNCx cells, which are C57MG cells transfected with an empty control vector, have been described previously [17,21] To evaluate the sustained Wnt-1 expression in C57MG⁄ Wnt-1 cells we tested the presence of the Wnt-1 transcript by reverse transcription RT-PCR We used the following primers for RT-PCR: 5¢-CATCGAGTC CTGCACCTG-3¢; 5¢-TGGGCGATTTCTCGAAGTAG-3¢ The GSK-3b can be inhibited by lithium ions (Li+), which are added to the medium of cultured cells [22,23] To test the effects of the Wnt pathway activation by Li+, we incubated C57MG cells with 20 mm LiCl for 24 h As a negative con-trol we used untreated cells or cells treated with 20 mm KCl L-M(TK-) cells (L cells) and Wnt-3a transfected L-M(TK-) cells (L Wnt-3a cells) from murine subcutaneous connective tissue were purchased from ATCC L Wnt-3a cells express and secrete biologically active Wnt-3a [24] For preparation
of conditioned media from L cells or from L Wnt-3a cells we followed the protocol from ATCC (distributed by LGC Pro-mochem, Wesel, Germany) C57MG, C57MG-pLNCx and C57MG/Wnt-1 cells were kindly provided by J Kitajewski (Columbia University, NY, USA)
Analysis of differential gene expression Total RNA was purified by standard methods from semi-confluent C57MG⁄ PLNCx and from C57MG ⁄ Wnt-1 cells Target preparation and hybridization to Affymetrix MG-U74Av2 microarrays (Affymetrix, Santa Clara, CA, USA) was performed as described [25] On the MG-U74Av2 chip, the major part of all known murine genes and 6000 murine ESTs are represented Gene chip array analysis and data evaluation were performed according to protocols provided
by the microarray manufacturer Northern blot analysis was performed to validate the results of the microarray analysis Denaturing RNA electrophoresis, blotting and detection of membrane immobilized RNA fragments by hybridization with radioactively labeled probes were done by following standard protocols and as described [26] The relative levels of proteins were measured by western blot of cell lysates or of concentrated culture medium Cell lysates were prepared by sonification after re-suspension in NaCl⁄ Piwith protease inhibitors (Roche, Mannheim, Germany) at 4C
followed by centrifugation at 16 000 g for 15 min The
supernatants were calibrated to equal protein concentrations
by dilution with NaCl⁄ Pi and analyzed by SDS⁄ PAGE and western blot Poly(vinylidene difluoride) (PVDF) mem-branes (Amersham Biosciences Europe, Freiburg, Germany) were probed with polyclonal rabbit antibodies against Lcn2 antibody (kindly provided by S.-T Chu, Institute of Biological Chemistry, Academia Sinica, Taiwan), Pedf (kindly provided by G Maik-Rachline, The Weizmann Institute of Science, Rehovot, Israel), Mrp⁄ Plf (kindly provided by M Nilsen-Hamilton, Iowa State University, Ames, IA, USA), Ccl2 (eBioscience, San Diego, CA, USA), Ramp3 (Santa Cruz, CA, USA) or b-catenin
Table 3 Number of transcription factor binding sites in genes
up-regulated in Wnt-1 transfected cells and statistical evaluation
Num-ber of transcription factor binding sites in the gene regions )3000
to +100 bp relative to the transcription start site (TSS) All known
genes identified as up-regulated by microarray analysis are
inclu-ded TO, transcription orientation; NA not analysed.
Chromosome⁄
genomic contig
TSS on
Transcription factor binding sites TCF MYC AP-1 PEA3 Gene
Mtap6 NA
Trang 10(BD-Transduction, Heidelberg, Germany) The expression
of the Mrp⁄ Plf genes in the used cell line was analyzed by
RT-PCR using the following primers: forward 5¢-CTCTG
CAGAGATGCTCCCTTC-3¢, reverse 5¢-CATGATATTTC
AGAAGCAGAGCAC-3¢ These primers border a genomic
fragment of 715 bp, which is part of the genes Plf1 and
Plf2, and also a fragment of 721 bp, which itself is part of
the genes Mrp3 and Mrp4 [27] (Fig 6) To discriminate
between the four different Mrp⁄ Plf isoforms, the RT-PCR
products were digested with the enzymes BstX1 and⁄ or
Fnu4H1 (New England Biolabs, Beverly, MA, USA) and
analyzed by agarose gel electrophoresis
Evaluation of microarray data Array images were processed to determine signals (Avg Diff) and absolute calls (Present, Absent, Marginal) for each probe set using the Affymetrix Microarray Suite 4.0 software (mas 4.0; empirical algorithm) Data on quality control of the microarray data are listed in the supplementary file (supplementary table S1) Scaling across all probe sets of a given array to an average intensity of
1000 was performed to compensate for variations in the amount and quality of the cRNA samples and other experi-mental variables of nonbiological origin Affymetrix probe
Table 4 Statistical analysis of transcription factor binding sites in the genes up-regulated in Wnt-activated cells Analysed gene region is rel-ative to the TSS; Nu, number of motifs in up-regulated candidate target genes; Tu, theoretically possible number of motifs in this region; Qu, number of motifs in the control genes; N c , possible number; T c , relation of N c to T c ; Q c , relation of Q u to Q c , Q; P, probability that the ana-lysed regions of candidate genes and control genes comprise the same number of binding sites as calculated from Z-test.
Analysed gene region
(bp)
Candidate genes (n ¼ 34)
N u
Possible number
T u
Nu⁄ T u
· 10000
Q u
Control genes (n ¼ 100)
N c
Possible number
T c
Nc⁄Tc
· 10000
Q c Q u ⁄ Q c P (Z-test) TCF ⁄ Lef
MYC
AP )1
PEA3