ABSTRACT
The importance of ornithine decarboxylase (ODC) to cell proliferation is
underscored by the complex array of cell-specific mechanisms invoked to regulate its synthesis and activity.
Misregulation of ODC has severe negative consequences on normal cell function,
including the acquisition of tumorigenic growth properties by cells
overexpressing ODC. We hypothesize that ODC gene expression is a candidate
target for the anti-proliferative function of certain tumor suppressors. Here we show that the
Wilms' tumor suppressor WT1 binds to multiple sites within the human ODC
promoter, as determined by DNase I protection and methylation interference
assays. The expression of WT1 in transfected HCT 116, NIH/3T3 and HepG2 cells
represses activity of the ODC promoter controlling expression of a luciferase
reporter gene. In contrast, WT1 expression enhances ODC promoter activity in
SV40-transfected HepG2 cells. Both the extent of modulation of ODC gene
expression and the mediating WT1 binding elements are cell specific. Constructs
expressing WT1 deletion mutants implicate two regions required for repressor
function, as well as an intrinsic activation domain. Understanding the
regulation of ODC gene expression by WT1 may provide valuable insights into the
roles of both WT1 and ODC in development and tumorigenesis.
Ornithine decarboxylase (ODC, EC 4.1.1.17), a key regulatory enzyme in polyamine
biosynthesis, is exquisitely responsive to a wide range of mitogenic agents,
including growth factors, hormones, regenerative stimuli and tumor promoters
(reviewed in
1
-
3
). This proliferation-associated enzyme is subject to a complex array of cell-specific regulatory mechanisms that govern transcription, mRNA
stability and translation and enzyme turnover (
4
-
7
). Misregulation of ODC expression has severe negative consequences on cell
growth and differentiation. The disruption of ODC function by pharmacological agents or mutagenesis results in
arrest of cell proliferation (
8
,
9
) and aberrant embryonic development (
10
). Conversely, ODC overexpression is intimately associated with cell transformation and carcinogenesis.
Levels of ODC mRNA and enzyme activity are constitutively elevated in
transformed cell lines and virtually all animal tumors (
11
,
12
). The induction of ODC is also essential for tumor promotion in a variety of
experimental models of skin, breast and colon carcinogenesis (
13
). Lastly, it has been recently demonstrated that overexpression of human ODC
confers transformation-related growth advantages, such as loss of contact inhibition, anchorage-independent growth and increased tumorigenicity, on ODC-transfected NIH/3T3 cells (
14
-
16
). Since ODC overexpression disrupts growth control, cells must be able to down-regulate ODC gene expression for normal differentiation and development to
occur. The mechanisms responsible for ODC repression are not known, but we
hypothesize that ODC gene expression is a candidate target for the anti-proliferative function of at least some types of tumor suppressor
proteins.
One such tumor suppressor, WT1, has been implicated in the etiology of Wilms'
nephroblastoma, an embryonic renal malignancy (
17
,
18
). Wilms' tumor is thought to arise from aberrant differentiation and,
ultimately, malignant transformation of abnormally persistent renal stem cells.
WT1 is a member of the early growth response (EGR) family of transcription
factors based on its binding to the EGR consensus sequence GCGGGGGCG (
19
) via highly conserved zinc finger motifs (
20
,
21
). However, WT1 has been shown to repress transcription of promoters which are
activated by EGR proteins (
22
-
24
). Recently WT1 was shown to contain transcriptional activating as well as
repressing domains (
25
,
26
). These characteristics are consistent with the proposed role for WT1 in normal
kidney development of repressing genes driving blastemal cell proliferation
and/or activating genes involved in blastemal differentiation (
27
). To date several genes involved in mesenchymal cell proliferation have been
reported to be transcriptionally repressed by WT1, including the genes encoding
insulin-like growth factor II (
28
), the A-chain of platelet derived growth factor (PDGF) (
23
), colony stimulating factor-1 (
29
), bcl-2 and c-myc (
30
).
We have cloned and characterized the human ODC gene to elucidate the molecular
mechanisms controlling ODC gene expression in normal and neoplastic cells (
31
,
32
). In this report we show that the human ODC gene promoter is a target for both
negative and positive transcriptional modulation by WT1. Both the extent of
modulation and the mediating WT1 binding elements are cell specific.
Understanding WT1 regulation of ODC gene expression may provide valuable
insight into the role of both ODC and WT1 in genitourinary development and
tumorigenesis.
The human ODC-luciferase reporter constructs phODCluc(-1491) and phODCluc(-378) and the parental plasmid pAAlucA (
32
) were kindly provided by Peggy Farnham. (University of Wisconsin). Deletion
mutant constructs phODCluc(-1215), phODCluc(-880) and phODCluc(-206) were made by digesting phODCluc(-1491) with
Spe
I, which cuts vector sequences immediately 5' of the ODC promoter insert, and either
Bgl
II,
BSS
H2 or
Mlu
I respectively, which cut within the promoter sequence, and religating plasmid
termini. Promoter constructs containing a 38 bp deletion including binding site
WT1A were generated from parental construct phODCluc(-206n WT1A). This plasmid was made by digesting phODCluc(-206) with restriction enzymes
Xma
III and
Bal
I, making the ends blunt and religating the large fragment. WT1 expression
vector CMV-WT1 and Egr1 expression vector CMV-Egr1 were as described previously (
33
). The Sp1 expression vector pSVSp1-F (
34
) and parental plasmid pSV2A101 were the generous gifts of Dr Jeffrey D.Saffer
(Batelle Pacific Northwest Laboratories). Mutant WT1 expression constructs (
25
) were generously provided by Dr Zhao-Yi Wang (Princeton University, Princeton, NJ).
Human colon carcinoma cell line HCT 116 and NIH/3T3 murine fibroblasts (American
Type Culture Collection, Rockville, MD) were maintained in Dulbecco's modified
Eagle's medium (DMEM) with high glucose and 10% fetal bovine serum (FBS). HepG2
(
35
), Nu2 (derived from a 2.2.15-induced tumor) (
36
) and G2P9T2 (
37
) cell lines obtained from Dr Judith Christman were cultured in minimal
essential medium (MEM) supplemented with Earle's salts, L-glutamine and 10% FBS. Plasmid DNA, purified either by two successive
cesium chloride density gradient centrifugations or by Qiagen-tip chromatography (Qiagen Inc., Chatsworth, CA), was transfected into
cells by the calcium phosphate protocol (
38
) using CellPhecttm transfection kits. Briefly, cells seeded in 60 mm
2
plates were transfected with 2 [mu]g of the appropriate phODCluc DNA and 2 [mu]g pCH110 DNA (Pharmacia Biotech, Piscataway, NJ) encoding the enzyme [beta]-galactosidase under transcriptional control of the SV40
early promoter. Expression constructs and the corresponding control vectors
were co-transfected as described for each experiment so that the total DNA
transfected was 20 [mu]g. After 6 h incubation with calcium phosphate-precipitated cells were shocked for 30 s with 15% glycerol in HEPES-buffered saline, pH 7.12, at room temperature. The cells were
then cultured for 40 h, washed three times with phosphate-buffered saline and lysed on the plate by adding 200 [mu]l 1* lysis reagent (Luciferase Assay System; Promega Corp., Madison, WI) containing 25 mM Tris-phosphate, pH 7.8, 2 mM dithiothreitol (DTT), 10 mM 1,2-cyclohexanediamine tetraacetic acid, 10% glycerol
and 5% Triton X-100 and incubating at 25oC for 15 min. The lysates were harvested, microcentrifuged briefly
and the supernatants collected for enzyme assays.
Luciferase activities were determined by injecting 100 [mu]l luciferin reagent (Promega Corp., Madison, WI) into 20 [mu]l cell extract and measuring the luminescence for 10 s in a Berthold
Lumat LB9501 luminometer (Wallac Inc., Gaithersburg, MD). [beta]-Galactosidase activity was assayed as described (
39
) except cell extracts were diluted with 0.25 mM Tris-HCl, pH 7.8, instead of 1* lysis reagent to avoid precipitate formation. Transfection
efficiencies were normalized with respect to [beta]-galactosidase activity. The experiments shown represent a minimum of
three independent transfections performed with at least two different plasmid
preparations.
Footprinting assays were done essentially as published (
40
) using the Hotfoot DNase I Footprinting kit (Stratagene, La Jolla, CA).
Overlapping restriction fragments from either phODCluc(-1491) or phODCluc(-378) were radiolabeled on one strand with 200 [mu]Ci [[alpha]-
32
P]dNTPs (3000 Ci/mmol) catalyzed by Klenow DNA polymerase I (Bethesda Research Laboratories, Gaithersburg, MD). The fragments were gel purified using Prep-A-Gene DNA purification matrix (BioRad). For each reaction ~80 000 c.p.m. DNA were incubated on ice for 15 min with 25-500 ng purified WT-ZF, WT-ZF(ins KTS) or Sp1 (Promega Corp.) protein
in 50 [mu]l binding buffer containing 10 mM HEPES-KOH, pH 7.5, 50 mM KCl, 10 mM ZnSO
4
, 10% glycerol, 0.1% NP-40, 1 mM DTT, 4% polyvinyl alcohol and 1 [mu]g poly(dI[middot]dC). Samples were incubated at room temperature for 2 min, 50 [mu]l DNase I buffer (Stratagene) were added and then the
samples were digested with empirically determined amounts of DNase I for 2 min.
The reaction was terminated by addition of 100 [mu]l stop buffer (Stratagene) and extraction with an equal volume of
phenol/chloroform (1:1). Samples were precipitated with 2.5 vol. ethanol in the
presence of 0.3 M NaOAc, washed with 70% ethanol and resuspended in formamide
dye (Stratagene), then run on a 8% acrylamide-7 M urea gel. Sequencing reactions were run as size markers. The gels
were dried and exposed to X-Omat film.
The binding of WT-ZF to ODC promoter regions WT1-A and WT1-F was analyzed by the methylation interference assay as
described (
41
). Briefly, ODC promoter DNA fragments
Bam
HI-
Bgl
I (-1346 to -1222) and
Mlu
I-
Nco
I (-250 to -50) were
32
P-radiolabeled at one end,
Bam
HI and
Mlu
I respectively, gel purified and methylated according to the protocol. Each DNA
was then subjected to a preparative electrophoretic mobility shift assay (EMSA)
to isolate WT-ZF bound and unbound probe for binding site analysis. Approximately 2 * 10
5
c.p.m. probe DNA was incubated with 250 ng WT-ZF protein in 25 mM HEPES-KOH, pH 7.5, 50 mM KCl, 10 mM ZnSO
4
, 10% glycerol, 0.1% NP-40, 0.2% bovine serum albumin, 1 mM DTT, 2 [mu]g poly(dI[middot]dC) for 20 min at room temperature. Bound and unbound DNA
probes were fractionated on a non-denaturing 5% acrylamide gel in 0.5* TBE buffer, visualized by autoradiography and eluted from the gel.
DNA was then cleaved with 1 M piperidine, heated to 95oC for 30 min and repeatedly frozen and lyophilized to remove residual
piperidine. The samples were then electrophoresed in 8% acrylamide-7 M urea sequencing gels and exposed to X-Omat film.
To determine whether WT1 modulates ODC gene expression HCT 116 human colon
carcinoma cells and NIH/3T3 murine fibroblasts were co-transfected with WT1 expression vector CMV-WT1 (
32
) and gene fusion constructs phODCluc(-1491) and phODCluc(-378). These constructs contain 1491 and 378 bp respectively of ODC
promoter DNA and 77 bp of exon 1 upstream of a luciferase reporter gene. ODC
promoter function was quantitated by measuring luciferase activity in crude
lysates 48 h post-transfection. WT1 represses the ODC promoter activities exhibited by
phODCluc(-1491) and phODCluc(-378) in HCT 116 cells to levels that are 24 and 43% respectively
of controls (Fig.
1
). In comparison, co-expression of WT1 in NIH/3T3 cells reduced the ODC promoter activities
from these two constructs to virtually the same extent, which was ~70% of controls. Luciferase activity levels were not inhibited by WT1 in
control transfections using parental plasmid pAAlucA or pAAlucA controlled by the SV2 early promoter (data not shown). These results indicate
that the human ODC gene promoter is a target for the
trans
-repressing activity of the WT1 tumor suppressor. Also, the extent of
repression by WT1 and the mediating
cis
-elements utilized appear to be cell specific.
WT1 binding elements in the ODC promoter were identified by DNase I protection
footprinting.
32
P-Radiolabeled DNA restriction fragments extending from nt +77 in exon 1 to
nt -1491 in the ODC promoter were incubated with varying concentrations of WT-ZF (
33
) or WT-ZF(ins KTS) proteins (
24
). WT-ZF contains the DNA binding zinc finger region of wild-type WT1 synthesized from an
Escherichia coli
expression vector and purified by nickel chelate affinity chromatography. WT-ZF(ins KTS) protein, also synthesized
in vitro
, contains the naturally occurring insertion of lysine, threonine and serine
within the zinc finger domain of WT1 which interrupts binding to the consensus
element GCGGGGGCG. DNase I digestion resulted in protection of four sites with relatively high binding affinities for WT-ZF (WT1-A, -D, -E and -F) and several sites with lower affinities (Fig.
2
). WT-ZF(ins-KTS) does not bind to sequences in the ODC promoter. Two WT-ZF protected regions, WT1-A and WT1-F, are GC-rich, each with two possible WT1 binding
sites and overlapping Sp1 consensus sequences (Fig.
3
). We performed methylation interference studies to determine more precisely the
guanine bases contacting WT-ZF within footprints WT1-A and WT1-F (Fig.
4
). The ODC promoter sequences protected by WT-ZF, the nucleotide positions of each element relative to the transcription
start site and the methylated bases in WT1-A and WT1-F that most severely disrupt WT-ZF binding are shown in Figure
3
.
5'-Terminal deletions were generated from phODCluc(-1491) to produce a panel of subclones each containing 77 bp
of exon 1 and extending 1490, 1215, 880, 378 and 206 bp respectively into the
ODC promoter. These constructs were transfected into HCT 116 cells to ascertain
the relative abilities of the major WT1 binding elements to mediate WT1
repression of the ODC promoter (Fig.
5
A). WT1-A and WT1-F mediate inhibition of ODC promoter activity by 62 and ~15% respectively. There is no significant statistical
difference in promoter function of constructs extending -206, -378, -880 and -1215 nt into the ODC promoter (
P
> 0.15 for each pair). The difference between constructs extending -1490 and -1215 nt is statistically significant (
P
= 0.012). Based on these studies it appears that binding sites WT1-A (-94 to -116) and WT1-F (-1231 to -1247) are responsible for most of the
inhibition of ODC promoter activity by WT1 in HCT 116 cells.
Each WT1 footprint mapped to the ODC promoter includes a putative Sp1 binding
site which exactly or nearly so matches the consensus sequence GGGCGG (Fig.
3
). Interestingly, there are multiple Sp1 elements within the proximal ODC
promoter that are not protected from DNase I digestion by WT1 (data not shown).
Overlapping WT1 and Sp1 binding sites suggested that competition for binding by
transactivating factor Sp1 could regulate the effect of WT1 on promoter
function. Using the functional WT1A binding site with a perfect Sp1 consensus
element we examined the possibility that Sp1 and WT1 competition influences
transcription from the proximal ODC promoter. We first determined whether Sp1
does, in fact, bind to sequences within WT1-A. DNase I protection footprinting shows that purified Sp1 protein binds
and protects a subset of WT1-A sequences (-98 to -116) (Fig.
6
). AP-2, which also binds GC-rich sequences, does not bind to WT1-A (data not shown). Because the patterns of protected
sequences generated by these two proteins are distinguishable, we were able to
perform competitive footprinting assays to examine the relative binding
affinities of WT1 and Sp1 for WT1-A. The simultaneous incubation of Sp1 and WT1-ZF with ODC promoter DNA produced an Sp1 protection pattern with
Sp1:WT-ZF ratios of from 0.5 to 5 (Fig.
6
). Thus the binding of WT1 and Sp1 to WT1-A is mutually exclusive and Sp1 competes more efficiently than WT-ZF for binding to WT1-A
in vitro
.
Since WT1 and Egr1 bind the same consensus sequence, we examined the effect of
Egr1 expression on WT1-A regulation of ODC promoter activity. HCT 116 and NIH/3T3 cells were co-transfected with phODCluc(-206) and CMV-Egr1 expression vector (Fig.
7
B). Egr1 expression had contrasting effects on ODC promoter activity in these
two cell lines. In HCT 116 cells ODC promoter activity from phODCluc(-206) was repressed to 64% of controls not transfected with CMV-Egr1. In comparison, Egr1 expression enhanced ODC promoter activity
from this same construct by 50%. Thus, as seen in HCT 116 cells, both WT1 and
Egr1 are capable of down-regulating ODC promoter activity via the same proximal binding element. It
is also clear from the disparate results with WT1 and Egr1 in NIH/3T3 cells
that the two factors are differentially regulated and have opposing effects on
ODC promoter activity within a given cell type.
We transfected parental HepG2 hepatoblastoma cells and HepG2 cells stably
transfected with genomic DNA from either hepatitis B virus (designated Nu2
cells) or SV40 DNA tumor virus (designated G2P9T2 cells) with phODCluc(-1491) and phODCluc(-378) and CMV-WT1 to examine the generality of ODC gene repression by WT1.
WT1 expression represses phODCluc(-1491) promoter activity by 50 and 37% in HepG2 and Nu2 cells respectively
(Fig.
8
). Distal elements contribute to down-regulation of the ODC promoter in these two cells to varying extents, but
based on the repression of phODCluc(-378) it appears that most of the effect of WT1 is mediated by a proximal
cis
-element(s). In contrast to HepG2 and Nu2 cells, ODC promoter activity was
elevated 200% in G2P9T2 cells by WT1 co-expression. This represents a 4-fold change in ODC promoter activity relative to the ODC promoter response to WT1 expression in parental HepG2 cells. The activation of ODC promoter activity
was mediated primarily by a proximal WT1 binding element(s). Thus the ODC
promoter is subject to activation by WT1, as well as repression.
WT1 loss of function has been strongly implicated in a variety of genitourinary
developmental anomalies, as well as in the etiology of Wilms' nephroblastoma (
42
). However, the exact role WT1 plays in normal development and tumorigenesis is
not well understood. The identification of target genes regulated by WT1 is
crucial to understanding the consequences of WT1 function at the molecular and
cellular levels. We have shown that the human ODC gene promoter is a target for
both negative and positive transcriptional modulation by WT1 in transient co-transfections. WT1 binds to multiple DNA
cis
-elements within the ODC promoter and represses ODC transcription in HCT
116 colon carcinoma, HepG2 hepatoblastoma and NIH/3T3 fibroblast cell lines. In
contrast, ODC promoter activity is enhanced by WT1 expression in SV40-transfected HepG2 cells.
Our results do not demonstrate that WT1 regulates the ODC promoter
in vivo
. However, the importance of ODC and polyamines to cell proliferation is widely
reported and is consistent with the hypothesis that WT1 may act during normal
development to suppress the expression of genes controlling cell proliferation
(
27
). Failure to repress ODC due to loss of WT1 function may block blastemal cell
differentiation by interfering with the cessation of proliferation. However,
WT1 regulation of ODC may have a more direct impact on blastemal cell
differentiation. There is limited but convincing evidence that polyamine
metabolism is critical for some aspects of differentiation and embryogenesis (
10
). However, the cellular effects of polyamines are complex, as both increases
and reductions in ODC and polyamine levels have been shown to induce cell
differentiation, depending on the experimental system. At present the role of ODC in organogenesis and, specifically, in the mesenchymal-epithelial transition common to genital and urinary tract formation is unknown. Focusing on the regulation of ODC by WT1,
which is specifically expressed in embryonic kidney, gonad and mesothelium, may
help define the relationship of ODC to genitourinary development.
The ODC gene is only weakly expressed in most adult tissues, as reflected in the
extremely low abundance of ODC mRNA and protein in non-proliferating tissues. Furthermore, the ODC promoter appears to be
repressed in differentiated or quiescent cells, as suggested by studies of
transgenic mice carrying the human ODC gene (
43
) and mammalian cells transiently transfected with ODC promoter/reporter fusions
containing various lengths of the ODC gene promoter (
44
,
45
; M.Flanagan, personal communication). We hypothesize that the ODC promoter in
most terminally differentiated cells exists in a repressed state. In support of
this several studies show that stimulation of ODC expression in quiescent cells
can occur without prior protein synthesis, suggesting involvement of short-lived repressor proteins in ODC gene regulation (
46
,
47
; M.Flanagan, personal communication). Different cell types may utilize a
variety of repressors which function similarly to WT1 but which are themselves
regulated in a tissue-specific manner to control ODC expression. From this study WT1, Egr1 and,
to a limited extent, Sp1 each affect down-regulation of ODC promoter activity from the WT1A element. In addition, Li
et al.
(
47
) have reported that multiple nuclear proteins, including Sp1 and a repressor
termed NF-ODC
1
, bind to sequences in the ODC promoter corresponding to the WT1A element.
Similarly to WT1 and EGR1 in this study, Li
et al.
have shown that the response of the ODC promoter to Sp1 and NF-ODC
1
varies with cell type. These observations, coupled with growing evidence that
the multiple members of the EGR/WT1 (
24
) and Sp1 (
48
) families of transcription factors are subject to different cell-specific and growth-specific regulation, suggest that WT1-A may serve as a `hot-spot' for mediating ODC repression.
WT1 has been shown to repress transcription via a functional repressor domain (
22
,
25
) or by competing for either Sp1 (
31
) or Egr1 binding (
23
,
28
). We examined the possibility that WT1 regulation of ODC gene expression
involved competition for binding to DNA element WT1-A between WT1 and either Sp1 or Egr1 transcription factors. Sp1 exhibited
a higher affinity for binding to WT1-A than WT-ZF
in vitro
, but the results of co-transfection of Sp1 and WT1 expression vectors in two cell lines indicated
that Sp1 did not compete efficiently with WT1
in vivo
. The difference in the results of the two assays may simply reflect the fact
that the
in vitro
binding studies utilized truncated WT-ZF protein, whereas entire WT1 was expressed in the transfection studies.
Alternatively, competition between ectopically expressed Sp1 and WT1 for
binding to this element may not be responsible for the differential response of
the ODC promoter to WT1 in NIH/3T3 and HCT 116 cell lines. That multiple
endogenous factors bind to the region including WT1A and modulate ODC promoter
function in a cell-specific manner (
47
) suggests the possibility that any one or more of these unidentified proteins
may interact differentially with WT1. With regard to Egr1, the inhibition in
HCT 116 cells and activation in NIH/3T3 cells of ODC promoter function by Egr1
expression suggests that competition by Egr1 and WT1 for binding to WT1A could
play a part in the differential response of the ODC promoter to WT1.
The enhancement of ODC promoter activity by WT1 expression in G2P9T2 cells
suggests that other factors may disrupt or alter WT1 repressor function.
Mutations in cellular factors that modulate WT1 function could result in
blastemal cell transformation despite the presence of wild-type WT1 and may explain why more WT1 mutants have not been identified in
Wilms' tumors. The positive and negative modulation of ODC expression by both WT1 and
Egr1 underscore the importance of understanding the interactions of WT1 with putatative regulatory sequences. Our studies of
mutant WT1 in HepG2 and G2P9T2 cells suggest the presence of a candidate
regulatory domain between amino acid residues 85 and 98 of WT1 that may be
responsive to factors capable of modulating repressor function. The loss of N-terminal residues 1-84 had no effect on either WT1 repression or activation of ODC
promoter activity in HepG2 and G2P9T2 cells respectively. However, removal of
the next 13 residues resulted in the loss of WT1 repression in HepG2 cells.
That residues 85-98 are important, but not sufficient, for WT1 repressor activity is
indicated by the results obtained with mutant WT1n 179-294. This mutant exhibits the highest levels of WT1 activating function
in both cell lines, despite containing an intact region surrounding and
including residues 85-98. These findings suggest that the deleted region 179-294 in this construct must also function in conjunction with
region 85-98 for maximum WT1 repressor activity.
The mapping of WT1 domains with the ODC promoter in HepG2 and G2P9T2 cells
agrees to a certain extent with a similar study using the PDGF-A promoter in NIH/3T3 cells (
25
). Both studies show that the N-terminal 84 amino acids do not contribute to transcriptional modulation by
WT1. In addition, Wang
et al.
(
25
) conclude that the WT1 repressor domain is located between amino acid residues
84 and 179. Our studies show an essential and modulatable repressor domain
between amino acids 85 and 97. The differences in assigning the boundaries of this domain are probably due to
the different set of WT1 mutants utilized in the two studies. The two studies
differ significantly, however, with regard to defining the boundaries of the
WT1 activation domain. Wang
et al.
showed that residues 180-294 are required for activation of the PDGF-A promoter, whereas this study implicates residues 98-179 in activation of the ODC promoter. The discrepancy may
be partially explained by differences in the broadly mutated WT1 constructs
used. It is more likely, however, that these differences reflect cell-specific and/or promoter-specific differences in transcription factors that modulate WT1
activity. In any case, activation of the ODC promoter by wild-type WT1 in G2P9T2 cells is matched by expression of mutant WT1 n1-294 containing only the DNA binding domain. Thus the intrinsic
activation domains within WT1 are not responsible for stimulating ODC promoter
function in these cells. Rather, the difference in ODC promoter response to WT1
in HepG2 and G2P9T2 cells may be due to the physical interference of promoter-bound WT1 with the binding and/or function of an unknown, cell-specific repressor.
We have demonstrated that the protein product of the WT1 tumor suppressor gene
modulates transcription of the ODC promoter in transient transfection assays.
Clearly this approach provides important insight into the mechanisms underlying
the differential effects of WT1 on a responsive promoter. However, the
physiological significance of the interaction between WT1 and ODC is not clear.
The identification of candidate target genes, such as ODC, should provide a
useful direction to ongoing investigations seeking to clarify the role of WT1
in the complex processes leading to both normal and abnormal genitourinary
development.
This research was supported by grant 6-FY94-0278 from The March of Dimes Foundation for Birth Defects, The
Department of Internal Medicine, Funds for Medical Research and Education,
Wayne State University School of Medicine and by grants CA52009 and CA47983,
core grant CA10815 from the National Institutes of Health and grants from the
W.W.Smith Charitable Trust, the Hansen Memorial Foundation and the Mary
A.H.Rumsey Foundation. FJR is a Pew Scholar in the Biomedical Sciences.
REFERENCES
Return