The Human Sex-determining Gene SRY Is a Direct Target of WT 1 *

The product of the Wilms’ tumor gene, WT1, is essential for male sex determination and differentiation in mammals. In addition to causing Wilms’ tumor, mutations in WT1 often cause two distinct but overlapping urogenital defects in men, Denys-Drash syndrome and Frasier syndrome. In this study we investigated the regulation of the sex determination gene SRY by WT1. Our results showed that WT1 up-regulates the SRY gene through the proximal early growth response gene-1-like DNA-binding sequences in the core promoter. Mutant WT1 proteins in Denys-Drash syndrome patients were unable to activate this promoter. These mutants did not act in a dominant negative manner, as expected over the wild-type WT1 in this promoter. We also found that WT1 could transactivate the endogenous SRY gene. These observations, together with the overlapping expression patterns of WT1 and SRY in human gonads, led us to propose that WT1 regulates SRY in the initial sex determination process in humans and activates a cascade of genes ultimately leading to the complete organogenesis of the testis.

The product of the Wilms' tumor gene, WT1, is essential for male sex determination and differentiation in mammals. In addition to causing Wilms' tumor, mutations in WT1 often cause two distinct but overlapping urogenital defects in men, Denys-Drash syndrome and Frasier syndrome. In this study we investigated the regulation of the sex determination gene SRY by WT1. Our results showed that WT1 up-regulates the SRY gene through the proximal early growth response gene-1-like DNA-binding sequences in the core promoter. Mutant WT1 proteins in Denys-Drash syndrome patients were unable to activate this promoter. These mutants did not act in a dominant negative manner, as expected over the wild-type WT1 in this promoter. We also found that WT1 could transactivate the endogenous SRY gene. These observations, together with the overlapping expression patterns of WT1 and SRY in human gonads, led us to propose that WT1 regulates SRY in the initial sex determination process in humans and activates a cascade of genes ultimately leading to the complete organogenesis of the testis.
The Wilms' tumor gene WT1 is involved in tumorigenesis in humans, and its mouse homolog, Wt1, has a distinct role in the development of several organs during embryogenesis (1,2). The gene was originally identified by positional cloning and sequencing on the basis of its association with the rare childhood kidney tumor, Wilms' tumor (3)(4)(5). Specific mutations in the Wilms' tumor gene cause two different types of genitourinary abnormalities called Denys-Drash syndrome (DDS) 1 (6) and Frasier syndrome (7). Patients often showed male to female sex reversal, male pseudohermaphroditism, and cryptorchidism.
WT1 is a zinc finger containing DNA-binding protein and acts as a transcriptional activator or repressor depending on the cellular or chromosomal context (8,9). It has four major isoforms, due to the insertion of three amino acids (KTS) between zinc fingers 3 and 4, and the insertion of an alternatively spliced 17-amino acid segment encoded by exon 5 in the middle of the protein (Fig. 1A) (10). These four isoforms are conserved among mammals. WT1 binds to the highly GC-rich canonical early growth response gene-1 (EGR-1) DNA-binding motif GCGGGGGCG (8). Many genes are proposed to be regulated by WT1. Most of them are involved in growth regulation and cellular differentiation as a result of the tumor suppressor action of WT1 (11)(12)(13). A second group is involved in sex development and includes Mü llerian inhibiting substance (Mis) (14). Nachtigal et al. (14) reported that SF1 and WT1 synergistically activate the Mis promoter. DNA binding of WT1 is not essential for this synergistic activation. Mutations in MIS or its receptor cause male pseudohermaphroditism in humans, and deletion of Mis or the Mis receptor also causes pseudohermaphroditism in mice (15,16). Thus the regulation of MIS by WT1 has physiological effects and suggests that pseudohermaphroditism appears in DDS patients as a result of deregulation of MIS. Another common trait of DDS is sex reversal (17), and our objective is to identify the cause of sex reversal in DDS patients.
One possible target gene of WT1 is the sex-determining gene SRY. SRY is an HMG box containing transcription factor. It is expressed during the critical period of male sex determination and is able to direct male sex determination in a genetically XX mouse (18). Little is known about the regulation of SRY during sex development (19). The level of expression of Sry in the mouse is important for sex determination (20). Characterization of the regulation of SRY has been hampered by the poor conservation of the upstream sequences of SRY across species. It is known that Sry is expressed only for a short period during male sex determination in the Sertoli's cell precursor in mouse (21)(22)(23). However, SRY expression in humans is less tissueand stage-specific (24,25). Gonadal expression of both WT1 (25,26) and SRY (25) overlaps in humans. WT1 first appears at a low level at 31 days post-ovulation in the gonadal ridges of both male and female embryos. Expression levels significantly increase at 41 days post-ovulation, when SRY expression begins in the gonadal ridge. SRY is most strongly expressed at 44 days post-ovulation, and both WT1 and SRY expression persist later in fetal development (25,26). Our hypothesis is that because mutation of both genes causes sex reversal in humans and WT1 is expressed before SRY in the gonadal ridge, WT1 may regulate the expression of SRY. Our preliminary data also showed that the SRY promoter is positively regulated by WT1 (27). In this report, we provide clear evidence that WT1 activated the SRY gene and initiated a regulatory gene cascade in the male sex determination and differentiation pathway.

EXPERIMENTAL PROCEDURES
Plasmids-The expression vectors with the four isoforms of WT1 (pCBWT1Ϫ/Ϫ, ϩ/Ϫ, Ϫ/ϩ, and ϩ/ϩ) were described previously (28). pcDNA3WT1 constructs were made by PCR amplifications of wild-type WT1 from CMVWT1Ϫ/Ϫ expression vector. The resulting PCR products were subsequently subcloned into a modified pcDNA3 vector (29) harboring the 5Ј-untranslated region of the herpes simplex virus-thymidine kinase gene. All WT1 mutants were created from this pcDNAWT1Ϫ/Ϫ background by site-directed mutagenesis, and mutations were verified by sequencing.
The plasmids in which the SRY upstream sequence drove transcription of the luciferase gene, a 1979-and 313-base pair fragment from SRY-CAT1 and SRY-CAT2 (30), were isolated and cloned in front of the luciferase gene pGL3 basic vector (Promega) and were designated Ϫ1938SRYP and Ϫ272SRYP, respectively. The SRY promoter deletion constructs were generated by PCR amplification using Ϫ272SRYP as template. Resulting PCR fragments were subcloned into the pGL3 basic vector. All constructs were verified by sequencing from both directions.
All site-directed mutagenesis of the SRY promoter constructs was performed with the Quick Change site-directed mutagenesis kit (Stratagene). Details of PCR primers used for mutagenesis can be obtained from the authors. All mutations were confirmed by sequencing from both directions.
Cells were seeded at a density of 50,000 -70,000 cells/well in 12-well plates 16 -18 h before transfection. The cells were cotransfected with expression and reporter plasmids as indicated in the figure legends. CMV-␤-galactosidase (10 ng) was cotransfected as an internal control to normalize the transfection efficiency. The transfections were carried out using LipofectAMINE-Plus reagent according to manufacturer's recommendations (Life Technologies, Inc.), and the cells were harvested after 40 -48 h. Luciferase activity was measured with a luciferase assay kit (Tropix) and a Lumat LB9507 luminometer (EG & G Berthold). ␤-Galactosidase was measured with the Galacto-Light plus kit (Tropix).
Gel Shift Assay-Gel shift reactions were performed in a total volume of 20 l on ice. Radiolabeled probes were prepared by end labeling with [␥-32 P]ATP, and 100 pmol of each labeled probe and 2.5 l of in vitro translated (IVT) protein were used for each reaction. For competition with wild-type or mutant oligonucleotides, a 100-fold excess of unlabeled oligonucleotides was added to the reaction mixture before addition of the labeled probe. Thirty minutes later, the reaction mixture was loaded onto a 5% polyacrylamide gel in Tris glycine buffer, and electrophoresis was performed at 150 V for 3 h. In the supershift with antibody assay, reaction mixture without labeled probe was incubated with 2.0 g of anti-WT1 antibody (C-19, Santa Cruz Biotechnology) for 15 min at room temperature, and labeled probe was added with further incubation on ice for 30 min.
Generation of Cells with Inducible WT1 Expression-A full-length FIG. 1. WT-KTS isoforms specifically activated the SRY promoter in different cell lines. A, isoforms of WT1. Alternative splicing of exon 5 and the alternative usage of two different splice-donor sites at the 3Ј end of exon 9 produce four different splice forms. ϩ and Ϫ indicate inclusion and exclusion, respectively, of the sites. B, NT2D1 and TM4 cells were transfected with 0.2 g of the promoter from human SRY driving the luciferase reporter constructs and 0.2 g of pCB6WT1Ϫ/Ϫ or the empty expression vector, pCB6ϩ, as a control. Two forms of the SRY promoter were used, one containing nt Ϫ1938 to ϩ41 (Ϫ1938SRYP) and the other containing nt Ϫ272 to ϩ41 (Ϫ272SRYP). Ten nanograms of CMV promoter-driven ␤-galactosidase gene, CMV-␤gal (Promega), was cotransfected with each sample to control the transfection efficiency. The assay was performed 40 h after transfection. All results are expressed as mean Ϯ S.D. of at least three experiments. C, HeLa cells were transfected with different amounts of pCB6.WT1Ϫ/Ϫ with Ϫ272SRYP and 10 ng of CMV-␤gal. Empty pCB6ϩ was added to keep the plasmid concentration to 0.4 g. The assay was performed 40 h after transfection. All results are expressed as mean Ϯ S.D. of at least three experiments. D, HeLa cells were transfected with different isoforms of WT1 in pCB6 with Ϫ272SRYP and 10 ng of CMV-␤gal. The assay was performed 40 h after transfection. All results are expressed as mean Ϯ S.D. of at least three experiments.
WT1 construct was assembled from pcDNA3WT1, cloned downstream from the tetracycline-regulated promoter in pEC1214A (31), and confirmed as wild-type by sequencing the entire coding region. NT2D1 cells were used to generate stably transfected clones with tightly regulated inducible expression of wild-type WT1.
NT2D1 cells stably transfected with a tetracycline-repressible WT1 expression vector and control vector were grown to about 70% confluency. Tetracycline analog doxycycline was removed from the cells. The cells were harvested after 12 h, and total RNA was isolated with the RNeasy mini Kit (Qiagen). RT was performed with random hexamers and the Superscript II kit (Life Technologies, Inc.). The products from the RT were aliquoted and directly subjected to PCR amplification. The following forward (F) and reverse (R) primers were used for PCR amplification: SRY-F GAAGATCAGGGGCTGGCAGA), SRY-R (CAGAG-GCGCAAGATGGCTCT), GAPDH-F(CCAGCCGAGCCACATCGCTC), and GAPDH-R (ATGAGCCCCAGCCTTCTCCAT). PCR was performed with the high fidelity PCR kit (Roche Molecular Biochemicals), and the following conditions were used: 1 cycle at 94°C for 2 min and 22 cycles of denaturation at 94°C, annealing at 60°C, and extension at 72°C. PCR products were separated on a 1.5% agarose gel.

Ectopic Expression of WT1 Activates the SRY Promoter in Different Cell
Lines-To test whether the SRY promoter is regulated by WT1, we used two luciferase reporter gene constructs. One of these, Ϫ1938SRYP, contains a large region upstream of the SRY promoter, and the other, Ϫ272SRYP, contains the minimal promoter and 5Ј sequence. Both constructs have similar promoter activities in TM4 and Ltk cells (30). Fig. 1B shows that both Ϫ1938SRYP and Ϫ272SRYP were activated severalfold in response to WT1Ϫ/Ϫ (Fig. 1B) in the human teratocarcinoma cell line NT2D1 derived from testicular tumor and the mouse Sertoli cell line TM4. The NT2D1 and TM4 cell lines have been shown to express most of the genes known to be involved in mammalian sex determination (32). For further studies, we chose the cervical carcinoma cell line HeLa, because most (if not all) male sex-specific genes and WT1 are not expressed in HeLa cells, and HeLa is highly transfectable. We found that the basal level of reporter gene activity was very low in HeLa cells compared with NT2D1 and TM4 cells. Both Ϫ1938SRYP (data not shown) and Ϫ272SRYP were robustly activated (Ͼ50-fold over the control) by WT1Ϫ/Ϫ in HeLa cells in a dose-dependent manner (Fig. 1C). We used Ϫ272SRYP for most of our subsequent studies, because both Ϫ1938SRYP and Ϫ272SRYP showed similar transactivational activities. The transactivation of the SRY promoter was dependent on the ϪKTS isoforms, WT1Ϫ/Ϫ and WT1ϩ/Ϫ. Both ϩKTS isoforms, WT1ϩ/ϩ and WT1Ϫ/ϩ, were unable to transactivate the SRY promoter (Fig. 1D). These results are consistent with the hypothesis that ϪKTS isoforms are involved in transcriptional activity, whereas ϩKTS isoforms are associated with the spliceosome and are involved in the regulation of certain genes at the post-transcriptional level (33). These data also clearly indicate that WT1 is able to transactivate the human SRY promoter.
Activation of the SRY Promoter Required nt Ϫ32 to ϩ41-To determine the location of the WT1-responsive element in the SRY promoter, we made a series of deletion constructs as follows: Ϫ272SRYP, Ϫ221SRYP, Ϫ147SRYP, Ϫ71SRYP, and Ϫ32SRYP ( Fig. 2A). All of these constructs were similarly activated by WT1 (Fig. 2B). However, empty pGL3 basic vector was not activated over the basal level by WT1 (Fig. 2B). The deletion constructs mapped the WT1-responsive element (WTE) to the region between nt Ϫ32 and ϩ41, within the SRY promoter.
WT1-responsive Element Is Essential for the Activation of the SRY Promoter-Analysis of this 73-base pair minimal WT1responsive region revealed only one EGR-1-like potential WT1responsive element (GAGGGGGGTG) (Fig. 2A). To test the functional significance of this site, single mutations were introduced at several positions (Fig. 3A). These constructs were assayed for their responsiveness to WT1 in HeLa cells and compared with wild-type Ϫ272SRYP. Substitution of any G to T did not significantly reduce the reporter gene activity (data not shown). Single nucleotide substitution of the A to a G (M2) had little or no effect on reporter gene activity. On the other hand, substitution of the T with a G (M1) was reduced in reporter gene activity by 50%. Moreover, substitution of the both the A and T to Gs (M3) completely blocked transactivation by WT1.
The WT1-responsive sequence was further characterized by gel shift assays. The gel shift assay in Fig. 4 shows that WT1 synthesized in vitro binds to a 32 P-labeled WTE oligonucleotide probe. Binding was competed completely by unlabeled wildtype probe and also by unlabeled EGR-1 probe (Fig. 4, lanes 3  and 5). The mutants M1 and M2, which showed partial or no reduction in reporter gene activity, also competed with the wild-type probe (Fig. 4, lanes 8 and 9). The mutant M3 in which both the A and T were changed to Gs did not compete with the wild-type probe and has no transactivation abilities in reporter gene assays. This indicates that both A and T residues at positions 2 and 8 were very important for WT1-mediated transactivation of the human SRY promoter. As a positive control, a 32 P-labeled EGR-1 consensus probe was used in a WT1-mediated DNA binding assay (Fig. 4B, lane 6). Antibody against WT1 supershifted WT1-DNA complexes, indicating that WT1 was present in that complex. These data suggest that the WTE sequence in the human SRY promoter is essential for the binding of WT1 to this promoter as well as for WT1-mediated transactivation of the SRY promoter.

Different WT1 Mutants Had Different Effects on the SRY Promoter-
The DDS phenotype often arises as a result of alteration of one allele by a missense point mutation, usually in the zinc finger DNA binding domain of WT1. One mutational hot spot is codon 394, in which arginine 394 is changed to tryptophan. To determine how these mutations affect the SRY promoter, we created two types of mutations. One group of mutants contained changes in the zinc finger region (C330Y, R366C, R366H, H377Y, and R394W), and the other group contained (F154S, P180S, and S273A) changes outside that region. These mutants were tested for their ability to transactivate the SRY promoter in HeLa and NT2D1 cells. The most common DDS mutants R366C, R366H, H377Y, and R394W failed to activate the SRY promoter in reporter gene assays. In contrast, the mutants with mutations outside the zinc fingers region (F154S, P180S, and S273A) had transactivational potentials similar to that of wild-type WT1. Another mutant (C330Y) retained some transactivational ability even though its mutation was in the zinc finger region. Western blot data from HeLa cell extracts showed that all the mutant proteins were expressed at similar levels (Fig. 5C).
The DNA binding abilities of these mutants were assessed by gel shift assays. Mutations in the DNA binding region (the zinc finger regions) caused loss of DNA binding ability (Fig. 6). However, the mutant C330Y, which has a mutation in the DNA binding region, retained some transactivational ability as well as DNA binding ability (Fig. 6). This implies that the cysteine in codon 330 in WT1 is not crucial for DNA binding, although the mutant did lose some of its activity. On the other hand, mutations outside the DNA binding region did not have major effects on DNA binding to the SRY promoter (Fig. 6).
DDS Mutants Failed to Act in a Dominant Negative Manner on the SRY Promoter-To determine the molecular mechanisms underlying the pathogenesis of DDS, we determined whether DDS mutants act in a dominant negative manner by sequestering wild-type WT1. There are conflicting reports about this issue (14,34). We tested this hypothesis in the context of SRY promoter activation. HeLa cells were cotransfected with the wild-type WT1Ϫ/Ϫ with or without a 5-fold excess of the different mutants in the presence of Ϫ272SRYP. The mutants (F154S, P180S, S273A, and C330Y) that had transactivation ability had additive effects both in HeLa and NT2D1 cells (Fig. 7). However, the DDS mutants (R366C, R366H, H377Y, and R394W) did not have additive or dominant negative effects, i.e. there was no significant reduction of promoter activation by wild-type WT1 (Fig. 7). This experiment was repeated at least three times with different plasmid preparations, and similar results were obtained. These data clearly indicated that the DDS mutants did not act in a dominant negative manner, at least in the activation of the SRY promoter.
Increase in Endogenous SRY Expression by the Inducible Expression of WT1 in NT2D1 Cells-To investigate the effect of WT1 on endogenous SRY expression, we constructed a system for tetracycline-regulated inducible expression of WT1Ϫ/Ϫ in the human teratocarcinoma cell line NT2D1. We chose NT2D1 because it expresses SRY at very low levels but does not express WT1 (Fig. 8A). Also, we assumed that other components for the expression of SRY are present in this cell line. It has been reported that some clones of NT2D1 constitutively express high amounts of SRY (35). However, our lines constitutively expressed low amounts. WT1Ϫ/Ϫ was placed under the control of the tetracycline transactivator. In the presence of the tetracycline analog doxycycline, WT1 expression was turned off. Withdrawal of doxycycline from the medium caused strong expression of WT1, as determined by Western blotting (Fig.  8A). Fig. 8B shows that NT2D1 cell expressing WT1 did upregulate SRY mRNA, as determined by RT-PCR. To confirm these results, we performed Western blottings using the whole cell extracts of the induced and uninduced NT2D1 cells and anti-SRY antibody. As expected, WT1-expressing NT2D1 cells expressed severalfold more SRY protein (Fig. 8C). These results showed clearly that WT1 was capable of up-regulating the endogenous SRY gene in its native chromosomal context. DISCUSSION The molecular mechanisms underlying the pathogenesis of DDS and how WT1 participates in the sex determination pathway have not been well understood. That heterozygous germ line mutations of WT1 cause DDS in humans has been known for a long time. Numerous genes have been proposed as target genes of WT1, but the physiological relevance of those genes remains unclear. By a candidate gene approach, we identified SRY as one of the direct target genes of WT1. SRY fulfills the criteria of a direct target of WT1 in the sex determination pathway for the following reasons. First, mutations in both SRY and WT1 cause male to female sex reversal in genetically XY individuals. Second, the expression patterns of SRY and WT1 overlap in the human gonads. Third, WT1 can positively regulate the SRY promoter and can directly bind to a cis-acting element in this promoter. Finally, WT1 can directly transactivate endogenous SRY expression in its native chromosomal context. Thus, we conclude that SRY is a physiologically relevant target of WT1 in the sex determination pathway.
WT1 was identified as a tumor suppressor and was predicted to be a transcriptional repressor on the basis of its tumor FIG. 7. DDS mutants failed to act in a dominant negative manner. Cells were transfected with Ϫ272SRYP (0.2 g) and different point mutants of WT1 in modified pcDNA3 (0.5 g) with or without wild-type pcDNA3WT1Ϫ/Ϫ (0.1 g). Empty pcDNA3 was added to keep the plasmid amounts equal in each well. The assays were performed 40 h after transfection. Reporter gene activity was normalized to ␤-galactosidase activity. All results are expressed as mean Ϯ S.D. of at least three experiments. suppressor action. However, some of its biological roles cannot be explained by its transcriptional repressor activity. Later it was found that WT1 has at least two independent transactivation domains in its N-terminal region. Transcriptional activation of the SRY promoter is not an unusual case. Recently, many genes have been shown to be up-regulated by WT1 (11,12,14). Amphiregulin has been identified as an in vivo target of WT1 by DNA microarray screening (12). In the same study, they found other genes were also up-regulated to different extents by WT1; however, no genes were down-regulated by WT1. It seems that transcriptional activation of the down stream genes is a major function of WT1 in vivo.
The WTE described in this study consists of high affinity WT1-binding sequences located near the transcription initiation site of the SRY promoter. WT1 can bind to DNA sequences showing variations in the EGR-1 consensus-binding site, GXGXGGGXG (13). This WTE (GAGGGGGTG) is similar to the consensus EGR-1-binding site. However, WTE displays higher affinity than the EGR-1 consensus site (Fig. 4). There is another WT1-binding site far upstream from the WTE in the SRY promoter. Mutational analysis revealed that the second binding site does not have a significant role in the transactivation of the SRY promoter. Full transcriptional activation of the Mis promoter requires at least two factors (14,32). It is likely that additional factors may also be important for the regulation of the SRY promoter. Several TCF-binding sequences are present in the SRY promoter. The contribution of these sites to the activation of the SRY promoter is being investigated. The WT1binding sequence in the SRY promoter is conserved among closely related primates (data not shown). There are virtually no similarities in the 5Ј-untranslated and promoter regions between the mouse and human genes. The WT1-binding site in the human SRY promoter is not conserved in the mouse, but there are similar sequences in the mouse Sry promoter. It will be interesting to see whether Wt1 also regulates mouse Sry during sex determination in the mouse. SRY has been known for its pivotal role in sex determination for a decade. Transcriptional regulation of SRY is relatively uncharacterized. Partial sex reversal has been found resulting from mutations in the 5Ј and 3Ј region of the SRY gene (36). It indicates that regulation of the SRY gene is important in male sex determination. De-regulated expression of the SRY may have pleiotropic effects in the critical period of sex determination. In addition to SRY and WT1, other genes also play an important role in male sex development (19). SRY initiates the differentiation of the indifferent gonads to the male pathway and creates an environment for other critical factors to carry on this process. Thus the regulation of SRY by WT1 is an important step in this pathway. Since DDS mutant proteins were unable to activate the SRY promoter, it is likely that SRY expression is affected in DDS patients.
DDS usually results from point mutations in the WT1 zinc finger region that abrogate DNA binding. The resulting mutant proteins are predicted to act in a dominant negative manner to inhibit the action of the wild-type protein by sequestering it from the transcriptional machinery (6). However, the validity of this hypothesis is in question for several reasons. If DDS mutants act in a dominant negative fashion, then that would inactivate the protein product of the wild-type allele, creating a WT1-null situation during embryogenesis. Wt1-nulls are embryonic lethals in mice, suggesting that WT1-null situations in human embryos would likewise not be viable. However, heterozygous WT1 lethality does not occur and has not been seen in DDS patients. Moreover, there is a mouse model of DDS (37). Clear elucidation of the molecular mechanism of DDS was compromised by many complexities; however, it was clear from the mouse model of DDS that simple squelching of the wildtype protein by the mutant did not occur. In our study we did not see dominant negative effects of DDS mutations in the SRY promoter. Our observations are in good agreement with others (14,34). Discrepancies with others are most likely due to the amount of the DNA used in the transfection experiments; we used nanogram quantities, whereas they used microgram quantities. Thus, the dominant negative effects were probably an artifact of the in vitro system and varied under different experimental conditions. Our model for DDS is that heterozygous point mutations cause functional loss of one allele of WT1 and thus haploinsufficiency. This reduces the amount of transcriptionally active ϪKTS isoform of WT1. As a result, expression of downstream target genes such as SRY and MIS is affected during the critical period of sex determination and differentiation in male gonads.