The Kidney-expressed Winged Helix Transcription Factor FREAC-4 Is Regulated by Ets-1

In this paper we show that the kidney-expressed winged helix transcription factor FREAC-4 is regulated by Ets-1, another kidney-expressed transcription factor. Through transfection experiments three Ets-1 cis-elements are identified within the first 152 nucleotides upstream of the transcription start in thefreac-4 promoter. These sites are confirmed in a DNase Iin vitro protection assay using recombinant Ets-1 protein. In cotransfection experiments using an Ets-1 expression vector, the induction of freac-4 reporter gene activity is attenuated approximately 6-fold when the three Ets-1 binding sites are mutated. Furthermore, we demonstrate that overexpression of Ets-1 in the human embryonic kidney cell line 293 is sufficient to increasefreac-4 mRNA levels. These results are compatible with the hypothesis that Ets-1 acts as an upstream regulator of FREAC-4 expression during kidney development.

The forkhead family of transcription factors belongs to the "winged helix" superfamily of DNA binding proteins (1, 2), a name derived from the x-ray crystallography data on HNF-3␥ bound to DNA (3). Forkhead proteins bind to DNA through a highly conserved region of some 105 amino acids, the forkhead motif (4). This domain was first identified in a Drosophila gene, forkhead (fkh), named after a homeotic mutation in which proper formation of terminal structures, most striking gut formation, is affected (5). The forkhead motif has since been identified in over 40 genes isolated from a vast range of organisms, among them man (6). Less is known about the biological function of the various forkhead genes. However, important roles in tumorigenesis (7)(8)(9)(10)(11), embryonic development (12)(13)(14)(15)(16), and regulation of tissue-specific gene expression (17)(18)(19)(20) have been demonstrated for members of this gene family. Highly complex functions, such as specification of visual projection maps in the retina, have been shown to depend on correct spatial expression of the forkhead genes CBF-1 and CBF-2 (21). It has also been shown that the Drosophila gene fkh directly participates in hormonal regulation of gene expression (22). The ecdysone responsive unit, controlling the stage-specific responses of sgs-4 (salivary gland secretion protein gene) to 20-hydroxyecdysone, interacts with the gene product of fkh in vivo (22). Recently, forkhead proteins have been implicated as nuclear targets for transforming growth factor-␤ and insulinlike signaling (23)(24)(25).
We have earlier reported that freac-4 (26) and freac-9 (27) are two human forkhead genes that are predominately expressed in the kidney. The mouse homologue of freac-4, BF-2, has been shown to be essential for stromal mesenchyme differentiation during kidney morphogenesis (28). In this paper we demonstrate that the kidney expressed transcription factor Ets-1 is essential for regulation of freac-4. We also discuss a possible role for Ets-1 as an upstream regulator of freac-4 during kidney development.

MATERIALS AND METHODS
Cell Culture, Transfections, and Reporter Gene Assays-Cells were obtained through the American Tissue Culture Collection: COS-7 (monkey transformed kidney; CRL-165) and 293 (human embryonic kidney; CRL-1573). We have previously demonstrated that these cell lines express freac-4 transcripts (26). All cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum, 100 IU/ml penicillin, and 100 g/ml streptomycin (Life Technologies, Inc.). For transfections, different FREAC-4 luciferase constructs were used as described previously (26). A 1.6-kilobase pair HindIII restriction fragment of the human Ets-1 cDNA was subcloned into pCB6ϩ to be used as an expression construct for Ets-1 protein in transfection assays. A typical transient transfection contained 100 ng of luciferase reporter plasmid and 160 -200 ng of cotransfected expression plasmid. These plasmids were diluted into 560 l of OptiMEM together with 2 g of LipofectAMINE (Life Technologies, Inc.) and added to cells cultured in a gelatin-coated 16-mm tissue culture well. Cell harvest and luciferase assay were performed according to Promega Corp. (Technical Bullentin 101). To compensate for differences in transfection efficiency, 10 ng of a ␤-galactosidase-expressing plasmid, pCMV␤gal (CLONTECH), was added to each transfection. ␤-Galactosidase activity was measured using a Lumi␤-galactosidase assay (CLONTECH). A typical stable transfection contained 20 g of linearized Ets-1 expression plasmid, diluted into 8 ml of OptiMEM together with 50 g of Lipofectin (Life Technologies, Inc.) and added to cells cultured in a gelatin-coated 82-mm tissue culture dish. Medium was changed after ϳ20 h, and after another 24 h 800 g of G418 sulfate/ml (LifeTechnologies, Inc.) was added to the medium. After 10 -15 days resistant foci appeared.
In Vitro Mutagenesis-A genomic DNA fragment spanning from nucleotides 2122 to 2485 was subcloned into pBluescript SK Ϫ (Stratagene). The three potential Ets-1 binding sites in this region were mutated with QuikChange™ Site-directed Mutagenesis Kit (Stratagene) using the following primers: site number 1, CGAGAAGGGCTG-ATTTAATAGGCTTGCTTTCC and GGAAAGCAAGCCTATTAAATCA-GCCCTTCTCG; site number 2, CCTAGGCTTGCTTTAATTCCCTCGG-CAGCG and CGCTGCCGAGGGAATTAAAGCAAGCCTAGG; and site number 3, GCTATAAGCCGATTAAGGTCCGCCCTCTCC and GGAG-AGGGCGGACCTTAATCGGCTTATAGC. Mutagenesis was verified by sequencing. The mutant promoters were then cloned into pGL2-Basic.
Protein Expression and Purification-An Ets-1 expression vector, ⌬N331 (30), was used to express the protein under the control of a T7 promoter in Escherichia coli BL21(DE3);pLysS cells as has earlier been described (29). In brief, T7 polymerase expression was induced with 1 mM isopropyl-␤-D-thiogalactopyranoside for 1 h at 37°C, cells were lysed by sonication in a buffer containing 50 mM Tris, pH 7.9, 1 mM EDTA, 1 M KCl, 1 mM dithiothreitol, and 1 mM phenylmethylsulfonyl fluoride. The soluble fraction was cleared by centrifugation, dialyzed into 20 mM sodium citrate, pH 5.3, 150 mM KCl, 1 mM dithiothreitol, 1 mM EDTA, and 0.2 mM phenylmethylsulfonyl fluoride and applied to a DEAE-cellulose column. Purity of Ͼ90% was determined by Coomassie Blue staining.
DNase I Footprinting-DNase I footprinting was performed with labeled restriction enzyme fragments corresponding to nucleotides 2251-2426 ( Fig. 1 in Ref. 26). Plasmids were linearized with XhoI and 5Ј-labeled with [␣-32 P]dATP and [␣-32 P]dCTP using the Klenow enzyme. The probes were cut out from the plasmids with SacI and gel purified. 20,000 cpm Cerenkov of probe was added to a binding reaction with a final volume of 50 l in the presence of various amounts of Ets-1 protein extract or bovine serum albumin (25 mM Tris⅐HCl, pH 7.8, 50 mM KCl, 0.5 mM EDTA, 0.5 mM dithiothreitol, 6.25 mM MgCl 2 , 2% polyvinyl alcohol, 10% glycerol, 5 g of poly[d(I⅐C)]/ml) and incubated for 15 min in room temperature. DNase I digestion and work-up procedure followed the method of Jones et al. (30).
RNase Protection Assays-A 257-base pair SacII restriction fragment spanning nucleotides 3719 -3976 ( Fig. 1 in Ref. 26) was used as a specific probe for freac-4. T3 RNA polymerase and [␣-32 P]CTP were used to label a cRNA antisense probe. Labeled antisense probe, approximately 170,000 cpm Cerenkov for each reaction, was added to 50 g of total RNA in a hybridization buffer (80% formamide, 100 mM sodium citrate, pH 6.4, 300 mM sodium acetate, pH 6.4, 1 mM EDTA) at 50°C overnight. After digestion with RNase A and RNase T1, the protected fragment was electrophoresed on a 6% sequencing gel. A 262-base pair XhoI/SacI restriction fragment of the human cDNA for Ets-1 was subcloned and used to make cRNA antisense probe. The assay was carried out as outlined for freac-4 with the exception that 25 g of total RNA was used. The ␤-actin probe used was derived from the plasmid pTRI-Actin-Mouse (Ambion). RNA samples were treated with DNase I free of RNase activity.

Cotransfections with an Ets-1 Expression Vector Induces an
Increase in freac-4 Reporter Gene Activity-Even though it has been clearly demonstrated that BF-2 (the mouse homologue of the human freac-4 gene) is essential for differentiation of the condensed mesenchyme into tubular epithelium during kidney formation (28), very little is known about upstream and downstream signaling in the freac-4/BF-2 pathway. We became interested in the transcription factor Ets-1 as a possible regulator of such a pathway because Ets-1 has been implicated as a regulator of gene expression in mesodermal cells that are involved in morphogenic processes such as organ formation (31). Ets-1 is widely expressed in the murine and chick embryo during kidney formation, and its expression pattern has been shown to include tubular structures of the mesonephric kidney as well as glomeruli in the developing kidney (32,33). In a first  (Fig. 1). A dose-response pattern is present in the range of 0 -50 ng of Ets-1 expression plasmid, whereas 200 ng of the same plasmid in no significant way changed the activation profile as compared with 50 ng of expression vector.
The Ets-1 Inducibility Is Mapped to a Ϫ152 freac-4 Reporter Gene Construct That Contains Three Potential Ets-1 Binding Sites-Two derivatives of the Ϫ2273 construct (FREAC-4-luc), Ϫ527-luc and Ϫ152-luc extending 527 and 152 nucleotides upstream of the transcription start (see Fig. 3A), were used to study the inducibility conferred by Ets-1 in cotransfection experiments. As can be seen in Fig. 2 FREAC-4-luc, Ϫ527-luc and Ϫ152-luc are all induced approximately 40-fold. Because the luciferase vector void of any freac-4 promoter sequence is not induced by Ets-1 cotransfections, we conclude that the freac-4 promoter sequence in Ϫ152-luc most likely contains binding sites for Ets-1 (Fig. 3B). A nucleotide frequency profile, derived from repetitive cycles of selection and amplification of a double stranded oligonucleotide template using recombinant Ets-1 protein (34), was used to identify potential Ets-1 binding sites in the Ϫ152-luc construct. Three potential Ets-1 binding sites are depicted in Fig. 3. The variation in induction seen between independent cotransfection experiments, e.g. ϳ15-fold ( Fig. 1; see also Fig. 5) to ϳ40-fold (Fig. 2), was found to correlate with the total amount of DNA used in the transfections, i.e. a total amount of 200 ng (experiment in Fig. 2) gave reproducibly a ϳ40-fold induction, whereas 260 ng (see Fig. 5) to 300 ng (Fig.  1) in a similar manner gave a ϳ15-fold induction.
When the Potential Ets-1 Binding Sites in the freac-4 Promoter Are Mutated the Ets-1 Induction Is Attenuated-To confirm the presence of Ets-1 binding sites within the Ϫ152 region   FIG. 3. A, the three reporter constructs used extend to Ϫ2273, Ϫ527, and Ϫ152 respectively, relative to the start of transcription. The Ϫ152 construct contains three potential Ets-1 binding sites located at Ϫ95 to Ϫ104, Ϫ77 to Ϫ86, and Ϫ24 to Ϫ15; nucleotide numbering is as described by Ernstsson et al. (26). TATA motif and location of transcription start as described by Ernstsson et al. (26). B, the three potential Ets-1 binding sites in the freac-4 promoter show a high degree of sequence similarity to a consensus sequence based on sequences known to bind Ets-1 (34). (rEts-1) and bovine serum albumin (BSA). Three regions are protected: Ϫ12 to Ϫ23, Ϫ77 to Ϫ86, and Ϫ93 to Ϫ102. These three regions show a high sequence similarity with sequences known to interact with Ets-1 (see Fig. 3B). For details see "Experimental Procedures." of the freac-4 promoter, we set up a DNase I in vitro protection assay using a probe spanning from Ϫ152 to ϩ23, relative the start of transcription ( Fig. 3B; corresponding to nucleotides 2251-2426 of Fig. 1 in Ref. 26). Purified recombinant Ets-1 protein (rEts-1) was used (see "Materials and Methods"), and three protected regions could be demonstrated Fig. 4. These three regions, Ϫ12 to Ϫ23, Ϫ77 to Ϫ86, and Ϫ93 to Ϫ102, are all centered over the conserved GGAA motif in the three predicted Ets-1 sites as discussed above. To further analyze the role these sites play in Ets-1 induction of reporter gene activity derived from the Ϫ152-luc construct, we introduced mutations in each of these sites. We chose to change the obligatory purine dinucleotide GG of the GGAA core motif for Ets-1 binding sites into the pyrimidine dinucleotide TT (34,35). As can be seen in Fig. 5 each of these mutations drastically reduced the Ets-1mediated induction of reporter gene activity with ϳ60 -70%. When all three mutations were combined, a further decrease in reporter gene activity was noted to less than 20% as compared with that of the wild type construct. Similar levels of reduction in reporter gene activity, after mutation of Ets-1 binding sites, have been reported for the rat prolactin promoter (36) as well as for the NFB1 promoter (37).

FIG. 4. DNase I footprint using recombinant Ets-1 protein
Expression of Ets-1 in 293 Cells, a Human Embryonic Kidney Cell Line Induces freac-4 Expression-In an attempt to find out if an increased Ets-1 expression would be sufficient per se to up-regulate freac-4 mRNA levels, in a kidney-derived cell line, we stably transfected 293 cells with an Ets-1 expression vector also encoding Neo R to facilitate selection of Ets-1 expressing clones. In Fig. 6 we demonstrate that cell clones with an increased Ets-1 mRNA level also have a higher level of freac-4 mRNA. Despite the fact that this experimental approach not can differ between indirect or direct effects on the freac-4 promoter, these results are compatible with the view that Ets-1 acts as a positive upstream regulator of freac-4 expression and that the Ets-1-dependent activation is mediated through Ets-1 sites present in the freac-4 promoter.

DISCUSSION
The Ets-1 proto-oncogene, the founding member of the Etsfamily of transcription factors, was originally described as the cellular homologue of the v-ets oncogene, which is translated as a 135-kDa gag-myb-ets fusion protein from the replicationdeficient retrovirus E26 in chickens (38 -40). Members of this family play important roles in regulating gene expression in response to multiple developmental and mitogenic signals (41)(42)(43). Previous studies in mammals have demonstrated that Ets-1 is a transcription factor important during development of lymphoid cells and their subsequent activation (44,45). Other sites of Ets-1 expression are vascular structures, the central nervous system (including the closed neural tube) as well as the developing kidney (32,33). Freac-4 and its mouse homologue BF-2 have a more restricted expression pattern, as compared with Ets-1, including testis, spinal cord, brain, and the developing kidney (26,28,46,47).
In this paper we have demonstrated the presence of at least three Ets-1 binding sites within the freac-4 promoter (Figs. 3B and 4). When an Ets-1 expression plasmid is cotransfected with various freac-4 promoter reporter constructs a clear induction is noted (ϳ15-40-fold; Figs. 1 and 2). When the three Ets-1 binding sites present on the Ϫ152-luc construct are mutated, the Ets-1 induction is reduced by a factor of ϳ6 as compared with the wild type promoter (Fig. 5). This is consistent with Ets-1 as a major regulator/activator of freac-4 gene expression, other cis-elements and trans-activators are also likely to contribute, because we have previously shown that both p53 and WT-1 (Wilms' tumor suppressor gene-1) are involved in the regulation of freac-4 (26).
In transfection assays, for practical reasons, only limited amounts of promoter sequence can be used. The expression level of transactivators used in cotransfection assays is also a factor to consider when interpreting the results of such experiments. To investigate whether or not an increased level of wild type Ets-1 mRNA expression would be sufficient to up-regulate freac-4 expression, we stably transfected the human embryonic kidney cell line 293 with an Ets-1 expression vector. Total RNA was then used in a RNase protection assay, and we could demonstrate an increased level of freac-4 mRNA in cell clones with an increased level of Ets-1 mRNA (Fig. 6). We would also like to point out that untransfected 293 cells have a low level of Ets-1 expression (a faint band in lane A of Fig. 6). Taken together, these experiments show that it is possible that Ets-1 could act as an upstream regulator of FREAC-4 expression in kidney cells. This notion gains further support from the fact that Ets-1 is expressed during early kidney development in the tubular structures of mesonephros day E10.5 pc (33), whereas the early stages of kidney development in BF-2 Ϫ/Ϫ mice remain unaffected (28). BF-2 is expressed at E12.5 pc in a population of cells that surround the condensation of nephrogenic mesenchyme (28). Thus the temporal appearance of Ets-1 and BF-2 in the developing kidney does not exclude Ets-1 as a possible upstream regulator in the BF-2/Freac-4 pathway. In Fig. 7 we have outlined a hypothetical regulative pathway in which Ets-1, p53, WT-1, and BF2/freac-4 participate. This pathway is based on the experiments described in this paper as well as work by others: (i) p53 is known to repress expression of both Ets-1 (48) and freac-4 (26); (ii) p53 is expressed in the developing kidney during embryogenesis (49); (iii) transgenic mice expressing wild type p53, under control of the mouse mammary tumor virus promoter, undergo progressive renal failure due to defective kidney differentiation with small kidneys with about half of the normal number of nephrons (50); (iv) WT-1 up-regulates freac-4 expression (26); and (v) in WT-1 Ϫ/Ϫ mice metanephrogenic mesenchyme remains uninduced and no kidney is formed (51). Clearly more research is needed to elucidate the pathway through which BF-2/FREAC-4 participates in the regulation of nephrogenesis, and we hope that the results presented in this paper will contribute to a better understanding of this regulative pathway.