The zinc finger domain of Wilms' tumor 1 suppressor gene (WT1) behaves as a dominant negative, leading to abrogation of WT1 oncogenic potential in breast cancer cells

Abstract

Introduction: There is growing evidence that the Wilms' tumor 1 suppressor gene (WT1) behaves as an oncogene in some forms of breast cancer. Previous studies have demonstrated that the N-terminal domain of WT1 can act as a dominant negative through self-association. In the studies presented here we have explored the potential for the zinc finger domain (ZF) of WT1 to also have dominant-negative effects, and thus further our understanding of this protein.

Methods: Using full-length and ZF-only forms of WT1 we assessed their effect on the WT1 and c-myc promoter using luciferase and chromatin immunoprecipitation assays. The gene expression levels were determined by quantitative real-time RT-PCR, northern blot and western blot. We also assessed the effect of the ZF-only form on the growth of breast cancer cell lines in culture.

Results: Transfection with WT1-ZF plasmids resulted in a stronger inhibition of WT1 promoter than full-length WT1 in breast cancer cells. The WT1-ZF form lacking the lysine-threonine-serine (KTS) insert (ZF - KTS) can bind to the majority of WT1 consensus sites throughout the WT1 promoter region, while the ZF containing the insert (ZF + KTS) form only binds to sites in the proximal promoter. The abundances of endogenous WT1 mRNA and protein were markedly decreased following the stable expression of ZF - KTS in breast cancer cells. The expressions of WT1 target genes, including c-myc, Bcl-2, amphiregulin and TERT, were similarly suppressed by ZF - KTS. Moreover, WT1-ZF - KTS abrogated the transcriptional activation of c-myc mediated by all four predominant isoforms of WT1 (including or lacking alternatively spliced exons 5 and 9). Finally, WT1-ZF - KTS inhibited colony formation and cell division, but induced apoptosis in MCF-7 cells.

Conclusion: Our observations strongly argue that the WT1-ZF plasmid behaves as a dominant-negative regulator of the endogenous WT1 in breast cancer cells. The inhibition on proliferation of breast cancer cells by WT1-ZF - KTS provides a potential candidate of gene therapy for breast cancer.

Figure 1

Inhibition of transcriptional activity of the WT1 promoter by its zinc finger domain. (a) Schematic drawing for luciferase reporter constructs. The predominant transcription start site (+1) is indicated by an arrow. Small solid triangles indicate Wilms' tumor 1 suppressor gene (WT1) binding sites. (b) MCF-7 cells and (c) MDA468 cells were co-transfected with plasmids expressing the WT1 proteins (either pcDNA₃ control (CON), aminoterminal-only construct N-WT1, zinc finger domain lacking or with lysine–threonine–serine (ZF - KTS or ZF + KTS) or full-length WT1 vectors A~D), and either the pGL2 vector control or the differential WT1 promoter-driven luciferase constructs (horizontal axis), respectively. Luciferase activity was normalized with β-galactosidase activity and is expressed in relative luciferase activity as compared with the luciferase vector control (vertical axis). Results are the average of three experiments.

Figure 2

WT1–ZF protein specific binding to WT1 consensus motifs in WT1 promoter in vivo/in vitro. (a) Chromatin immunoprecipitation/PCR analysis for the cross-linked sheared chromatin from control MCF-7 cells (CON) and MCF-7 cells expressing HA-tagged zinc finger domain (HA/ZF-KTS, HA/ZF+KTS) of Wilms' tumor 1 suppressor gene zinc (indicated on right). Upper panel, schematic representation of PCR-amplified fragments P1 and P2 in endogenous human WT1 promoter. Input, mock control and immunoprecipitation with preimmune serum (PI), anti-WT1, anti-HA or anti-acetyl-histone H4 antibody are indicated. Left lane, 1 kb DNA ladder as marker. (b) Same chromatin immunoprecipitation/PCR analysis as (a) with aldehyde dehydrogenase 1 family member A2 (ALDH) promoter primers. (c) Coomassie blue-stained gel. The bacteria-expressed GST and GST/WT1–ZF fusion proteins (shown as GST, G/ZF-KTS and G/ZF+KTS) were purified, fractionated in a 10% SDS-PAGE and visualized by Coomassie blue stain. The molecular sizes of standard protein markers (M) are shown. (d) Competition electrophoretic mobility shift assays were performed using either GST only or GST/WT1–ZF fusion proteins (G/ZF-KTS, G/ZF+KTS) with the ³²P-labeled oligonucleotide of human WT1 proximal promoter in the absence (-) or presence of 20-fold or 100-fold molar excess of WT1 binding site wild-type or site mutant cold oligonucleotides (WT, Mut) as indicated. Migration of DNA-binding complexes induced by GST/ZF-KTS protein and free probe is shown by an arrow. The lane GST displayed no DNA-binding associated with purified GST protein. KTS, lysine–threonine–serine.

Figure 3

Decrease of endogenous WT1 abundance by overexpression of zinc finger domain in breast cancer cells. (a) Northern blot: total RNAs were isolated from the stable transfectants of MCF-7 cells with either pcDNA3 control (CON) or the zinc finger domain lacking the lysine–threonine–serine insert (ZF - KTS) of Wilms' tumor 1 suppressor gene (WT1) expression plasmids as indicated, and were analyzed with WT1 and β-actin probes. Overexpressed WT1–ZF, endogenous WT1 (FL WT1) or β-actin RNA is indicated. (b) Western immunoblot: whole cell extracts were prepared from the stable transfectants of MCF-7 cells or MDA468 cells with either pcDNA₃ control or ZF - KTS expression plasmids, and were analyzed by western blot with either anti-WT1 and β-actin antibodies. Overexpressed WT1–ZF, endogenous WT1 (FL WT1) and β-actin proteins are indicated.

Figure 4

Alteration of expression of WT1 target genes in zinc finger domain lacking KTS-transfected MCF-7 cells. (a) Transient and (b) stable transfectants of MCF-7 cells were analyzed for the expression of Wilms' tumor 1 suppressor gene (WT1) target genes or nontarget genes (horizontal axis) with quantitative real-time PCR assays. The relative expression (vertical axis) of these genes in the transfected MCF-7 cells was calculated as compared with the pcDNA₃ control. Results are the average of three experiments. AREG, amphregulin; BASP1, brain acid soluble protein 1; KTS, lysine–threonine–serine; WT1, Wilms' tumor 1 suppressor gene; ZF, zinc finger domain.

Figure 5

Abrogation of WT1-mediated c-myc transactivation by the zinc finger domain lacking KTS in MCF-7 cells. The MCF-7 cells were transiently co-transfected with the reporter constructs (either pGL2 vector, XNM/Luc or Xmut/Luc), with expression plasmids of Wilms' tumor 1 suppressor gene (WT1) isoforms (either pcDNA₃ vector, A, B, C or D) and with WT1–zinc finger domain (ZF) (either pcDNA₃ control (CON) or ZF lacking the lysine–threonine–serine insert (ZF-KTS)) as well as β-galactosidase expression plasmid, respectively. Relative luciferase activity (vertical axis) is indicated. Histograms show average of three independent experiments.

Figure 6

Retarded cell division in the zinc finger domain lacking lysine–threonine–serine-transfected MCF-7 cells. A typical result of three individual assays for the carboxy-fluorescein diacetate succinimidyl ester (CFSE) labeling profiles of gated transfectants of MCF-7 cells. The transfectants (CON and zinc finger domain lacking the lysine–threonine–serine insert (ZF - KTS)) and the days post G418 selection (day 1, day 4, day 7 and day 10) are indicated.

Figure 7

Accelerated apoptosis in the zinc finger domain lacking lysine–threonine–serine-transfected MCF-7 cells. Dot plots of a representative experiment of flow cytometric analysis for staining of Annexin V–PE (horizontal axis) and 7-amino-actinomycin D (7-AAD) (vertical axis) in nontransfected MCF-7 cells (CON) and transfected MCF-7 cells (Vec, zinc finger domain lacking the lysine–threonine–serine insert (ZF - KTS) or full-length Wilms' tumor 1 suppressor gene (WT1)). Percentages of cell populations in the different quadrants are indicated. Days after G418 selection are shown.