Interaction between transactivation domain of p53 and middle part of TBP-like protein (TLP) is involved in TLP-stimulated and p53-activated transcription from the p21 upstream promoter

Abstract

TBP-like protein (TLP) is involved in transcriptional activation of an upstream promoter of the human p21 gene. TLP binds to p53 and facilitates p53-activated transcription from the upstream promoter. In this study, we clarified that in vitro affinity between TLP and p53 is about one-third of that between TBP and p53. Extensive mutation analyses revealed that the TLP-stimulated function resides in transcription activating domain 1 (TAD1) in the N-terminus of p53. Among the mutants, #22.23, which has two amino acid substitutions in TAD1, exhibited a typical mutant phenotype. Moreover, #22.23 exhibited the strongest mutant phenotype for TLP-binding ability. It is thus thought that TLP-stimulated and p53-dependent transcriptional activation is involved in TAD1 binding of TLP. #22.23 had a decreased transcriptional activation function, especially for the upstream promoter of the endogenous p21 gene, compared with wild-type p53. This mutant did not facilitate p53-dependent growth repression and etoposide-mediated cell-death as wild-type p53 does. Moreover, mutation analysis revealed that middle part of TLP, which is requited for p53 binding, is involved in TLP-stimulated and p53-dependent promoter activation and cell growth repression. These results suggest that activation of the p21 upstream promoter is mediated by interaction between specific regions of TLP and p53.

Figure 1. TLP binds to p53 in solution.

(A) Detection of p53-biding ability of TLP. TLP and TBP were examined for p53 binding by a GST pull-down assay, and affinities of both proteins against p53 were roughly determined by a competitive pull-down assay. (a) FH-p53 was challenged to GTS-tagged TBP (lane 1) or TLP (lane 3) as indicated and a simple pull-down assay was performed. TBP/TLP and FH-p53 were detected by α-GST antibody and α-53 antibody, respectively. No signal was detected when only GST tag was used (data not shown). FH-TLP (lane 2) and FH-TBP (lane 4) were co-applied to the GST-fused protein-adsorbed beads together with FH-p53, respectively, as competitors for GST proteins. (b) Relative band intensities of lane 2 (TBP+TLP), lane 3 (TLP) and lane 4 (TLP+TBP) to that of lane 1 (TBP) of panel (a) are displayed. (B) Comparison of p53-binding affinities of TLP and TBP. GST pull-down assays of lane 1 and lane 2 of panel A-a were performed with increasing amounts of GST-TBP (a) and GST-TLP (b), respectively. input: input protein corresponding to experimental (pull-down) materials. (c) Relative band intensity for p53 protein of panel (a). Results of 0.05 and 0.1 pmole of GST proteins of panel (a) are shown again in the magnified graph.

Figure 2. Function of p53 mutants in TLP-stimulated transcriptional activation.

(A) Schematic representation of the structure of human p53. (a) Positions of TAD (transactivation domain), DBD (DNA-binding domain) and TD (tetramerization domain) are indicated with AA positions. Positions of mutation in the examined mutants are shown by vertical triangles. (b) AA residues of TAD1 in the TAD region. 22L and 23W have been reported to be critical for transactivation (TA), and 18T and 20S are phosphorylated (PH) amino acids (6). (B) Analysis of TLP-stimulated function for individual p53 mutants by an overexpression experiment. Cells were co-transfected with p21 upstream promoter-carrying reporter plasmid and expression plasmid for p53/mutant alone or p53/mutant+TLP. Results are shown as relative luciferase activities (RLA). Ratio represents RLA of p53/mutant expression to RLA of p53/mutant+TLP expression. Some data were examined by statistical analysis. Since the control experiment (ctr) was performed with a vacant effector plasmid, ratios could not be obtained because measured faint luciferase activities are meaningless. (C) Analysis of TLP-stimulated function of representative p53 mutants by a knockdown experiment. TLP siRNA and scrambled (control) siRNA were used as depicted in the figure, and promoter activity was determined as described in panel B.

Figure 3. TLP-binding ability of p53 mutants.

(A) In vitro binding of various p53 mutants. A GST pull-down assay was performed as described in the legend of Fig. 1 by using several representative p53 mutants. (B) Binding between TLP and p53 or its mutants in cells was examined by a mammalian two-hybrid assay. Binding was monitored by luciferase reporter assay. Plasmids for TLP-containing bait (BIND) and p53/mutant-containing prey (ACT) were introduced into cells as indicated. Since TLP is a transcriptional activator with poor DNA-binding capacity, experiments with bait alone brought significant luciferase activity. (C) Immunoprecipitation to detect in vivo binding of TLP and p53. FH-TLP and HA-tagged p53 or its mutant (#22.23) were overexpressed in cells and immunoprecipitataied with M2 beads. Immunoprecipitates were analyzed for TLP-associating p53, TLP and GAPDH.

Figure 4. Effect of #22.23 mutation on gene expression from endogenous p21 promoters.

(A) Two kinds of major p21 transcripts produced from the human p21 gene. Position of exons of p21 alt-a and p21 variant-1 transcripts and genomic DNA around the two p21 promoters are schematically illustrated. Open and solid boxes represent non-coding and coding exons, respectively. Two primer sets indicated by thick arrows were used for RT-PCR to detect variant-1 and alt-a, respectively. (B) p53^−/− cells were transfected with expression vectors for wild-type and mutant (#22.23) p53, and two species of p21 transcripts were determined by RT-PCR. Vector: vacant vector. RNAs of endogenous β-actin, p53 and TLP were also analyzed. (C) Assays for TLP-stimulated function of wild-type p53 and #22.23. (a) Experiments were performed as described in panel B. Cells were transfected with a TLP expression plasmid in addition to a p53 expression plasmid as indicated. ctr and vec: corresponding vacant vectors. (b) Amounts of intracellular p53 and #22.23 proteins were also detected by immunoblotting in addition to GAPDH and endogenous and exogenous TLPs. (c) Degree of increase in alt-a transcripts stimulated by exogenous TLP in p53-expressing cells. Ratios of band intensities of alt-a of panel (a) in vacant vector-introduced cells to that in TLP overexpressed cells were calculated for three kinds of cells.

Figure 5. Effect of #22.23 mutation on cell growth and etoposide-induced cell death.

(A) Five-hundred thousand p53^−/− cells in a dish were cultured for 24 hr. Cells were transfected with an expression plasmid for p53 (WT) or #22.23 (mut) together with a TLP expression plasmid. After 24 hr, 8×104 cells were replated and maintained. Cell numbers were counted every 24 hr (panels a–c). ctr: vacant plasmid. (d) Cell numbers at each time shown in panels a–c are displayed as ratios to the initial cell number. (B) Experiments were performed as described above, but replated cells were maintained in a medium containing 30 µM etoposide to examine the effect of TLP on apoptotic cell death (a–c). Numbers of remaining viable cells were counted. (d) Data are summarized as described above.

Figure 6. Examination of mutant TLPs on transcriptional activation and p53 binding.

(A) Structural relationship between TBP and TLP. Amino acid numbers are indicated from N-termini. TLP covers the evolutionally conserved region of TBP. A putative p53-binding region in TBP deduced from deletion analyses and its TLP counterpart (from 63 to 115) are depicted as a gray area. Positions of AAs of the TLP mutants used in this study (R86S, F100E, and F114E) are indicated with vertical arrowheads. (B) Transcription activation function of wild-type (WT) and mutant TLPs were assayed in native (a) and p53^−/− (b) cells. (C) Binding of TLP and p53. Wild-type and F100E TLPs were analyzed for the p53-bidnding ability by two-hybrid assay.

Figure 7. Effect of F100E mutation of TLP on the expression of endogenous p21 gene and cell growth.

(A) Wild-type (a) and p53^−/− cells (b) were transfected with expression vectors of wild-type and mutant (F100E) TLPs, and two species of p21 transcripts were determined by RT-PCR as described in a legend of Fig. 4. (B) Wild-type and mutant TLP-transfected native (a) and p53^−/− (b) cells were cultured for 24 hr. Cells (1×10⁵) were replated and cell numbers were counted every 24 hr. ctr: vacant plasmid.