Chaperoning of mutant p53 protein by wild-type p53 protein causes hypoxic tumor regression

Abstract

Mutant (Mt) p53 abrogates tumor suppression functions of wild-type (WT) p53 through mutant-specific, gain-of-function effects, and patients bearing Mt p53 are chemoresistant. The dominant negative effect of p53 mutants results from their aggregation propensity which causes co-aggregation of WT p53. We explored the mechanism of p53 inactivation in hypoxia and hypothesized whether WT p53 could rescue Mt p53 in hypoxic tumors. WT p53 exists in mutant conformation in hypoxic core of MCF-7 solid tumors, and its conformation is oxygen-dependent. Under simulated hypoxia in cells, WT p53 undergoes conformational change in acquiring mutant conformation. An in vivo chaperone assay shows that WT p53 functions as a molecular chaperone in rescuing conformational and structural p53 mutants in cancer cells both at the transcription and proteome levels. WT p53 chaperone therapy is further shown to cause significant regression of tumor xenografts through reconversion of the mutant phenotype to wild-type p53. The chaperone function of WT p53 is directly linked to the induction of apoptosis in both cancer cells and tumor xenografts. As oncogenic p53 mutants are linked to chemoresistance in hypoxic tumors, p53 chaperone therapy will introduce new dimensions to existing cancer therapeutics. We propose that in cancer cells, WT p53 chaperoning may either exist as a cellular event to potentially reverse the dominant negative effect of its oncogenic mutants or to stabilize yet unidentified factors.

FIGURE 1.

p53 conformation is dependent upon oxygen concentration. a, oxygen concentration in the core and peripheral regions of MCF-7 tumors (14 day, 2 cm³) was analyzed using EPR oxymetry. Core tissue (CT) and peripheral tissue (PT) have 1.8% and 14.3% O₂, respectively (n = 5). b, left, in MCF-7 tumors CT and PT were excised; CT was separated into whole cell (WC), nuclear fraction (NF) and cytoplasmic fraction (CF) to study p53 level. IP showed high p53 level in NF of CT (second lane). p53 conformation was analyzed with p53 conformational antibodies pAb1620 (WT) and pAb240 (Mt); in PT, p53 was in 1620 (fifth lane), and in CT it was exclusively in 240 conformation (seventh lane). b, right, hypoxic (1.8% O₂) MCF-7 cells were divided into WC, NF, and CF; p53 was in high level in NF (second lane) and in 240 conformation (ninth lane). c, in vivo ELISA of hypoxic p53-CFP H1299 cells shows that p53 undergoes conformational change from 1620 to 240 conformation (0–72 h). 1620 conformation was stabilized in early hypoxia (6–12 h) and then decreased with increasing hypoxia exposure; total p53 (black), 1620 (green), and 240 (red) (n = 11, S.D. (error bars), ANOVA). d, model depicts hypothesis that CFP-p53 may undergo conformational change in hypoxia. e, florescence intensity of p53-CFP H1299 cells was analyzed under normoxia (black) and hypoxia (red) using flow cytometry; stable p53-CFP fluorescence is shown in normoxic cells, with a rise in early hypoxia followed by a decrease in fluorescence (48–72 h). f, live cell imaging of p53-CFP H1299 cells shows that fluorescence increased between 6 and 12 h (compare c and d) and then gradually decreased over 72 h (refer to MFI plot) (n = 15, S.D., ANOVA). Reoxygenation induced a sharp increase in p53-CFP florescence (last image) (n = 15), suggesting O₂-dependent conformational change in p53.

FIGURE 2.

WT p53 functions as a molecular chaperone. a, model explains the design for in vivo chaperone assay, which shows various GAL4-Ch and p53-DBS-luciferase constructs (right) b, in the luciferase assay, GAL4BD-p21 5′-p53-DBS was co-transfected with WT p53 cDNA and GAL4-(NTD^1–125), p53, HSP90, HSP70, CHIP, p23, PIN1, BP1 constructs in H1299 cells and luciferase activity was measured, in H1299 cells. The luciferase activity induced by GAL4BD-p21 5′-p53-DBS and p53 cDNA (right) (second lane) was taken as base line. In control, HSP90, HSP70, CHIP, p23, and PIN1 were able to chaperone p53 bound to the p21 5′ site and increase the luciferase activity (p53 inhibitor BP1; negative control). GAL4BD-p21 5′-p53-DBS was then co-transfected with p53-GAL4 or NTD^1–125-GAL4 (free NTD) or with GAL4-p53 or GAL4-NTD^1–125 (NTD constrained); results show that NTD^1–125-GAL4 (second lane) and p53-GAL4 (fifth lane) chaperoned p53 anchored upon p21 5′-p53-DBS, whereas GAL4-NTD^1–125 (fourth lane) and GAL4-p53 (sixth lane) failed (n = 10, S.D. (error bars), ANOVA). Chaperone activity of p53-GAL4 was higher than HSP90. c, ChIP was performed on the GAL4-p21 5′-p53-DBS luciferase vector using pAb421 (p53-Cter), pAb1620 (WT), and pAb240 (Mt) in H1299 (+WT-p53 cDNA), hypoxic MCF-7, and DU-145 cells. ChIP shows that both mutant and wild-type p53 were present in equal ratio on p21 5′-p53-DBS in H1299 cells (+wt-p53 cDNA) (top row, lanes 4 and 5); in hypoxic MCF-7 and normoxic DU-145 cells (Input, GAL4 Ab ChIP; positive controls) p53 exclusively present in mutant (240) conformation (lane 5). Addition of HSP90-GAL4, CHIP-GAL4, NTD^1–125-GAL4, and p53-GAL4 increased p53-1620 conformation considerably upon promoter (lanes 8, 11, 14, and 17). d, results were repeated using real-time ChIP (n = 8, S.D., ANOVA).

FIGURE 3.

WT p53 rescues both conformational and structural Mt p53. a, in vivo ELISA of hypoxic p53-CFP H1299 cells was conducted to observe p53 conformation shift upon addition of p53, NTD^1–125, and CHIP; pAb421 (p53-C-ter) (black), pAb1620 (WT) (green) and pAb240 (Mt) (red). Cisplatin failed to induce change in p53 conformation. b–d, p53, NTD^1–125, and CHIP increased the 1620 conformation by 2–2.50 (n = 11, S.D. (error bars), ANOVA). e, IP with pAb1620 (a) and pAb240 (b) of MCF-7 cells shows that 1620 level is maintained. In hypoxic MCF7 cells (0–72 h), 1620 level declined and was absent at 72 h; only 240 form was present. Addition of CHIP, NTD^1–125, and p53 increases 1620 conformation (black arrow) under hypoxia (Mdm2 negative control) (n = 8). f, IP of normoxic A-431 (p53^R273H, mutant cancer cell line), DU-145 (heterozygous, p53WT/Mt (p53^P223L/V274F), and WRO (homozygous, Mt/Mt) cells with pAb1620 and pAb240 show absence of 1620. Addition of CHIP, NTD^1–125, and WT p53 cDNA (5 μg of DNA) induced conversion of 240 form to 1620 form conformation both in WRO and DU-145 cells (n = 8).

FIGURE 4.

WT-p53 rescues Mt-p53 and causes regression of hypoxic tumor. a, apoptosis was measured in MCF-7 cells in hypoxic condition in presence of cisplatin, UV, CHIP, NTD^1–125, and p53. Cisplatin and UV did not induce apoptosis in hypoxic MCF-7 cells (sixth lane); p53 chaperoning by CHIP, NTD^1–125, and WT p53 caused induction of apoptosis >2.5-fold compared with hypoxia (n = 10) (S.D. (error bars), ANOVA). b, MCF-7 tumors (2 cm³) were treated with molecular chaperones (CHIP, NTD^1–125, and WT p53) (n = 7, each group). Core tissue of control and treated mice were excised to study p53 conformational status. In pretreatment tumors, p53 exclusively exists in 240 conformation. Cisplatin has no effect upon p53 conformation. WT p53, NTD^1–125, and CHIP induced a significant increase in 1620 conformation. c, in vivo ELISA was conducted to validate IP result. d and e, MCF-7 tumors were treated with cisplatin and molecular chaperones (CHIP, NTD^1–125, and WT p53), and MRI was conducted to study tumor volume in pre- and post-treatment (48 h) groups. Cisplatin did not affect the tumor volume whereas transfection of p53 resulted in regression of tumor (64%); CHIP and NTD^1–125 showed regression by 37 and 29%, respectively.