p53 transactivation and the impact of mutations, cofactors and small molecules using a simplified yeast-based screening system

Abstract

Background: The p53 tumor suppressor, which is altered in most cancers, is a sequence-specific transcription factor that is able to modulate the expression of many target genes and influence a variety of cellular pathways. Inactivation of the p53 pathway in cancer frequently occurs through the expression of mutant p53 protein. In tumors that retain wild type p53, the pathway can be altered by upstream modulators, particularly the p53 negative regulators MDM2 and MDM4.

Methodology/principal findings: Given the many factors that might influence p53 function, including expression levels, mutations, cofactor proteins and small molecules, we expanded our previously described yeast-based system to provide the opportunity for efficient investigation of their individual and combined impacts in a miniaturized format. The system integrates i) variable expression of p53 proteins under the finely tunable GAL1,10 promoter, ii) single copy, chromosomally located p53-responsive and control luminescence reporters, iii) enhanced chemical uptake using modified ABC-transporters, iv) small-volume formats for treatment and dual-luciferase assays, and v) opportunities to co-express p53 with other cofactor proteins. This robust system can distinguish different levels of expression of WT and mutant p53 as well as interactions with MDM2 or 53BP1.

Conclusions/significance: We found that the small molecules Nutlin and RITA could both relieve the MDM2-dependent inhibition of WT p53 transactivation function, while only RITA could impact p53/53BP1 functional interactions. PRIMA-1 was ineffective in modifying the transactivation capacity of WT p53 and missense p53 mutations. This dual-luciferase assay can, therefore, provide a high-throughput assessment tool for investigating a matrix of factors that can influence the p53 network, including the effectiveness of newly developed small molecules, on WT and tumor-associated p53 mutants as well as interacting proteins.

Figure 1. Generation of a small volume format for p53 functional assays.

(A) Relative transactivation capacity of WT p53 and a carboxy terminal deletion measured in permeabilized cell cultures and normalized to optical density OD. p53 proteins were induced at different levels by varying the amount of galactose, as indicated. Three different p53 response elements (REs) that differed in relative transactivation capacity from very strong (CON, an optimized consensus sequence), to strong (P21, corresponding to the p21-5′ site) and to moderate (GADD45). OD of the cultures was used as normalizing factor. Presented are the average measurements and standard deviations of three biological replicates. (B, C) Small-volume yeast cultures can determine p53 transactivation capacityRelative transactivation capacity of WT and the R282Q p53 have been compared towards four different REs obtained with the traditional assay based on 2ml liquid cultures in individual tubes (B) and with the permeabilized assay format based on 100 µl cultures prepared directly in 96-well plates (C). p53 proteins were induced at different levels by varying the amount of galactose, as indicated. A strong (P21), two moderate (PUMA, GADD45) and a weak RE (AIP1) were compared. Cells collected from the two different culture protocols were used for the measurement of luciferase activity as described in the Materials and Methods section. Presented are the average fold-induction of luciferase by p53 proteins relative to the activity obtained with an empty vector; included is the standard deviations of three replicates. (D) Impact of genetic modifications of the ABC-transporter systems on yeast growth. Overnight liquid cultures in synthetic medium containing glucose were washed and resuspended in fresh medium containing raffinose (2%) as the carbon source and low levels of galactose (0.0032%) (time zero) to induce p53 protein expression. Cultures were diluted to ∼0.1 OD_600nm, as measured by a plate reader. OD was measured at the 6, 12, 24hr time intervals. Error bars plot the standard deviations of three biological replicates. The average absorbances are also presented to the right of the graph.

Figure 2. Either Firefly or Renilla luciferase can function as p53-dependent reporters.

(A) The ability of Firefly and Renilla cDNAs to serve as reporters for p53 transactivation was examined by placing them downstream from the moderate p53 RE derived from the PUMA promoter in isogenic strains. The values indicate the fold induction measured over an empty vector. Presented are average and standard deviations of three replicates relative to optical density of the cultures measured at different times (T in hrs) after switching cultures to galactose-containing medium. (B) Dual luciferase reporter assay with a strain expressing WT p53 and containing the Firefly luciferase as p53 reporter gene and the Renilla luciferase as constitutive reporter. Presented are the average and standard error of the Firefly luciferase activities normalized for Renilla and compared to empty vector at various time points after shifting 100 µl yeast cultures to galactose-containing media in the 96-well plate format. (C) Comparison of relative induction using measurement of protein from 2 ml cultures vs direct permeabilization of cells in a 384 well format following transfer from a 96-well growth plate, as described in the text and the Materials and Methods section. Relative transactivation capacities of WT p53 and the R282Q mutant in the “2 ml vial”experimental set-ups were measured using either protein extraction or permeabilization. Experiments were conducted using 0.032% galactose inducer, unless specified otherwise. Error bars plot the standard error of four biological replicates.

Figure 3. MDM2 co-expression reduces WT and mutant p53-dependent transactivation and can impact p53 protein level and stability.

(A) The functional interaction between p53 and MDM2 was examined using two different reporter strains, as indicated. Transformants were cultured in 0.012% galactose to achieve low expression of p53 for 16 hours. MDM2 is expressed under the constitutive PGK1 promoter. Besides WT p53, several mutants at Ser/Thr in the p53 N-ter were tested. The activity of each p53 mutant was set to one to better focus on the relative impact of MDM2 co-expression on p53 transactivation capacity. The relative transactivation potential of the various p53 mutants is presented in Supporting Information S1; 4D refers to a quadrupole mutant with S15D, T18E, S20D, S33D changes in p53 . 6A indicates a multiple mutant with alanine changes at S15, T18, S20, S33, S37, S46. 6D indicates a multiple mutant with aspartic acid changes at S15, S20, S33, S37, S46 and a glutamic acid change at T18. Presented are the average fold-inductions by p53 proteins compared to empty vector and normalized using the Renilla control luciferase. These assays were conducted with diploid strains that were obtained by crossing the indicated yLFM- p53 reporter strains with the BY4704 strain (see Materials and Methods section) using the permeabilized format. (B) Western blot analyses of p53 and MDM2 protein levels. O/N cultures in synthetic glucose medium (GLU) were washed and shifted to medium containing raffinose and 0.012% galactose (GAL) to achieve low expression of p53. The p53 was expressed under the inducible GAL1 promoter while MDM2 was expressed at constitutive levels from a moderate PGK1 promoter. After 16 hrs of growth in galactose-containing medium, cells were washed and transferred to glucose medium to repress the GAL1 promoter. Samples were collected at the indicated time points to prepare protein extracts for western blot. 100 µg (MDM2 and actin, top panel) and 20 µg (p53 and actin, lower panel) of extract was loaded in each lane. The DO-1, SMP14 and I-19-R antibodies (Santa Cruz) were used for the immunodetection of p53, MDM2 and actin, respectively.Actin levels were used as a normalization factor to estimate relative MDM2 and p53 amounts. Consistent with a previous study , we observed that MDM2 expression under the PGK1 promoter was affected by the culture state and was particularly reduced when cell approached the stationary phase (O/N in glucose; T8 and T12 time points; at T12 cells were diluted for the additional 12 hr time point). The relative changes in MDM2 and p53 protein amounts compared to the level observed in glucose cultures are indicated above the immunoblot. (C) Quantification of p53 expression relative to the amount observed after 16 hrs in 0.012% galactose, normalized to actin levels. A 10% reduction in steady-state p53 protein amount due to the co-expression of MDM2 was observed in the galactose-induced cultures. To better visualize the impact of MDM2 on the estimated p53 half life (EHL) the relative amount of p53 observed after 16 hrs in galactose was set to 100%, both for extracts of cells expressing only p53 or p53 + MDM2.

Figure 4. Functional interactions between partial function p53 mutants and MDM2 or 53BP1.

Mutant p53 expression was under the control of the GAL1 promoter while MDM2 or 53BP1 (a clone containing a N-ter deletion of the first 970 amino acids) were expressed at constitutive levels under the PGK1 and ADH1 promoters, respectively. p53 expression was induced for 16 hrs in medium containing 0.012% galactose. Presented are results describing the impact of MDM2 or 53BP1 on transactivation of various p53 mutants that are capable of partial transactivation toward the PUMA RE. To better visualize the impact of MDM2 and 53BP1, the activity of each p53 mutant alone is set to 100%. The relative light units of the various mutants in this experiment were WT p53, 2.1×10⁵; A119V, 1.3×10⁵; R181L, 0.86×10⁵; P219L, 0.87×10⁵; R282Q, 0.79×10⁵; R283H, 0.53×10⁵. Significant differences in activity relative to p53 alone are shown (*: p<0.01; ∧: p<0.05, Student's t-test).

Figure 5. Functional interactions between wild type p53 and MDM2 or 53BP1 and the impact of Nutlin and RITA.

WT p53 was expressed at low-level achieved by culturing cells in medium containing 0.012% galactose for 16 hrs in the 96-well plate format. MDM2 was expressed from the moderate PGK1 promoter. (A) The impact of MDM2 on p53-dependent transactivation was examined in the presence of different concentrations of Nutlin added to the medium at the time of the switch to galactose-containing medium using a reporter strain containing the moderate PUMA p53 RE. The average transactivation relative to the basal level of reporter activity measured in cells that do not express p53 and standard deviations of three biological repeats are presented. Significant differences in activity relative to p53 alone are shown (*: p<0.01, Student's t-test). (B) Firefly luciferase activities normalized using the control luciferase Renilla are presented for empty vector and wild type p53 in the presence of different amounts of RITA. (C) Nutlin and RITA impact on the functional interactions between p53 and MDM2 or 53BP1. Nutlin (20 µM) or RITA (0.5 µM) were added at the time of switching cultures to galactose-containing medium. The luciferase activity by wild type p53 alone, normalized using the Renilla control luciferase, is set at 100%. Both MDM2 and 53BP1 co-expression reduced p53-dependent transactivation. Nutlin partially relieved the functional impact of MDM2, but not that of 53BP1. RITA partially relieved p53 from the inhibition by both MDM2 and 53BP1. Significant differences are shown (*: p<0.01; ∧: p<0.05, Student's t-test). (D) MDM2 and p53 immunoblot in mock-, RITA- and Nutlin-treated yeast cells. Proteins were prepared from cells grown in medium containing 0.012% galactose for 16 hrs and treated with DMSO solvent control 0.5 µM RITA or 20 µM Nutlin. 25 µg were loaded to detect p53 and 100 µg of protein extracts were loaded to probe for MDM2. Actin was used as a loading control.