Yeast As a Chassis for Developing Functional Assays to Study Human P53

Summary

Presented here are four protocols to construct and exploit yeast Saccharomyces cerevisiae reporter strains to study human P53 transactivation potential, impacts of its various cancer-associated mutations, co-expressed interacting proteins, and the effects of specific small molecules.

Abstract

The finding that the well-known mammalian P53 protein can act as a transcription factor (TF) in the yeast S. cerevisiae has allowed for the development of different functional assays to study the impacts of 1) binding site [i.e., response element (RE)] sequence variants on P53 transactivation specificity or 2) TP53 mutations, co-expressed cofactors, or small molecules on P53 transactivation activity. Different basic and translational research applications have been developed. Experimentally, these approaches exploit two major advantages of the yeast model. On one hand, the ease of genome editing enables quick construction of qualitative or quantitative reporter systems by exploiting isogenic strains that differ only at the level of a specific P53-RE to investigate sequence-specificity of P53-dependent transactivation. On the other hand, the availability of regulated systems for ectopic P53 expression allows the evaluation of transactivation in a wide range of protein expression. Reviewed in this report are extensively used systems that are based on color reporter genes, luciferase, and the growth of yeast to illustrate their main methodological steps and to critically assess their predictive power. Moreover, the extreme versatility of these approaches can be easily exploited to study different TFs including P63 and P73, which are other members of TP53 gene family.

Introduction

Transcription is an extremely complex process involving dynamic, spatial, and temporal organization of transcription factors (TFs) and cofactors for the recruitment and modulation of RNA polymerases on chromatin regions in response to specific stimuli¹. Most TFs, including the human P53 tumor suppressor, recognize specific cis-acting elements in the form of DNA sequences called response elements (REs), which consist of single (or multiple) unique motifs ~6-10 nucleotides long. Within these motifs, individual positions may show various degrees of variability², usually summarized by position weight matrices (PWM) or logos^3,4.

The yeast S. cerevisiae is a suitable model system for studying different aspects of human proteins through complementation assays, ectopic expression, and functional assays, even when an orthologous yeast gene is not present^5,6,7. Due to the evolutionary conservation of basal components of the transcriptional system⁸, many human TFs (when ectopically expressed in yeast cells) can modulate the expression of a reporter gene by acting through promoters engineered to contain appropriate REs. The transcription model system presented here for human P53 is characterized by three major variables whose effects can be modulated: 1) the modality of expression and type of P53, 2) the RE sequence controlling P53-dependent transcription, and 3) the type of reporter gene (Figure 1A).

Concerning the modality of P53 expression, S. cerevisiae allows the choice of inducible, repressible, or constitutive promoters^9,10,11. In particular, the inducible GAL1 promoter allows basal (using raffinose as a carbon source) or variable (by changing the amount of galactose in the media) expression of a TF in yeast. In fact, the finely tunable expression represents a critical development for studying not only P53 itself but also other P53 family proteins^12,13.

Regarding the type of REs controlling the P53-dependent expression, S. cerevisiae allows the construction of different reporter strains possessing unique differences in the RE of interest in an otherwise isogenic background. This goal is reached using an adaptation of a particularly versatile genome editing approach developed in S. cerevisiae, called delitto perfetto^12,14,15,16.

Furthermore, different reporter genes (i.e., URA3, HIS3, and ADE2) can be used to qualitatively and quantitatively evaluate transcriptional activities of human TFs in S. cerevisiae, each with specific features that can be tailored to experimental needs^{17,18,19,20,21}. The expression of these reporter genes confers uracil, histidine, and adenine prototrophy, respectively. The URA3 reporter does not allow the growth of cells in the presence of 5-FOA as well, and thus it can be counterselected. The ADE2 reporter system has the advantage that, besides nutritional selection, it allows the identification of yeast cells that express wild-type (i.e., functional on ADE2 expression) or mutant (i.e.,not functional on ADE2) P53 from the colony color.

For example, yeast cells expressing the ADE2 gene generate normally sized white colonies on plates containing limiting amounts of adenine (2.5-5.0 mg/L), while those that poorly or do not transcribe it appear on the same plate as smaller red (or pink) colonies. This is due to accumulation of an intermediate in the adenine biosynthetic pathway (i.e., P-ribosylamino-imidazole, which has been previously called amino-imidazole ribotide or AIR), which is converted to form a red pigment. The qualitative color based ADE2 reporter gene has since been replaced with the quantitative Firefly Photinus pyralis (LUC1)^12,22. More recently, the ADE2 reporter has been combined with the lacZ reporter in an easy-to-score, semi-quantitative, double reporter assay that can be exploited to sub-classify P53 mutants according to their residual level of functionality²³.

Fluorescent reporters such as EGFP (enhanced green fluorescent protein) or DsRed (Discosoma sp. red fluorescent protein) have also been used for the quantitative evaluation of transactivation activity associated with all possible missense mutations in the TP53 coding sequence²⁴. Lastly, the chance of combining tunable promoters for P53 allele expression with isogenic yeast strains differing for the RE and/or reporter gene has led to the development of a data matrix that generates a refined classification of cancer-associated and germline mutant P53 alleles^25,26,27.

The approaches described above are used to measure the transcriptional activity of the P53 protein. However, the expression of wild-type P53 in the yeast S. cerevisiae²⁸ and Schizosaccharomyces pombe²⁹ can cause growth retardation, which has been associated with cell cycle arrest^28,30 or cell death³¹. In both cases, yeast growth inhibition is triggered by high P53 expression and has been correlated with potential transcriptional modulation of endogenous yeast genes involved in cell growth. Supporting this hypothesis, the loss-of-function mutant P53 R273H did not interfere with yeast cell growth when expressed at similar levels as wild-type P53³². Conversely, the expression in yeast of the toxic mutant P53 V122A (known for higher transcriptional activity compared to wild-type P53) caused a stronger growth inhibitory effect than wild-type P53³².

Additionally, it was demonstrated that human MDM2 was able to inhibit the human P53 transcriptional activity in yeast, promoting its ubiquitination and subsequent degradation³³. Accordingly, the ability of human MDM2 and MDMX to inhibit P53-induced yeast growth inhibition was demonstrated^32,34. In an additional study, a correlation between P53 transcriptional activity and actin expression levels was established, with the identification of a putative P53 RE upstream on the ACT1 gene in yeast³². Consistently, actin expression was enhanced by wild-type P53 and even more so by P53 V122A, but not by mutant P53 R273H. Conversely, actin expression by P53 decreased in the co-presence of P53 inhibitors MDM2, MDMX, or pifithrin-α (a small-molecule inhibitor of P53 transcriptional activity), consistent with results based on the yeast-growth assay. Importantly, these results established a correlation between P53-induced growth inhibition and degree of its activity in yeast, which has been also exploited to identify and study small molecules modulating P53 functions^28,34,35.

Protocol

1. Construction of ADE2 or LUC1 reporter yeast strains containing a specific RE (yAFM-RE or yLFM-RE)

1. Streak a yAFM-ICORE or yLFM-ICORE strain^12,14 (ICORE = I, ISce-I endonuclease under GAL1 promoter; CO = counter selectable URA3; RE = reporter KanMX4 conferring kanamycin resistance; Table 1) from a 15% glycerol stock stored at -80 °C on a YPDA agar plate (Table 2). Let it grow for 2-3 days at 30 °C.
2. Take one yeast colony from the fresh plate (no more than 3 weeks old) and place it in a small flask containing 5 mL of YPDA (Table 2). Incubate at 30 °C overnight, shaking at 150-200 rpm.
3. The next day, to remove all traces of dextrose, pellet the cells for 2 min at 3,000 x g and discard the supernatant by inversion of the tube.
   NOTE: Perform all centrifugations in this protocol at room temperature (RT).
4. Resuspend the cell pellet in 30-50 mL of pre-warmed CM (complete media) containing galactose (Table 2) and incubate for 4 h at 30 °C, shaking at 150-200 rpm (necessary for the induction of I-SceI).
5. Centrifuge the cells for 2 min at 3,000 x g and discard the supernatant by inversion of the tube.
6. Resuspend the cell pellet in 30-50 mL of sterile water. Repeat step 1.5.
7. Resuspend the cell pellet in 10 mL of sterile water. Repeat step 1.5.
8. Resuspend the cell pellet in 5 mL of sterile LiAcTE (Table 3), an ionic solution that favors DNA uptake. Repeat step 1.5.
9. Resuspend the cell pellet in 250 µL of sterile LiAcTE and transfer the cells to a 1.5 mL tube. Repeat step 1.5 and resuspend the cell pellet in 300-500 µL of sterile LiAcTE.
10. During the washes, denature a 10 mg/mL solution of salmon sperm DNA carrier for 10 min at 100 °C and chill immediately on ice to maintain it as single-stranded DNA.
11. In a separate 1.5 mL tube, add 500 picomoles of the desired oligonucleotide (Table 3), 5 µL of the boiled salmon sperm carrier DNA, 300 µL of sterile LiAcTE PEG (Table 3), and 50 µL of the yeast cell suspension (from step 1.9).
12. Vortex the tubes for 10 s to mix and incubate for 30 min at 30 °C, shaking at 150-200 rpm. Place the 1.5 mL tubes on the side to favor shaking.
13. Heat shock the yeast cells for 15 min at 42 °C in a heating block, then centrifuge the cells for 20 s at 10,000 x g. Remove the supernatant and resuspend the cells in 1 mL of sterile water.
14. Spread 100 µL of the cell suspension on YPDA agar plate and incubate (upside down) for 1 day at 30 °C. To ensure well-separated colonies are obtained, also spread 100 µL of a 1:10 dilution.
15. The next day, replica plate using sterile velvets onto CM agar plate containing dextrose and 5-FOA (Table 2). Consider a second replica plate on a new plate if many cells are transferred (i.e., some growth of URA3 cells).
16. Three days later, replica plate on non-selective YPDA and YPDA containing G418 (an aminoglycoside antibiotic similar to kanamycin) agar plates (Table 2), marking each plate to facilitate their subsequent comparison. Incubate the plates overnight at 30 °C.
17. The next day, identify the candidate reporter strains from colonies that are G418-sensitive but grow on YPDA plates (e.g., yLFM- or yAFM-RE colonies). Streak the identified colonies (3-6 colonies) on a new YPDA plate to obtain single colony isolates and let them grow for 2 days at 30 °C.
18. Patch single yeast colonies on a YPDA plate to isolate the colonies for further analyses. After 24 h at 30 °C, test them by replica plating on a YPGA agar plate (Table 2) that prevent the growth of petite mutants (i.e., respiratory-deficient mutants). At the same time, replica plate on a new YPDA agar plate.
19. Test the correct patches (i.e., growth on YPDA and YPGA agar plates; 1-3 colonies) from step 1.18 for the presence of correct oligonucleotide integration by colony PCR. Assemble a reaction mix by adding 5 μL of 10x PCR buffer (1.5 mM MgCl₂), 2 μL of 10 picomoles/μL primers (Table 3), 4 μL of 2.5 mM dNTPs, 0.25 µL of 5 U/μL Taq polymerase, and water to a final volume of 50 μL. Multiply the reaction mix for the number of yeast colonies that need to be screened and aliquot 50 μL into each PCR tube. Using a pipette, add a very small amount of yeast cells from the YPDA agar plate into a single PCR reaction mix.
20. Perform the PCR reaction with the following program: 94 °C for 8 min followed by 35 cycles of denaturation for 1 min at 94 °C, primers annealing for 1 min at 55 °C, and extension for 2 min at 72 °C.
21. After the reaction is completed, load an aliquot of the PCR reaction (about one-tenth the volume) on an agarose gel to check the correct size (~500 bp) .
22. Sequence the PCR product after purification with a commercial kit to confirm the integration of the desired RE sequence using the same primers of step 1.19.
23. After the validation of the correct sequence, make a 15% glycerol stock of yAFM-RE or yLFM-RE strain culture (in YPDA) and store it at -80 °C.

2. Evaluation of P53 protein transactivation ability using the qualitative color-based ADE2 yeast assay

1. Repeat steps 1.1 and 1.2 using the yAFM-RE strain (Table 1).
2. The day after, dilute the cell culture (1:10) in 30-50 mL of pre-warmed YPDA and continue to incubate at 30 °C by shaking until the OD_600nm reaches 0.8-1.0 (~2 h).
3. Repeat steps 1.5-1.10.
4. In a separate 1.5 mL tube, add 300-500 ng of yeast P53 (or control) expression vector (Table 4), 5 µL of the boiled salmon sperm DNA carrier, 300 µL of sterile LiAcTE PEG, and 50 µL of yeast cell suspension.
5. Repeat steps 1.12 and 1.13 but resuspend the cell pellet in 300 µL of sterile water.
6. Spread 100 µL of the cell suspension on synthetic selective (for P53 expression or control vector) plates containing dextrose as a carbon source and high amount of adenine (Table 2), then incubate (upside down) at 30 °C for 2-3 days.
7. Streak single yeast transformant colonies (2-6 streaks per plate) on a new selective plate and let them grow overnight at 30 °C.
8. The day after, using sterile velvets replica plate onto new selective plates that allow for the assessment of the color phenotype (i.e., plates containing dextrose as carbon source but limiting the amount of adenine ; Table 2). Incubate the plates upside down at 30 °C for 3 days. Optionally, to evaluate temperature sensitivity of P53 protein, incubate at three different temperatures for 3 days: 24 °C, 30 °C, and 37 °C.
   NOTE: The same streak can be replica plated multiple times.
9. To evaluate the P53 protein transactivation ability, check the color-based phenotype of yeast colonies and compare the P53 protein phenotype with respect to P53 wild-type and empty vector phenotypes.

3. Evaluation of P53 protein transactivation ability using the quantitative luminescence based LUC1 yeast assay

1. Transform yeast cells with P53 (or control) expression vectors (Table 4) using the LiAc based method described in protocol 2. Use yLFM-RE strain (Table 1).
2. Patch single transformants on a new selective plate with the high amount of adenine containing glucose as the carbon source and let them grow at 30 °C overnight. For each transformation type, make 5-7 different patches.
3. After overnight growth, resuspend a small amount of yeast cells using a sterile toothpick or pipette tip from the plate in synthetic selective medium containing dextrose or raffinose as the carbon source (200 µL final volume in a transparent 96 well plate, with a round or flat bottom). If the experiment requires inducible P53 expression, add galactose to the raffinose medium to modulate the level of induction (Table 2).
   NOTE: These cell suspensions should have an OD_600nm of about 0.4 and no higher than 1.
4. Measure the absorbance of each well at OD_600nm after inducible P53 expression (4-8 h at 30°C with 150-200 rpm shaking) using a multilabel plate reader. Make sure the cell suspensions are homogeneous by mixing each well with a multichannel pipette.
5. Transfer 10-20 µL of cell suspension from the transparent 96 well plate into a white 384 (or 96) well plate and mix with an equal volume (10-20 µL) of lysis buffer. Incubate for 10-15 min at RT on a shaker (150-200 rpm) to achieve permeabilization of the cell to the luciferase substrate.
6. Add 10-20 µL of Firefly luciferase substrate and measure the light units (LU) by a multi-label plate reader.
7. To determine P53 protein transactivation ability, normalize the LUs of each well to the corresponding OD_600nm (relative light unit, RLU). Calculate average RLU and standard deviation from 3-4 patches of yeast transformant colonies.
8. Compare the P53 protein transactivation data with respect to P53 wild-type and empty vector, either by subtracting the values obtained with the empty vector or by dividing by the values obtained with the empty vector (i.e., computing fold of induction).
   NOTE: The P53 transactivation activity can be also evaluated using the same experimental set-up in the presence of P53-interacting proteins (i.e., MDM2 and MDMX) and/or including drug treatment.

4. Evaluation of P53 protein growth inhibition using the yeast phenotypic assay

1. Transform yeast cells with P53/MDM2/MDMX (or control) expression vectors (Table 4) using the LiAc-based method described in section 2. Use the CG379 strain (Table 1) and spread yeast transformants on minimal selective plates (Table 2).
2. Grow transformed cells in minimal selective medium (Table 2) to approximately 1 OD_600nm.
3. Dilute yeast cells to 0.05 OD_600nm in the selective induction medium (Table 2), and optionally, add a chosen small molecule to the appropriate concentration (or solvent only) to test its efficacy in reactivating mutant P53 or in inhibiting MDM2-/MDMX-P53 interactions.
4. Incubate cells at 30 °C under continuous orbital shaking (200 rpm) for approximately 42 h (time required by negative control yeast to reach mid-log phase, about 0.45 OD_600nm).
5. Spot 100 µL aliquots of yeast cell cultures on minimal selective plates (Table 2).
6. Incubate for 2 days at 30 °C.
7. Measure yeast growth by counting the number of colonies obtained in the 100 µL culture drops (colony forming unit, CFU counts). For example, calculate the mutant reactivating effect of compounds considering the growth of wild-type P53 expressing yeast as the maximal possible effect (set to 100%), while the growth of cells expressing mutant P53 (but exposed to solvent control) represents the zero level of reactivation.

Representative Results

Construction of ADE2 or LUC1 reporter yeast strains

Thedelitto perfettoapproach^12,14,15,16 has been adapted to enable the construction of P53 reporter yeast strains (Figure 1B). The method employs single- or double-stranded oligonucleotides containing, on both ends, at least 30 nucleotides homologous to the chosen locus of integration. Specifically, the homology corresponds to sequences flanking the site of integration of the double marker cassette ICORE, containing the KlURA3 and kanMX4 (reporter gene providing resistance to Geneticin, G418) previously positioned at the desired target site^15,16. This cassette also contains the yeast I-SceI endonuclease coding sequence, under the inducible GAL1 promoter and its cognate target site. Therefore, a switch in galactose-containing medium right before the transformation of yeast cells with the desired sequence results in I-SceI expression and consequent generation of a single double-strand break (DSB) at the site of ICORE integration. The presence of a DSB highly stimulates the frequency of targeting events, with more than 1,000 replacements obtained with a single transformation.

At the end of the targeting and selection process based on acquired resistance to 5-FOA and sensitivity to G418, candidate clones are confirmed by colony PCR based amplification of the modified locus and Sanger sequencing to verify the correct integration of the desired P53 RE (Figure 1C). At the end of the strain construction protocol, that can be completed in about a week, a panel of isogenic reporter yeast strains are obtained that can be used to evaluate how differences in the sequence of the REs influence the transactivation capacity of the P53 protein.

Functionally heterogeneous mutant P53s

In human tumors, the TP53 gene is mainly affected by single missense mutations that generally involve six major hotspot residues (R175, G245, R248, R249, R273, and R282) located in the central DNA-binding domain (DBD) of the P53 protein. However, more than 2,000 single P53 amino-acid substitutions have been recorded, including some that are infrequently observed²⁷. Mutant P53s can be classified as DNA contact or structural mutants, depending on the effects of amino-acid substitution on the DNA-protein contact (e.g., R273H) or protein structure (e.g., R175H)³⁶. Single TP53 missense mutations impact P53 functions, generating a wide range of functional diversity, which can influence important clinical features such as tumor aggressiveness, chemo-resistance, and metastatic potential^37,38,39.

The concept that mutants P53 are functionally heterogeneous clearly has emerged in the last 15 years through a large amount of experimental data available in TP53 mutation databases (e.g., <p53.free.fr//>)²⁷. Different functional assays based on yeast and/or mammalian reporter systems have been developed, highlighting the different properties of mutants P53s (i.e., transactivation, temperature sensitivity, dominant-negative potential, interference with the members of the P53 family and interactions with other TFs)^{13,24,40,41,42,43,44}. The most comprehensive functional study examined more than 2,300 mutants in a yeast-based assay by characterizing their transactivation activity towards eight different P53 REs²⁴.

The data confirmed that nearly all mutant P53s involving hotspot residues are loss-of-function (i.e., they completely lost transactivation activity). Conversely, mutant P53s that hit other positions of the P53 protein and are generally found at moderate-to-low frequency in cancer⁴⁵ are partial function mutants, which retain some level of transactivation activity on P53 REs^13,17. An example of using the ADE2 assay to study the described P53 proteins functional heterogeneity is presented in Figure 2A. In fact, while R175H is a loss-of-function mutant producing red colonies in every condition tested, R282W showed temperature sensitivity that was more evident using galactose-dependent P53 expression (i.e., the red sector at 37 °C in the P21-5' RE strain in the plate containing lower galactose, which becomes pink in the higher galactose plate).Figure 2B presents a summary of results for an extended panel of P53 mutants and reporter strains.

The existence of this P53 functional heterogeneity prompted exploration of whether such heterogeneity paralleled the heterogeneity observed at the clinical level in subjects with TP53 germline mutations. TP53 germline mutations underlie the molecular basis of a group of cancer predisposition disorders, including the more severe Li-Fraumeni (LFS) and Li-Fraumeni-like (LFL) syndromes, and the less severe non-syndromic predispositions with (FH) or without (no FH) family history⁴⁶.

The protocol highlighted genotype-phenotype correlations by matching the yeast assay-based transactivation abilities of all TP53 germline mutant alleles with corresponding clinical data from the IARC database <http://www-p53.iarc.fr/Germline.html>. Loss-of-function P53 mutants have been found in more severe cancer proneness syndromes, while partial function P53 mutants has been found in less severe cancer proneness conditions. This indicates that P53 residual transactivation ability influences clinical phenotype in patients who inherited TP53 mutations and developed cancer^25,26.

Nucleotide sequences variations within REs govern P53transactivation potential

Human P53 is a tetrameric (dimer of dimers) TF. Each P53 dimer recognizes a RE consisting of 10 nucleotides (RRRCWWGYYY, R = A or G; W = A or T; Y = C or T)^47,48. Two such half-sites can be either adjacent to or spaced apart by up to 13-20 nucleotides, constituting a functional P53 RE. Nevertheless, P53 REs with a spacer are characterized by lower P53 affinity and transactivation than P53 REs without a spacer^49,50, in different functional assays from various systems.

It was recently demonstrated that half-sites can recruit P53 tetramers that are bound hemispecifically⁵¹, along with functional assays and chromatin immunoprecipitation studies, these findings indicate that P53 can also act at non-canonical REs, comprising half-sites and three-quarter sites^48,52. Furthermore, the fact that the full consensus P53 RE motif is very degenerate (RRRCWWGYYY)₂ prefigures the possibility that individual REs can substantially differ in sequence and binding affinity. Considering that hundreds of P53 target genes have been identified in the human genome^41,48, virtually all P53 REs showed non-identical sequence between each other. Moreover, it has been hypothesized that REs with a distinct DNA binding affinity are selected in the promoters of P53 target genes involved in cell cycle arrest or apoptosis⁵³.

The analysis in yeast of many variants of the P53 RE at a constant genomic location established the impact of nucleotide variants, spacer and organization of RE half-sites on transactivation capacity⁴⁹. It even led to the discovery of polymorphic P53 REs with the two alleles markedly differing in the responsiveness of an associated promoter to P53^14,54,55. Recently, information deriving from P53 RE sequence features and yeast-based transactivation potential has been coded in p53 Retriever, a pattern search algorithm that is able to localize and rank both canonical and non-canonical REs, according to their predicted transactivation potentials throughout the whole human genome⁵².

As an example of using the assay to study sequence-specific P53 transactivation potential, presented here is a comparison between human wild-type P53 and the evolutionary distant P53 protein from Caenorhabditis elegans (Figure 3). These results are a follow-up of a recent study in which the evolutionary divergence of transactivation specificity among P53 proteins was explored⁵⁶. Isogenic yLFM-RE reporter strains were developed as described in the protocol. Specifically, the Cep-1 P53 RE derived from the ced13 gene⁵⁷, and four variants (V1-V4) were compared (Figure 3A). The RE variants were constructed to examine the impact of varying the length of the spacer that separates the two decameric motifs (which in ced13, is of 28 nucleotides) and to examine changes in the nucleotides flanking the CWWG core motif (which in ced13, are A/T rich; while in the highest affinity human P53 REs, are G/C rich)⁵⁸. The results clearly show how Cep1 and human P53 diverge in terms of transactivation specificity. In fact, while human P53-mediated transactivation is strongly inhibited by the presence of a spacer (Figure 3B), Cep-1 activity is inhibited by the removal of the spacer (Figure 3C). Furthermore, Cep-1-mediated transactivation is abolished when the RE is modified from A/T- into G/C-rich. Neither human P53 nor Cep-1 can transactivate from a single decamer derived from the ced13 RE.

Searching for small molecule disruptors of P53-MDM2/MDMX interactions and reactivators of mutant P53 activity

The established correlation between the growth inhibitory effect of human P53 and degree of its activity in yeast led to a simplified screening assay based on measurements of yeast cell growth to analyze the impact of interfering factors on P53 activity (Figure 4). The effectiveness of the yeast phenotypic assay to screen for potential anticancer agents has been demonstrated in several works. Using yeast cells co-expressing P53 and its inhibitors MDM2 and/or MDMX, new disruptors of these interactions were identified by their ability to inhibit the negative effect of MDMs on P53, thus re-establishing wild-type P53-induced yeast growth inhibition (Figure 4A). In particular, this assay led to the discovery of 1) pyranoxanthone 1 as the first P53-MDM2 interaction inhibitor with a xanthone scaffold³⁴ and 2) oxazoloisoindolinone inhibitors of the P53-MDM2 interaction⁵⁹. Additionally, with this assay, α-mangostin and gambogic acid were described for the first time as potential inhibitors of the P53-MDM2 interaction³⁰.

Later, the same assay demonstrated that prenylation of chalcones enhanced their ability to disrupt the P53-MDM2 interaction⁶⁰. Interestingly, this yeast assay also led to the discovery of two inhibitors of the P53-MDM2/MDMX interactions: the tryptophanol-derived oxazolopiperidone lactam OXAZ-1⁶¹ and tryptophanol-derived oxazoloisoindolinone DIMP53-1⁶².

The reduced impact of loss-of-function mutant P53 on yeast cell growth has also been explored to screen for mutant P53-reactivating drugs, characterized by the ability to restore wild-type-like growth inhibitory activity to mutant P53 (Figure 4B). With this yeast assay, a reactivator of mutant P53 R280K, the enantiopure tryptophanol-derived oxazoloisoindolinone SLMP53-1⁶³, was identified. Figure 4C,D shows representative results obtained in yeast with the expression of P53 or treatments with the inhibitor of the P53-MDM2 interaction Nutlin-3a or with the reactivator of P53 mutant Y220C PhiKan083.

Validation of the molecular mechanism of action of these compounds as P53-activating agents as well as their antitumor activity, both in human tumor cell lines^{34,60,61,62,63} and animal models^62,63, attests to the great potential of the yeast phenotypic assay in drug discovery.

Figure 1: Features of yeast based P53 transactivation assays and workflow for the construction of P53 reporter yeast strains.(A) The ease of genome manipulation and controlled ectopic gene expression renders yeast akin to an "in vivo test tube", in which several variables relevant to examine P53 transactivation functions are rigorously explored. These variables include the levels of expression of a wild-type or mutant P53 protein, the sequence of the RE, and type of reporter gene [qualitative/semi-quantitative (e.g., ADE2 and LACZ) or quantitative (e.g., LUC1)]. The consensus RE sequence is highly degenerated (RE = RRRCWWGYYY-n-RRRCWWGYYY; R = purine; W = A/T; Y = pyrimidine; CWWG = CORE sequence, n = 0-13 bps spacer). The numbers of mismatches with respect to the consensus, the CORE sequence, and number of base pair spacers will influence the transactivation activity. This panel has been modified from a previous publication⁶⁴. (B) Scheme of the delitto perfetto approach to introduce a desired P53 RE upstream of the minimal promoter driving the ADE2 or LUC1 reporter gene expression. The protocol was adapted from previous publications^12,15,16. (C) Example of the results of yeast colony PCR amplification of the region surrounding the oligo RE integration site. The image of electrophoresis presents the amplification result of two putative positive colonies (URA3 minus and G418 sensitive) with the corresponding PCR negative control next to the 1 kb DNA ladder (Promega). The ~500 nt band expected using primers described in Table 3 can be sequenced to confirm correct promoter editing, as shown in the electropherogram. Please click here to view a larger version of this figure.

Figure 2: P53 proteins show RE-specific and temperature-sensitive transactivation potential. Wild-type P53 and the indicated TP53 missense mutations were tested using the qualitative color-based reporter assay in yeast and exploiting reporter strains that harbor different P53 REs controlling the ADE2 gene expression. (A) An example of the colony color phenotype for P53 wild-type, R175H, and R282W with the PUMA (BBC3) and the P21-5' REs at 30 °C and 37 °C is shown. Moderate (0.008% galactose) and high (0.12% galactose) P53 expression is compared. (B) Results obtained with an extended set of P53 mutants and yeast reporter strains examined for the P53-dependent colony color phenotype at three temperatures (24°C, 30 °C, and 37 °C). Results exemplify functional heterogeneity of mutant P53 alleles and are summarized in a table format. For instance, R175H is a loss-of-function mutant, while G199H is wild-type P53-like. K139E and R282W retain partial function and exhibit cold sensitivity and heat sensitivity, respectively. Please click here to view a larger version of this figure.

Figure 3: Human P53 and credited ortholog Cep-1 exhibit highly diverged transactivation specificity. (A) Sequences of the: human P53 RE consensus sequence, Cep-1 P53 RE derived from the ced13 gene, and four variants (V1-V4) that were tested in the functional assays. (B) Transactivation was measured after inducing human P53 expression for 8 h using three different carbon sources to achieve low, moderate and high expression (0.008%, 0.032%, 0.12% galactose added to selective medium containing 2% raffinose). Results are expressed as average relative light units (RLU) and standard error of four replicates. Reporter activity for cell transformed with empty expression vector was subtracted out. (C) Transactivation specificity of Cep1 as measured in (B). For both proteins, transactivation levels were proportional to the amount of galactose. The higher light units measured when human P53 is expressed can be dependent on different levels of protein amounts produced, due to difference in protein length and potentially also in codon usage. The two proteins are characterized by different transactivation specificity towards the different REs tested and this property is not affected by possible differences in relative protein amount. Please click here to view a larger version of this figure.

Figure 4: Small molecule activators of P53 identified using the yeast phenotypic assay. (A) Yeast cells co-expressing human wild-type P53 and MDM2 or MDMX have been used to search for inhibitors of the P53-MDM2 and P53-MDMX interactions. In this system, interaction inhibitors restore P53-induced yeast growth inhibition in yeast cells co-expressing P53 and MDM2 or MDMX. This approach has allowed identification of inhibitors of the P53-MDM2 interaction and identification of dual inhibitors of P53-MDM2 and P53-MDMX interactions. (B) Yeast cells expressing human mutant P53 have been used to search for mutant P53 reactivators and are able to restore the wild-type-like P53-induced yeast growth inhibition in mutant P53-expressing yeast cells. (C) Representative images of the plate spot assay showing the impact of wild-type P53 or mutant Y220C on yeast colony growth compared to the empty vector (see step 4.5). (D) Quantitative results of the growth assay: 10 µM Nutlin-3a restores P53-induced yeast growth inhibition in cells co-expressing MDM2; 50 µM PhiKan083 restores wild-type-like P53-induced yeast growth inhibition to mutant P53 Y220C. Both effects are plotted relative to the yeast growth inhibitory effect of wild-type P53, which was set as one. Please click here to view a larger version of this figure.

Strain name and genotype	Specific use	Protocol number
yAFM-ICORE: MATa leu2-3,112 trp1-1 his3-11,15 can1-100; ura3-1; ICORE::pCyc1::ADE2	It is used for the construction of the  yAFM-RE strain (MATa leu2-3,112 trp1-1 his3-11,15 can1-100; ura3-1; ICORE::pCyc1::ADE2) that is exploited in the qualitative, color-based ADE2 assay.	Protocols 1,2
yLFM-ICORE: MATa leu2-3,112 trp1-1 his3-11,15 can1-100; ura3-1; ICORE::pCyc1::LUC1	It is used for the construction of the  yLFM-RE strain (MATa leu2-3,112 trp1-1 his3-11,15 can1-100; ura3-1; RE::pCyc1::LUC1) that is exploited in the quantitative luminescence-based LUC1 assay.	Protocols 1,3
CG379: MATa, ade5 his7-2 leu2-112 trp1-289aura3-52 [Kil-O]	It is used for the phenotypic growth assay.	Protocol 4

Table 1: Yeast strains.

Media name and recipe	Specific use	Protocol number
YPDA (2% yeast extract, 2% peptone, 2% dextrose, 200 mg/L adenine) + 2% agar	YPDA without agar is sterilized preferably by filtration. YPDA + agar is sterilized by autoclaving; it is possible to omit dextrose from the recipe and add dextrose from a 20% filter sterilized solution in water before pouring the plates.	Protocols 1-4
CM [0.17% Yeast Nitrogen Base without AA and Ammonium Sulfate, 0.5% Ammonium Sulfate, 5% (volume/volume, v/v) non-essential amino-acid solution, 20 mg/L histidine, 20 mg/L tryptophan, 20 mg/L uracil, 30 mg/L lysine, 100 mg/L leucine, 200 mg/L adenine] + 2% galactose	The media is sterilized by filtering. The following sterile stock solutions are prepared in water (with the exception of uracil, that is dissolved in 0.1M NaOH) by filtering: non-essential amino-acid solution (0.04% arginine, 0.04% methionine, 0.06% isoleucine, 0.1% phenylalanine, 0.2% glutamic acid, 0.2% aspartic acid, 0.3% valine, 0.4% threonine, 0.8% serine), 1% histidine, 1% tryptophan, 1% uracil, 1% lysine, 1% leucine, 0.5% adenine, 20% galactose. Adenine stock solution must not be stored in fridge due to the formation and the sedimentation of crystals. Tryptophan stock solution must be stored in the dark.	Protocol 1
CM  (see above) + 2% Dextrose + 1g/L 5-FOA + 2% agar	The media (containing Yeast Nitrogen Base without AA and Ammonium Sulfate, Ammonium Sulfate and Agar) is sterilized by autoclaving. The other components are added after autoclaving to preserve heat-labile ingredients; the 5-FOA powder can be added directly in the media once it has cooled to about 55°C.	Protocol 1
YPDA +  400 µg/mL G418 + 2% agar	G418 must be added after autoclaving once media has cooled to about 55°C. If starting from powder, a 50 mg/mL G418 stock solution can be prepared in water and sterilized by filtration.	Protocol 1
YPGA [2% yeast extract, 2% peptone, 2% glycerol (v/v), 200 mg/L adenine] + 2% agar	The media is sterilized by autoclaving.	Protocol 1
Synthetic selective plate: [0.17% Yeast Nitrogen Base without AA and Ammonium Sulfate, 0.5% Ammonium Sulfate, 5% (v/v) non-essential amino acid solution (0.04% arginine, 0.04% methionine, 0.06% isoleucine, 0.1% phenylalanine, 0.2% glutamic acid, 0.2% aspartic acid, 0.3% valine, 0.4% threonine, 0.8% serine), 20 mg/L histidine, 20 mg/L uracile, 20 mg/L tryptophan (depending on vector selection marker), 30 mg/L lysine, 100 mg/mL leucine (depending on vector selection marker), 5 mg/L or 200 mg/L adenine] + 2% dextrose + 2% agar	The media (containing Yeast Nitrogen Base without AA and Ammonium Sulfate, Ammonium Sulfate and Agar) is sterilized by autoclaving. The other components are added after autoclaving to preserve heat-labile ingredients. When using pLS- and pLSG-based vectors prepare plates without leucine. When using pTS- and pTSG-based vectors prepare plates without tryptophan.	Protocols 2-3
Synthetic selective media: [0.17% Yeast Nitrogen Base without AA and Ammonium Sulfate,  0.5% Ammonium Sulfate, 5% (v/v) non-essential amino-acid  solution (0.04% arginine, 0.04% methionine, 0.06% isoleucine, 0.1% phenylalanine, 0.2% glutamic acid, 0.2% aspartic acid, 0.3% valine, 0.4% threonine, 0.8% serine), 20 mg/L histidine, 20 mg/L uracile, 20 mg/L tryptophan (depending on vector selection marker), 30 mg/L lysine, 100 mg/mL leucine (depending on vector selection marker), 200 mg/L adenine] + 2% dextrose or raffinose + 0-2% galactose	The media is sterilized by filtration. When using pLS- or pTS-based vectors in drug treatment prepare media without leucine or tryptophan, respectively but containing 2% dextrose. When using pLSG- or pTSG-based vector in inducible P53 expression with or without drug treatment prepare media without leucine or tryptophan, respectively but containing 2% raffinose and variable amount of galactose to modulate P53 expression. The extent of P53 induction can also be modulated by varying the time of incubation.	Protocol 3
Minimal selective plates: 2%  glucose, 0.7%  yeast nitrogen base (without amino-acids and ammonium sulfate), 50 mg/L adenine, 50 mg/L uracil, 50 mg/L histidine, 50 mg/L tryptophan (depending on vector selection marker), 50 mg/L leucine (depending on vector selection marker), 2%  agar	The medium is sterilized by autoclaving. When using pLS89-based vectors prepare medium without tryptophan. Use pLS89-empty vector (not expressing any protein) as negative control. When using pLS89-based and pGADT7-MDM2 vectors (co-expression of P53 and MDM2), prepare medium without leucine and tryptophan. Use pLS89-empty and pGADT7 (both not expressing any protein) as negative controls. When using pLS76-based vectors prepare medium without leucine. Use pRS315 vector (not expressing any protein) as negative control.	Protocol 4
Minimal selective medium: as minimal selective plates but without agar	The medium is sterilized by filtration.	Protocol 4
Selective induction medium: as minimal selective medium but containing 2%  galactose instead of glucose	The medium is sterilized by filtration.	Protocol 4

Table 2:Yeast media.

Solution  name  and recipe or primer name and sequence	Specific features and  use	Protocol number
LiAc/TE: 10 mM Tris-HCl, pH 8.0, with 1 mM EDTA and 0.1M lithium acetate	The following sterile stock solutions in water are prepared by autoclaving: 1M Lithium acetate pH 7.5, 1 M Tris-HCl / 0.1 M EDTA pH 8.0 (TE buffer 10X).	Protocols 1-4
PEG/LiAc/TE: 40% Polyethylene glycol (PEG) in 10 mM Tris-HCl, pH 8.0, with 1 mM EDTA and 0.1 M Lithium acetate	The following sterile stock solutions in water are prepared by autoclaving: 1 M Lithium acetate pH 7.5, 1 M Tris-HCl / 0.1 M EDTA pH 8.0 (TE buffer 10X), 50% PEG (molecular weight ~3350).	Protocols 1-4
Homology tails for the RE oligonucleotides: 5’-gcggaattgactttttcttgaataatacat-RE-gcagatccgccaggcgtgtatatagcgtgg-3’	Provided is the sequence of the homology tails, corresponding to the chromosomal sequences flanking the integrated ICORE cassette; RE = RE sequence.	Protocol 1
Primers: Ade2 Fw: 5’-AAGTTGCCTAGTTTCATGAA-3’;                Ade2 Rv: 5’-GGAGCCATTAACGTGGTCAT-3’;                Luc1-Rv:  5’-CATAGCTTCTGCCAACCGAA-3’	For yAFM-RE strain (PCR and sequencing) combine Ade2 Fw with Ade2 Rv primer. For yLFM-RE strain (PCR and sequencing) combine Ade2 Fw with Luc1-Rv primer.	Protocol 1

Table 3: Solutions and oligonucleotides.

Types of vectors	Positive and negative controls	Protocol number
P53 (or control) vectors. pLS- or pLSG-based: P53 protein expression under constitutive ADH1 promoter or galactose inducible GAL1 promoter, respectively (LEU2 as selection marker). pTS- or pTSG-based: P53 protein expression under constitutive ADH1 promoter or  galactose inducible GAL1 promoter, respectively (TRP1 as selection marker)	pLS- and pLSG-based: use as positive control the pLS76 and the pLSG-P53 vectors (expressing wild type P53), respectively; use as negative control pRS315 vector (not expressing any protein). pTS- and pTSG-based: use as positive control the pTS76 and the pTSG-P53 vectors (expressing wild type P53), respectively; use as negative control pRS314 vector (not expressing any protein).	Protocol 2,3
P53 and MDM2 (or control) vectors. pLS89-based: wild type P53 protein expression under galactose inducible GAL1 promoter (TRP1 as selection marker).  pLS76-based: mutant P53 protein expression under constitutive ADH1 promoter (LEU2 as selection marker). pGADT7-based: MDM2 protein expression under constitutive ADH1 promoter (LEU2 as selection marker)	pLS89-based: use as positive control the pLS89-P53 vector (expressing wild type P53); use as negative control pLS89-empty (not expressing any protein). pLS76-based: use as positive control the pLS76-mutant P53 vector (expressing mutant P53); use as negative control pRS315 vector (not expressing any protein). pGADT7-based: use as positive control the pGADT7-MDM2 vector (expressing wild type MDM2); use as negative control pGADT7 vector (not expressing any protein).	Protocol 4

Table 4: Yeast expression vectors.

Discussion

Yeast-based assays have proven useful to investigate various aspects of P53 protein functions. These assays are particularly sensitive for evaluating P53 transactivation potential towards variants of RE target sites, including the evaluation of functional polymorphisms. The use of color reporters as well as miniaturization of the luciferase assay result in cost-effective and relatively scalable assays. Also, the growth inhibition test is potentially amenable to being used in chemical library screening, automating the quantification of yeast cell viability through the measurement of absorbance. While limited permeability due to the cell wall and action of ABC transporters has been considered a limitation in using yeast for drug screening, genetic modifications to improve uptake have been successfully developed^22,65.

Compared to mammalian cell-based assays, the yeast-based P53 assay may improve the identification of direct hits, given that P53 is isolated from the amazing complexity of cofactors and signaling cascade pathways that modulate its functions in higher eukaryotes. Yeast-based functional assays have also been partially successful in monitoring the interaction of P53 with protein cofactors (namely MDM2 and MDMX^22,32,33,34) and impact of small molecules (such as Nutlin-3a) on those interactions. However, the sensitivity and predictive power of yeast-based assays to investigate crosstalk between P53 and interacting proteins may be somewhat limited, depending on how those interactions in vivo depend on post-translational modifications and species-specific signaling cascades.

The systems described here can be easily adapted to other TFs. Indeed, yeast functional assays have been developed for P53-related P63 and P73 proteins^42,49 as well as members of the NF-κB family^66,67 [or even for some homeobox and basic helix-loop-helix (bHLH) TFs^68,69]. Human nuclear receptors have also been expressed in yeast and proven to work as ligand-dependent, sequence-specific TFs⁷⁰. One limiting factor has been identified in the weak capacity of some mammalian transactivation domain (TAD) to interact with the yeast basal transcription machinery⁸, a shortcoming that can be overcome by using chimeric constructs including an active heterologous TAD^68,69.

Materials

Name	Company	Catalog Number	Comments
L-Aspartic acid	SIGMA	11189
QIAquick PCR Purification Kit	QIAGEN	28104
L-Phenylalanine	SIGMA	78019
Peptone	BD Bacto	211677
Yeast ex+A2:C26tract	BD Bacto	212750
Difco Yeast Nitrogen Base w/o Amino Acids and Ammonium Sulfate	BDTM	233520
Lithium Acetate Dihydrate	SIGMA	517992
Bacteriological Agar Type A	Biokar Diagnostics	A1010 HA
G418 disulfate salt	SIGMA	A1720
Ammonium Sulfate	SIGMA	A2939
L-Arginine Monohydro-chloride	SIGMA	A5131
Adenine Hemisulfate Salt	SIGMA	A9126
Passive Lysis Buffer 5x	PROMEGA	E1941
Bright-Glo Luciferase Assay System	PROMEGA	E2620
5-FOA	Zymo Research	F9001
D-(+)-Galactose	SIGMA	G0750
L-Glutamic acid	SIGMA	G1251
Dextrose	SIGMA	G7021
L-Histidine	SIGMA	H8125
L-Isoleucine	SIGMA	I2752
L-Lysine	SIGMA	L1262
L-Leucine	SIGMA	L8000
L-Methionine	SIGMA	M2893
PEG	SIGMA	P3640
D-(+)-Raffinose Pentahydrate	SIGMA	R0250
L-Serine	SIGMA	S4500
L-Tryptophan	SIGMA	T0271
L-Threonine	SIGMA	T8625
Uracil	SIGMA	U0750
L-Valine	SIGMA	V0500