Domain organization of p53 isoforms.

Overview of the so far described p53 isoforms that are formed by a combination of four N-terminal (TAp53, Δ40p53, Δ133p53 and Δ160p53) and three C-terminal variants (α, β and γ), leading to 12 individual p53 isoforms. The remaining two isoforms p53ψ (or TAp53ψ) and Δp53 (or TAΔp53α) are generated by alternative splicing.

Activity, oligomerization and DNA binding.

Luciferase reporter assay of p53 isoforms and variants either on the pBDS-2 reporter (three repeats of the 14-3-3σ promoter RE) (A), or on the Mdm2 promoter (B), or on the p21 promoter (C). TAp53α carrying the cancer-related R175H mutation served as a negative control. H1299 cells (A, B) or Saos-2 cells (C) were transiently transfected with the respective luciferase reporter plasmids and the N-terminally Myc-tagged proteins. The plasmid encoding for TAp53α was titrated with the remaining DNA amount for transfection being filled up with empty vector. (D) Luciferase reporter assay of p53ψ on the pBDS-2 reporter. TAp53α carrying the cancer-related R175H mutation served as a negative control. H1299 cells were transiently transfected with the respective luciferase reporter plasmids and the N-terminally Myc-tagged proteins. (E) DNA pull-down assay of p53 isoforms with the 20bp REs of the human PUMA and p21 promoter as bait. TAp53α carrying the cancer-related R175H mutation served as a negative control. N-terminally Myc-tagged proteins were in-vitro translated using RRL. For the relative pull-down efficiency each pulldown signal was normalized to the input signal. (F) SPR affinity curves of purified Δ40p53β (red), p53 DBD-OD variants (purple and black) and the isolated p53 DBD (orange) binding to the 20bp p21 RE immobilized on a streptavidin (SA) chip. Data points were extracted by equilibrium analysis of sensograms as in (Supplementary Figure S1G), plotted and fitted with a non-linear, least squares regression using a single-exponential one-site binding model with Hill slope. p53 DBD-OD (41-356) encompasses the complete OD, while p53 DBD-OD (41-331) only has the N-terminal part of the OD common to all p53α, p53β and p53γ isoforms. (A-E) The bar diagram shows the mean values and error bars the corresponding SD (n=3). Statistical significance was assessed by ordinary one-way ANOVA (n.s.: P > 0.05, *: P ≤ 0.05, **: P ≤ 0.01, ***: P ≤ 0.001, ****: P ≤ 0.0001).

Probing the DBD Fold.

(A) Superposition of structures of the p53 DBD (grey) in complex with DNA (orange; PDB: 3TS8), HPV16 E6 (blue; PDB: 4XR8) and DARPin G4 (green; PDB: 7Z7E). Epitope of the pAB240 is marked in red and the zinc ion is depicted as a purple sphere. (B) E6 degradation assay of p53 isoforms, cancer-related mutants, and variants. ΔNp63α served as a negative control. N-terminally Myc-tagged proteins were in-vitro translated using RRL. Lysates were diluted in reaction buffer and supplemented with either 5 µM GST-tagged HPV16 E6 or GST only as control. Reactions were incubated for 4 h at 25°C and analyzed for protein levels by WB. Signals were normalized to the loading control vinculin and the relative protein level ratio between E6 and GST samples were calculated by setting the normalized signal of the GST samples to 1. (C) Conformation-specific immunoprecipitation (Conf-IP) of p53 isoforms and cancer-related mutants. H1299 cells were transiently transfected with empty vector or N-terminally Myc-tagged p53 variants. p53 was immunoprecipitated (IP) with either α-mouse IgG or an α-p53 antibody (pAB240). The latter binds an epitope in the DBD of p53, which is only exposed when the domain is unfolded. Consequently, pAB240 only recognizes intrinsically unfolded p53 mutants under native IP conditions. Input and IP samples were subsequently analyzed by WB using α-Myc antibody. For the relative Conf-IP efficiency each IP signal was normalized to the input signal. (D) DARPin pull-down assay of p53 isoforms and cancer-related mutants in-vitro translated using RRL or transiently expressed in H1299 (E) with immobilized DARPin G4. DARPin G4 only recognizes the folded DBD of p53. For the relative pull-down efficiency each pulldown signal was normalized to the input signal. (B,C,D,E) The bar diagram shows the mean values and error bars the corresponding SD (n=3). Statistical significance was assessed by ordinary one-way ANOVA (n.s.: P > 0.05, *: P ≤ 0.05, **: P ≤ 0.01, ***: P ≤ 0.001, ****: P ≤ 0.0001).

Aggregation and Solubility.

(A) Predicted aggregation propensity of TAp53α (black) and the C-terminal isoforms TAp53β (auburn), TAp53γ (purple) and TAp53ψ (pink). TAp53α is predicted to have a total of four APRs (APR1-4) sharing the first three with TAp53β and TAp53γ. TAp53γ and TAp53ψ each introduced a novel APR (APR5 and APR6) with their specific C-termini. Schematic illustration of the boundaries of all N- and C-terminal p53 isoforms is shown below the graph. Position and length of the novel sequences introduced by the C-terminal are shown with the corresponding colours. All in-silico predictions were run using the TANGO algorithm with the following parameters: pH 7.5, 150 mM ionic strength and at a temperature of 37°C. (B) Sequence alignment of p53α with the p53β and p53γ specific C-termini. The boundaries of the OD are indicated by the grey line and the predicted APRs from (A) are highlighted in the respective colours. (C) Sequence alignment of p53α the p53 ψ specific C-terminus. The boundaries of the DBD are indicated by the grey line and the predicted APRs from (A) are highlighted in the respective colours. (D) Analytical SEC of p53 isoforms and cancer-related mutants. H1299 cells were transiently transfected with the N-terminally Myc-tagged proteins and cell lysates were loaded onto a Superose 6 SEC column. Both, lysis and running buffer contained 20 mM CHAPS. Collected fractions were analysed for p53 by WB using an α-Myc antibody. An elution volume of 0.820 ml corresponds to the void volume of the column (2.4 ml bed volume). (E) BN-PAGE of p53 isoforms and the cancer-related R175H mutant. H1299 cells were transiently transfected with the N-terminally Myc-tagged proteins. Cell lysates were subsequently analysed by BN-PAGE and SDS-PAGE (shown in figure supplement 3B) followed by WB using α-Myc antibody for detection. High molecular weight species corresponding to aggregates are marked by ‘a’. (F) Solubility assay of p53 isoforms and cancer-related mutants. H1299 cells were transiently transfected with the indicated Myc-tagged proteins and lysed in a buffer supplemented with Triton X-100. Soluble and insoluble components were separated by centrifugation. The insoluble fraction in the pellet was solubilized with a buffer supplemented with SDS. Samples of both fractions were analysed by WB using α-Myc antibody. (G) Thioflavin T (ThT) fluorescence assay of the p53β (aa 322-341) and p53γ (322-346) C-termini. The peptides, solubilized in denaturation buffer, and denaturation buffer only as a control, were diluted 20-fold in assay buffer supplemented with ThT and incubate at 37°C for 45 min. The final concentration of peptides and ThT was 20 µM and 25 µM, respectively. (F, G) The bar diagram shows the mean ThT fluorescence and error bars the corresponding SD (n=3). Statistical significance was assessed by ordinary one-way ANOVA (n.s.: P > 0.05, ****: P ≤ 0.0001).

Chaperons.

(A) Co-Immunoprecipitation of p53 isoforms and cancer-related mutants with endogenous HSC/HSP70. H1299 cells were transiently transfected with either empty vector or the indicated N-terminally Myc-tagged p53 variants. HSC/HSP70 was immunoprecipitated (IP) with an α-HSC/HSP70 antibody. Input and IP samples were subsequently analysed by WB using α-HSC/HSP70 and α-Myc antibody to detect HSC/HSP70 and p53 variants, respectively. A light and dark exposure of the IP samples detected with α-Myc antibody is shown. Vinculin served as a loading control for the input samples. (B) Luciferase reporter assay of p53 isoforms and mutants as well as the indicated Scarlet fusion proteins on the HSP70 promoter. H1299 cells were transiently transfected with the respective luciferase reporter plasmids and the N-terminally Myc-tagged proteins. (B) p53 (..331) contains only the C-terminal part common to p53α, p53β and p53γ. The fluorescent protein Scarlet was fused with the C-termini of p53 isoforms. Scarlet alone served as a negative control. (C) Luciferase reporter assay of p53 isoforms and mutants as well as the indicated Scarlet fusion proteins on the HSE promoter (containing three repeats of heat shock element). H1299 cells were transiently transfected with the respective luciferase reporter plasmids and the N-terminally Myc-tagged proteins. (C) p53 (..331) contains only the C-terminal part common to p53α, p53β and p53γ. The fluorescent protein Scarlet was fused with the C-termini of p53 isoforms. Scarlet alone served as a negative control. (B,C) The bar diagram shows the mean fold induction relative to the empty vector control and error bars the corresponding SD (n=3). Statistical significance was assessed by ordinary one-way ANOVA (n.s.: P > 0.05, *: P ≤ 0.05, **: P ≤ 0.01, ***: P ≤ 0.001, ****: P ≤ 0.0001). (D) Chaperons and other proteins associated with binding un-/misfolded proteins are shown, which are significantly enriched in the mass spectrometry analysis of p53 isoforms. The red nods represent the identified proteins assigned by the grey lines to the p53 isoforms they were enriched for. Significant hits were determined by setting the parameters: log2 enrichment greater than or equal to 0.5 and p-value less than 0.05 and proteins were filtered for the keywords (‘chaperone’, ‘unfolded protein binding’ and ‘misfolded protein binding’) in the GOBP (Gene Ontology Biological Process) and GOMF (Gene Ontology Molecular Function) terms. The plot was generated using DiVenn (v2.0).

Inactivation of p53 family members.

(A) Luciferase reporter assay of TAp53α in combination with p53 isoforms and cancer-related mutants on the pBDS-2 reporter. H1299 cells were transiently transfected with the respective luciferase reporter plasmids, the N-terminally Flag-tagged TAp53α alone or together with the N-terminally Myc-tagged p53 variants. Luciferase reporter assay of TAp73α (B), TAp73β (C) or TAp63γ (D) in combination with p53 and p63 isoforms and cancer-related mutants on the pBDS-2 reporter. H1299 cells were transiently transfected with the respective luciferase reporter plasmids, the N-terminally HA-tagged TAp73α (B), TAp73β (C) or TAp63γ (D) alone, empty vector or together with the N-terminally Myc-tagged p53/p63 variants. (A-D) The bar diagram shows the mean fold induction relative to the empty vector control and error bars the corresponding SD (n=3). Statistical significance was assessed by ordinary one-way ANOVA (n.s.: P > 0.05, *: P ≤ 0.05, **: P ≤ 0.01, ***: P ≤ 0.001, ****: P ≤ 0.0001). (E) Luciferase reporter assay of TAp63γ in combination with the Δ40p53α isoforms and cancer-related mutants on the pBDS-2 reporter. H1299 cells were transiently transfected with the respective luciferase reporter plasmids, the N-terminally Myc-tagged TAp63γ alone or together with the N-terminally Myc-tagged Δ40p53α variants. (F) Conformation-specific immunoprecipitation (Conf-IP) of p53, p63 or p73 isoforms and p53 variant (Δ40p53α, Δ40p53β, Δ133p53α and Δ133p53β). H1299 cells were transiently transfected with N-terminally HA-tagged p53, p63 or p73 isoforms and N-terminally Myc-tagged p53 variants (Δ40p53α, Δ40p53β, Δ133p53α and Δ133p53β). p53, p63, or p73 isoforms were immunoprecipitated (IP) with α-HA. p53α isoforms mostly hetero-tetramerize with different p53α isoforms, but co-aggregate with remaining p53 isoforms. p63 and p73 isoforms interact with p53 isoforms by co-aggregation. Input and IP samples were subsequently analysed by WB using α-Myc antibody.

Mechanisms of inactivation within the p53 family.

p53 isoforms have the potential to inactivate p53 family members via different mechanisms. (A) Either p53α isoforms form inactive hetero-tetramers (and/or aggregate) with wtp53 (Δ40p53α, Δ133p53α, Δ160p53α) and therefore inhibit wtp53 or inactive wtp53 by promotor squelching (Δ40p53α). (B) Furthermore, Δ40p53α isoform can as well inactivate tetrameric p63/p73 via promoter squelching. DBD-truncated aggregating p53 isoforms can co-aggregated with tetrameric p63α and p73α isoforms, but not with the closed TAp63α dimer.