Fig 2: The mechanism of MEG3 in EC. MEG3 was found to downregulate MDM2 to repress p53 ubiquitination and degradation, activating p53 to elevate miR-149-3p expression, which inhibited regulatory T cell differentiation and immune escape in EC mice by targeting FOXP3

Fig 3: Single-cell quantification of p53-dependent transcription highlights distinct patterns of gene expression upon DNA damage Fluorescence microscopy images of smFISH probes CAL Fluor 610 (gray) overlayed with Hoechst 33342 staining (blue) at 3 h after 10 Gy IR or the indicated target genes in A549 cells. Scale bar corresponds to 10 µm distance; images were contrast and brightness enhanced for better visualization.We quantified RNAs per cell for the indicated target genes before (basal, gray) and 3 h (red), 6 h (blue) and 9 h (orange) after DNA damage (10 Gy IR). smFISH-based single-cell analysis of gene expression patterns highlights distinct RNA counts for p53 targets. RNA counts per cell are displayed as boxplots (see Data visualization section); lines indicate medians of distributions; boxes include data between the 25th and 75th percentiles; whiskers extend to maximum values within 1.5× the interquartile range. Notches represent 5% confidence intervals for the median. n: number of analyzed cells, fc: median fold of induction relative to time point basal (indicated by gray lines)Distributions of RNAs per cell for the indicated target genes before (basal, gray) and 3 h (red), 6 h (blue), and 9 h (orange) after DNA damage (10 Gy IR). Despite a clear change in median levels (m: median), single-cell analysis reveals a strong dispersion that overlaps for the different conditions, as shown by the strongly overlapping distributions of RNA counts per cells. For better visualization, probability density estimates (PDF) based on a normal kernel are shown (see Data visualization section, compare Appendix Fig S4, raw measurements available as figure source data). Source data are available online for this figure.

Fig 4: The interplay of p53's C-terminal lysine acetylation and methylation regulates transiently expressed target genes in response to IR AA schematic illustration of p53's C-terminal modifications and described functional implications, including key regulatory enzymes.BTotal p53, p53 acetylated at K382 and K370 as well as GAPDH were measured by Western blot at indicated time points in the context of different p53 dynamics: pulsing p53 (10 Gy IR), transient p53 (10 Gy IR + BML-277, central lanes), and sustained p53 (10 Gy IR + Nutlin-3, right lanes). See Fig 3 and Materials and Methods section for details.CThe relative change in p53 acetylation at K370 (light green) and K382 (dark green) was quantified from Western blot and normalized to the abundance 3 h post-IR. Means and propagated standard errors from three independent experiments are indicated. Acetylation increased over time in the context of sustained p53. See also Appendix Fig S12.DThe p53-K370 methylase Smyd2 was down-regulated in a clonal stable A549 cell line expressing a corresponding shRNA. Transcript levels were measured in wild-type and knockdown cells by qRT–PCR. Mean levels and standard deviation from technical triplicates are indicated.E, FPromoter activity of CDKN1A (E) and MDM2 (F) was quantified in Smyd2 knockdown cells before (basal, gray) and 3 h (red), 6 h (blue), and 9 h (orange) after DNA damage (10 Gy IR). Left panel: The percentage of cells with active TSS, subdivided into populations with strong (> 75% of TSS, solid colors) and weak (< 75% of TSS, shaded colors) activity, is shown as stacked bar graphs; the mean fraction of active promoters is indicated above each bar. Right panel: Distributions of calculated transcription rates at active TSS are presented for each time point as probability density estimates (PDF, see Data visualization section). We measured a higher fraction of active promoters upon damage compared to A549 wild-type cells (Fig 3), while transcription rates remained unchanged. See Fig EV5A for corresponding changes in p53 acetylation patterns.GThe p53-K382 methylase Set8 was down-regulated in a clonal stable A549 cell line expressing a corresponding shRNA. Transcript levels were measured in wild-type and knockdown cells by qRT–PCR. Mean levels and standard deviation from technical triplicates are indicated.H, IPromoter activity of CDKN1A (E) and MDM2 (F) was quantified in Set8 knockdown cells before (basal, gray) and 3 h (red), 6 h (blue), and 9 h (orange) after DNA damage (10 Gy IR). Left panel: The percentage of cells with active TSS, subdivided into populations with strong (> 75% of TSS, solid colors) and weak (< 75% of TSS, shaded colors) activity, is shown as stacked bar graphs; the mean fraction of active promoters is indicated above each bar. Right panel: Distributions of calculated transcription rates at active TSS are presented for each time point as probability density estimates (PDF, see Data visualization section). We measured a higher fraction of active promoters upon damage compared to A549 wild-type cells (Fig. 3), while transcription rates remained unchanged. See Fig EV5A for corresponding changes in p53 acetylation patterns. Source data are available online for this figure.

Fig 5: SmFISH-based analysis at the first and second p53 pulse after IR reveals gene-specific stochastic expression patterns Schematic illustration of the life cycle of an mRNA and the rate constants that influence RNA abundance due to stochastic bursting according to previously published models of promoter activity. While burst frequency (bf) describes the switching of a promoter between a transcriptionally active and inactive state with the rate constants k on and k off, the burst size (bs) describes the number of RNAs transcribed in an active period. Additionally, degradation (d) further influences RNA levels by reducing the cytoplasmic RNA pool.Illustration of promoter activity according to the random telegraph model. An increase in RNA levels per cell can be due to a higher burst frequency (more active promoter periods, a higher rate of transcription initiation), or an increase in burst size (a higher rate of RNA transcription in an active period). Additionally, also mixtures of both scenarios are possible.We used smFISH data to calculated promoter activity based on previously published models. An overview of the calculations characterizing stochastic gene expression is shown. X RNA: number of quantified RNAs/cell, n: number of genomic loci, f: fraction of active promoters (proxy for burst frequency bf), µ: transcription rate per cell [RNA/h] (proxy for burst size bs), dRNA: RNA degradation rate per cell [1/h], M: polymerase occupancy [RNAs/h], v: RNAP2 speed (estimated as 3 kb/min), l: gene length, TSS: active TSS at the moment of measurement. Further details can be found in Materials and Methods section.Quantification of stochastic gene expression for the indicated p53 target genes before (basal, gray) and 3 h (red), 6 h (blue), and 9 h (orange) after DNA damage (10 Gy IR). The fraction (f) of active promoters (proxy for burst frequency) increases, while the transcription rate (µ; proxy for burst size) at active TSS remains similar upon DNA damage for all time points. Left panel: The percentage of cells with active TSS is shown as stacked bar graphs. We subdivided the population in cells with strong TSS activity (> 75% of TSS active, solid colors) and those with partial TSS activity (at least one, but less than 75% of TSS active, shaded colors). The mean fraction of active promoters (ratio of all active TSS to the total number of genomic loci analyzed) is indicated above each bar. Right panel: Distributions of calculated transcription rates µ [RNAs/h] at active TSS are presented for each time point as probability density estimates (PDF, see Data Visualization section). The number of TSS analyzed is indicated in each plot (compare Fig EV2C).Mean degradation rates of indicated RNAs in transcriptionally active cells before (basal, gray) and 3 h (red), 6 h (blue), and 9 h (orange) after DNA damage (10 Gy IR) as calculated from smFISH data. RNA stability is not changing in the measured time frame upon DNA damage. The plot displays the average RNA degradation rate per cell [1/h] over time after DNA damage, calculated from model (C) in actively transcribing cells for each gene.Based on promoter activity, we allocated target gene promoters along three archetypical expression patterns illustrated by a schematic triangle.Amount of p53 bound to indicated target gene promoters before (basal, gray) and 3 h (red), 6 h (blue), and 9 h (orange) after DNA damage (10 Gy IR) as measured by ChIP. The amount of bound p53 was calculated as percentage of input and normalized to the time point of the first p53 peak at 3 h. Individual data points (mean values of triplicate quantification in qRT–PCR measurements) from 3 to 4 biological repeats are shown as dots; mean values are displayed as black horizontal lines. Dashed lines serve as guide to the eyes. We could not detect p53 binding above IgG controls at the published p53 response element in the PPM1D promoter (indicated by n.d.) Source data are available online for this figure.