p53-mediated adaptation to serine starvation is retained by a common tumour-derived mutant

Background In response to oncogenic stress, the tumour suppressor protein p53 can induce the elimination of cells through induction of cell death or senescence, helping to restrain malignant progression. Conversely, under nutrient stress, p53 can protect cells by supporting metabolic adaptation. Many cancers express mutant p53 proteins that have lost the cell-elimination properties of wild-type p53. However, a previous report showed that a tumour-derived mutant can retain the ability to support cells under glutamine starvation. Results We show that a commonly occurring p53 mutant, R248W, retains wild-type ability to support survival under serine starvation. R248W, but not R175H, can engage p21 and MDM2, which both function to limit oxidative stress and facilitate the switch to de novo serine synthesis. In vivo, the growth of R248W-expressing tumours is resistant to dietary depletion of serine and glycine, correlating with an increased capacity to limit ROS compared to tumours expressing R175H. Human cancers expressing this p53 mutant show a worse outcome. Conclusion Our work shows that mutant p53s can selectively retain wild-type p53 functions that allow adaptation to serine starvation through the activation of antioxidant defence pathways. Tumours containing this p53 mutation are resistant to serine-limited conditions and less responsive to therapy. Electronic supplementary material The online version of this article (10.1186/s40170-018-0191-6) contains supplementary material, which is available to authorized users.

Cell growth in Vogelstein HCT116 cells expressing WT p53, p53-null cells (KO), and the R248W/-p53 mutant cell line cultured in full media, full media containing the MDM2 inhibitor Nutlin-3A (10 µM), or serine and glycine-depleted media for 5 days. Total cell numbers were counted from triplicate wells per condition on days 2, 4, and 5 and tracked as cell growth relative to day zero counts. Data are represented as the mean of triplicate wells ± SEM for each condition. Representative experiment shown. Cell growth in HCT116 cells expressing WT, KO, R175H, R248W, or R273H p53 treated with non-targeting control (NT) or p21 siRNA and cultured in full media for 5 days. iRFP level per well relative to the day 2 reading shown for each cell line. Data are represented as the mean of triplicate wells ± SEM for each condition. Representative experiment shown. B.
Cell growth in HCT116 cells without p53 (KO), or expressing R175H, R248W, or R273H p53 treated with non-targeting control (NT) or MDM2 siRNA and cultured in full media for 5 days. iRFP level per well relative to the day 2 reading shown for each cell line. Data are represented as the mean of triplicate wells ± SEM for each condition. Representative experiment shown. C.
Cell growth in HCT116 cells with CRISPR KO p53 treated with either an HA-tagged MDM2 expression construct (MDM2) or a control vector (EV) and cultured in full media before being switched at day 1 to grow in serine and glycine-depleted media (or continued in full media). iRFP level in each well at end point in serine and glycine-depleted media (top) and iRFP intensity per well for all conditions at end point (bottom) shown for each cell line. Data are presented as mean ±SEM obtained from triplicate wells per condition. Representative experiment shown. D.
Cell growth in HCT116 cells with wild-type p53 treated with either an HA-tagged MDM2 expression construct (MDM2) or a control vector (EV) and cultured in full media before being switched at day 1 to grow in serine and glycine-depleted media (or continued in full media). iRFP level in each well at end point in serine and glycine-depleted media ( Circulating levels of serine or glycine in the serum of experimental animals at endpoint of xenograft experiment as determined by LC-MS analysis. Each data point represents the mean peak area of duplicate samples from each mouse on the indicated diet (control or -SG) and injected with indicated tumour cell line (HCT116 with R175H or R248W mutant p53). Black bars represent mean per sample group. Data were analysed using a 2-way ANOVA with Holm-Sidak's multiple comparisons test and multiplicity-adjusted p-values. B.
Tumour growth in HCT116 cells expressing either R175H or R248W injected SC into immunocompromised mice fed either control or -SG diet. Tumour growth was monitored by calliper measurement at regular intervals as shown. Data are from the same mice and tumours as depicted in figure 4A (where iRFP measurements are shown). Data presented as tumour volume relative to day 9 baseline reading (mean ±SEM) for each condition from N= 9 175 control, N=8 175 -SG, N=8 248 control, and N=10 248 -SG tumours. Data analysed using a 2-way ANOVA with Tukey's multiple comparisons test and multiplicity-adjusted p-values. P-value reported in figure from final measurements.