p53-mediated senescence impairs the apoptotic response to chemotherapy and clinical outcome in breast cancer

Abstract

Studies on the role of TP53 mutation in breast cancer response to chemotherapy are conflicting. Here, we show that, contrary to dogma, MMTV-Wnt1 mammary tumors with mutant p53 exhibited a superior clinical response compared to tumors with wild-type p53. Doxorubicin-treated p53 mutant tumors failed to arrest proliferation, leading to abnormal mitoses and cell death, whereas p53 wild-type tumors arrested, avoiding mitotic catastrophe. Senescent tumor cells persisted, secreting senescence-associated cytokines exhibiting autocrine/paracrine activity and mitogenic potential. Wild-type p53 still mediated arrest and inhibited drug response even in the context of heterozygous p53 point mutations or absence of p21. Thus, we show that wild-type p53 activity hinders chemotherapy response and demonstrate the need to reassess the paradigm for p53 in cancer therapy.

Figure 1

Superior response of p53 mutant MMTV-Wnt1 mammary tumors to doxorubicin treatment. (A–E) MMTV-Wnt1 mice bearing spontaneous, measurable, growing mammary tumors of approximately 500mm³ were injected once daily with 4mg/kg doxorubicin for T 5 consecutive days as indicated by arrowheads in graphs. Tumors with p53 wild-type (p53^+/+) (A), heterozygous mutant that lost the wild-type allele (p53^R172H/0) (B), or retained the wild-type allele (p53^R172H/+) (C), and homozygous mutant (p53^−/−) (D) were measured regularly and tumor volume calculated. Shown are 2 representative mice for each genotype. LOH was determined by sequencing DNA from tumors across the region of the knockin point mutation. The double peak observed in (C) shows p53 heterozygous status, while the single peak observed in (B) indicates LOH. (E) Summary of responses for MMTV-Wnt1 tumors with wild-type p53 or p53^R172H/+ that retained (n=22 and n=15, respectively, for tumor volume, and n=21 and n=15 respectively, for relapse) (“p53 wild-type”) versus tumors homozygous mutant or p53^R172H/+ that underwent LOH (n=2 and n=8, respectively, for tumor volume, and n=1 and n=4, respectively, for relapse) (“p53 mutant”). Tumors showing primary resistance (~20%) were not included in analysis of response. Time to relapse is the number of days post-treatment by which tumor exceeded maximum tumor volume. Values shown are +/− SEM. (F) To measure p53 activity, tumors were harvested 24 hr after the 5^th and final doxorubicin injection (Rx), or from untreated control mice (“C”) in p53 wild-type (WT) or homozygous mutant (Mut) backgrounds. Relative mRNA levels determined by reverse transcriptase real time PCR are shown for indicated genes, with mean of WT C set to 1. Horizontal line is the mean. (G) Tumors from p53^R172H/+ mice either retaining (Ret) or having lost (LOH) the 1 wild-type allele were harvested 24 hr after the final doxorubicin treatment (Rx), or from untreated control mice (“C”) and mRNA levels for indicated genes determined as in (F). Horizontal line is the mean. See also Figure S1.

Figure 2

Failure to arrest leads to aberrant mitoses and cell death in doxorubicin treated p53 mutant MMTV-Wnt1 tumors (A) p53 homozygous mutant and p53 wild-type MMTV-Wnt1 spontaneous tumors were harvested 24 hr following final treatment as in Figure 1F, and formalin fixed, paraffin embedded sections were stained for Ki67. Shown are representative images of 5 tumors from each group. Scale bar: 50μM. (B) An MMTV-Wnt1p53 wild-type tumor and mutant (p53^R172H/0) tumor were harvested and processed to single cell suspension by mincing and trypsinizing, and then 4×10⁶ cells were injected into the abdominal mammary fat pads of recipient female C57BL6 mice. Tumors that formed were left untreated or given the usual 5 treatments of doxorubicin and then 24 and 44 hr later, mice were injected with BrdU, and then harvested 4 hr later (48 hr after final doxorubicin treatment). Formalin fixed, paraffin embedded sections were stained for BrdU. Scale: 50μM. (C–D) H&E stained sections from transplanted tumors in (B) were examined for anaphases and scored for bridges. Shown are 3 representative images from the p53 wild-type tumor (C) and mutant (p53^R172H/0) tumor (D). Scale: 10μM. Bridges are indicated by yellow arrows. (E) Quantitation of anaphase bridges. “Clear” indicates no bridge observed. (F) p53 wild-type and p53 homozygous mutant MMTV-Wnt1 spontaneous tumors were harvested 24 hr following final treatment as in (A), and formalin fixed, paraffin embedded sections were stained for cleaved caspase 3 antibody or TUNEL as indicated in the figure. At least 4 random 200× high powered fields per tumor were counted, averaged and plotted, with mean +/− SEM shown as line with error bars. The one p53 wild-type tumor with a very high number of TUNEL positive cells was also the only p53 wild-type tumor that regressed during the treatment period (data not shown).

Figure 3

Wild-type p53 mediates senescence following doxorubicin treatment 1 of MMTV-Wnt1 tumors (A–C) Spontaneous tumors were harvested from untreated mice or from mice 5 to 7 days following final doxorubicin treatment and mRNA levels for senescence genes indicated in the figure were determined for p53 wild-type MMTV-Wnt1 tumors (p53^+/+) (A), p53 heterozygous mutant tumors that retained the wild-type allele (p53^R172H/+) (B) or lost the wild-type allele (p53^R172H/0) (C). p16 is shown on a separate axis due to the aberrantly high values in some tumors. All graphs (A–C) are relative to the mean of untreated p53 wild-type tumors set to a value of 1. For p53 mutant tumors, responding tumors harvested 5 days after treatment (5 day Rx) (the time point when p53 wild-type tumors express senescence markers) were used. ** p<0.005; * p<0.05 by student t-test, p values of non-significant comparisons are shown above the gene symbol. Horizontal line is the mean. (D) p21 protein levels in 6 of the tumors from (A) compared to growing, relapsed tumors as determined by western blot, with Vinculin (Vinc) used as a control. Student t-test of densitometric analysis of the two groups, p=0.0045. (E–F) Parallel transplanted tumors as in 2B were untreated (“C”) or doxorubicin treated (Rx) and harvested 48 hr or 5 days after the final treatment, or followed until relapse (Rel). mRNA levels for senescence genes indicated in the figure were determined for p53 wild-type (E) and mutant (F) transplanted tumors, in at least 4 tumors for each treatment group. Values in mutant tumors (F) are relative to untreated tumors in (E), and shown on a smaller scale so differences can be discerned. *p<0.05 by ANOVA and Newman-Keuls post test for comparison to untreated. Error bars: +/−SEM. (G) SAβGal staining in histological sections of untreated and treated orthotopic tumor transplants. Shown are representative sections from one p53 wild-type donor and 2 different p53^R172H/+ donors, with and without doxorubicin treatment. Similar results 1 were observed in a total of 3/3 wild-type and 3/3 p53^R172H/+ transplanted tumors. Scale: 50μM. See also Figure S2.

Figure 4

Dichotomous responses in doxorubicin treated p21-null MMTV-Wnt1 mice: subsets of tumors retain arrest capability while others continue proliferation. (A) p21-null MMTV-Wnt1 mice bearing spontaneous, measurable, growing mammary tumors of approximately 500mm³ were injected once daily with 4mg/kg doxorubicin for 5 consecutive days as indicated by red arrowheads in graphs. Shown are representative charts from mice with poor (top) and favorable (bottom) responses, and mean tumor regression, +/− SEM. (B) Ki67 IHC staining of untreated (“C”) and treated (“Rx”) p21-null MMTV-Wnt1 tumors harvested 24 hr following the final treatment. Shown are representative images from 7 treated and 6 untreated or relapsed tumors. Scale: 50μM. (C–D) Two different p21-null MMTV-Wnt1 tumors were transplanted, treated in parallel and BrdU incorporation was determined as in Figure 2B. Shown are representative images of BrdU staining (scale: 50μM), the corresponding tumor volume charts for parallel transplanted tumors that were followed (blue arrow indicates the 48 hr time point when parallel transplants were harvested for analysis), representative anaphases (scale: 5μM) and a table quantitating the anaphase data. (E) H&E stained sections from untreated (“C”) and treated (“Rx:”) tumors of the indicated genotypes were quantitated for mitotic figures. Each data point represents the average count of mitotic figures in ten 400× high power field views. p21-null MMTV-Wnt1 treated tumors were separated into arrested (open circles) and mitotic (open triangles) groups. Mean (horizontal line) with error bars (SEM) are shown. See also Figure S3.

Figure 5

G2 arrest with downregulated G2-M regulators in non-proliferating p21-null MMTV-Wnt1 treated tumors. (A–B) p53 wild-type and p21-null MMTV-Wnt1 tumors were transplanted and treated in parallel as in Figure 2B and harvested. Tumors as indicated in the figure were processed to single cell suspension by mincing and trypsinizing, followed by fixation and staining for DNA content with propidium iodide. (C) mRNA levels for G2-M regulatory genes as indicated in the figure were determined for untreated tumors (“C”) or tumors harvested 48 hr after final treatment (Rx), at least 3 tumors were analyzed per treatment group. Error bars: +/−SEM. ** p<0.005; * p<0.05 by student t-test. p values of non-significant comparisons and comparisons where treated tumors were higher are shown above the gene symbol. See also Figure S4.

Figure 6

Senescence associated cytokines are expressed in treated MMTV-Wnt1 mammary tumors. (A–C) p53 wild-type tumors (A), p53 heterozygous mutant tumors that retained the wild-type allele (B) or lost the wild-type allele (C), corresponding to spontaneous tumors from Figure 3A–C, were harvested and mRNA levels for genes indicated in the figure were determined. All graphs (A–C) are relative to the mean of untreated p53 wild-type tumors set to a value of 1. ** p<0.005; * p<0.05 by student t-test. p values of non-significant comparisons are shown above the gene symbol. Horizontal line is the mean. (D–G) Phospho-Stat3 and Ki67 staining in treated, senescent tumors. Serial sections from formalin fixed paraffin embedded tumors from Figure 3B were stained for phosphoTyr705-Stat3 (antibody clone D3A7) (upper panels) or Ki67 (lower panels). (D–G) Representative examples of positive/negative or negative/positive phospho-Stat3 and Ki67 from treated tumors. (F) Inverse staining of adjacent areas within the same tumor sections, and an area of double positivity marked by red dashed line. (G) Similar staining results with a second polyclonal phosphoTyr705-Stat3 antibody, with more intense phospho-Stat3 positive/Ki67 negative areas (indicated by black arrows) and less intense phospho-Stat3 staining/Ki67 positive areas (indicated by red arrows). Scale: 50μM (H) Tumor cells isolated from an MMTV-Wnt1 mammary tumor and cultured were plated at 8000 cells per well in a 24 well plate overnight, then media changed to serum free media (SFM), or 100ng/ml cytokine. EGF and insulin, known mitogens in breast cancer, were at 20ng/ml and 10μg/ml, respectively. Cell number was determined at day 4 by MTT assay. Error bars: +/− SEM. For comparison of SFM to Eotaxin, * p<0.05 by student t-test. For comparison of SFM, Cxcl5 and Rantes, ANOVA and Newman-Keuls p<0.005. See also Figure S5.

Figure 7

p53 or p21 knockdown improves doxorubicin response in p53 wild-1 type MCF-7 breast cancer cells (A) MCF-7 human breast cancer cells untreated or treated with 200nM doxorubicin for 24 hr and harvested 8 days later were fixed and stained for SAβGal. Scale bar: 500μM. (B) mRNA levels for cytokines indicated in the figure were determined for MCF-7 cells that were serum starved for 72hr (Q), growing in Log phase (L) or treated with doxorubicin as in (A) and harvested 1 day or 8 days later. (C) MCF-7 cells untreated or 8 days following treatment as in (A) were harvested 48hr after a media change, and western blots were performed with indicated antibodies. (D–F) MCF-7 cells were transfected with non-targeting siRNA (Control) or siRNA targeting p53 or p21. After 24 hr, cells were treated with doxorubicin for 24 hr, or untreated as indicated. (D) Cells were harvested 24 hr after treatment for western blot with p53, p21 or actin antibodies. (E) Cells were pulsed for 1 hr with BrdU 24 hr after doxorubicin treatment, fixed, stained with anti-BrdU FITC, and sorted by flow cytometry. Percent BrdU positive cells are indicated in the figure. (F) Four days following treatment and media change, cells were photographed using bright field microscopy (scale bar, 500μM). (G) Cell number was determined by MTT assay 9 days following treatment. Error bars: +/−SEM. ** p<0.005 ANOVA and Newman-Keuls post test. (H) MCF-7 cells were untreated or treated with 200nM doxorubicin for 24 hr, then fixed and stained for α-tubulin and DAPI 72 hr later. Scale bars, white: 40μM; yellow: 10μM. Yellow arrows indicate micronuclei and white arrows anaphase bridges. See also Figure S6.