Aspirin cooperates with p300 to activate the acetylation of H3K9 and promote FasL-mediated apoptosis of cancer stem-like cells in colorectal cancer

Abstract

Cancer stem-like cells (CSCs) have been proposed as a key driving force of tumor growth and relapse in colorectal cancer (CRC), and therefore, they are promising targets for cancer therapy. Epidemiological evidence has suggested that the daily use of aspirin reduces overall mortality of CRC and the risk of distant metastasis. We investigated the effect and mechanism of aspirin on CSCs in CRC. Methods: The ratio of CSCs was analyzed after aspirin treatment both in a cell model and patient samples. Chemically modified aspirin and immunoprecipitation were adopted to detect the target proteins of aspirin. A locus-specific light-inducible epigenetic modification system based on CRISPR technology was constructed to verify the causal relationship in these molecular events. In vivo characterization was performed in a xenograft model. Results: We found that aspirin induces apoptosis in enriched colorectal CSCs, inhibits tumor progression, and enhances the anti-neoplastic effects of chemotherapeutic agents. Furthermore, aspirin directly interacts with p300 in the nucleus, promotes H3K9 acetylation, activates FasL expression, and induces apoptosis in colorectal CSCs. Notably, these effects of aspirin are absent in non-CSCs since H3K9 is hypermethylated in non-CSCs and the effects are not induced by other NSAIDs. In addition, aspirin can suppress oxaliplatin-enriched CSCs and serve as an adjuvant therapy. Conclusions: Taken together, we revealed a unique epigenetic and cox-independent pathway (p300-AcH3K9-FasL axis) by which aspirin eliminates colorectal CSCs. These findings establish an innovative framework of the therapeutic significance of aspirin.

Keywords: Aspirin; FasL; apoptosis; cancer stem-like cells; colorectal cancer.

Figure 1

Aspirin treatment eliminates colorectal CSCs. (A-B) Representative images (left) and quantification (right) of immunostaining assays used to detect the percentage of colorectal CSCs (ALDH1+, DLCK1+) in tumor cells in the paraffin specimens of patients. Scale bar indicates 100 μm. CRC: colorectal cancer. The results are presented as the mean ± SEM, n=3. *, p<0.01, unpaired t-test. (C-D) Tumorsphere-forming assay. Representative images (E) and quantification (F) of TCs (tumor spheres) formed from three colorectal cancer cell lines (HT29, P1, or P2) following 14-day treatment with aspirin (Asp) at the indicated concentrations. Tumor spheres formed from the adherent cells (ACs) of each cell line were also included. The results are presented as the mean ± SEM, n=3. *, p<0.01, one-way ANOVA. (E) The activity of ALDH1 analyzed using the ALDEFLUOR assay; diethylaminobenzaldehyde (DEAB) is a negative control, which was used to inhibit the reaction of ALDH with the ALDEFLUOR reagent. ALDH1-positive cells and their average percentages in the total cell counts are indicated in each panel (mean ± SEM, n=3). (F) Quantitative PCR (qPCR) results from HT29 and P1 ACs, TCs, or TCs with 2-day Asp treatment (5 mM) (mean ±SEM, n=3, normalized to Gapdh mRNA expression). *, p<0.01, one-way ANOVA compared to TCs without Asp treatment. (G) Tumor growth curve from xenograft assays following subcutaneous injection of 1×10⁵ TCs of HT29 cells into 6-week-old female nude mice (mean ± SEM, n=6). In vivo effects of Asp analyzed by i.p. injection of 20 mg/kg Asp following the seeding of TCs. *, p<0.01, two-way ANOVA compared to untreated TCs at the indicated time points. (See also Figure S1, Table S1 and Table S2)

Figure 2

Aspirin selectively induces apoptosis in CSCs. (A) Proliferation analysis of the survival of cells following 24 and 48 h exposure to Asp at the indicated concentrations. The results of Asp treatment of TCs from HT29 and P1 cell lines are presented (mean ±SEM, n=3). *, p<0.01, one-way ANOVA. (B) ALDH1 expression analyzed via FCM assays of ACs from HT29 and P1 cell lines, following 2-day treatment with Asp at the indicated concentrations. ALDH1 immunoreactive cells and their average percentages of the total cell counts are indicated in each panel (mean ± SEM, n=3). (C-D) TUNEL assay. Representative images (C) and quantification (D) of apoptotic cells of TCs from HT29 or P1 cell lines following 2-day treatment with Asp at the indicated concentrations. Scale bar in (C) indicates 50 μm. (D) The results are presented as the mean ± SEM, n=3. *, p<0.01, one-way ANOVA. (E) Cell cycle distributions analyzed via FCM assays of HT29 ACs or TCs following treatment with 5 mM Asp at the indicated time points. Sub-G1 was designated the apoptotic proportion and is listed in each panel (mean ± SEM, n=3). (See also Figure S2)

Figure 3

Aspirin-induced apoptosis is mediated by FasL. (A-B) qPCR results of HT29 and P1 TCs (A) or ACs (B) following 2-day treatment with 5 mM Asp (mean ±SEM, n=3, normalized to Gapdh mRNA expression). *, p<0.01, one-way ANOVA. The control is set to 1 for every gene. (C-D) FasL and Fas expression in HT29 TCs (C) or ACs (D) analyzed via FCM assays. Immunoreactive cells and their average percentages of the total cell counts are indicated in the histogram and quantified (mean ± SEM, n=3). *, p<0.01, unpaired t-test. (E) Proliferation analysis of the survival of HT29 and P1 TCs following 48 h exposure to 5 mM Asp in the presence of an isotype-matched control antibody or FasL-blocking antibody (Nok-1) at the indicated concentrations (mean ±SEM, n=3). *, p<0.01, one-way ANOVA. (F) Representative results of HT29 TCs analyzed via FCM assays following 48 h exposure to 5 mM Asp with a control antibody or 2 μg/mL Nok-1. Sub-G1 was designated as the apoptotic proportion. Scale bar indicates 50 μm. (G-J) TUNEL assay (G, I) and FCM assay (H, J) for analyzing apoptotic cells and FasL expression, respectively, from HT29 TCs in the presence of Asp (5 mM), Asp (5 mM) and PGE2 (1 μM) co-treatment, and siRNA against COX-2 or NS-398 (75 μM). The results are presented as the mean ± SEM, n=3. *, p<0.01, one-way ANOVA. (See also Figure S3)

Figure 4

Aspirin promotes the acetylation of histone H3. (A) FCM assay of FasL expression in HT29 TCs in the presence of Asp (5 mM), salicylic acid (5 mM), ibuprofen (1 mM), sulindac (100 μM) or indomethacin (600 μM). The results are presented as the mean ± SEM, n=3. *, p<0.01, one-way ANOVA. (B) Chemical structure of aspirin. The red square indicates its acetyl moiety. (C-F) Expression of FasL in HT29 TCs (C-D) or ACs (E-F) in the presence of Asp (5 mM) or TSA (1 μM) analyzed via FCM assays. FasL immunoreactive cells and their average percentage of the total cell counts are indicated (C, E) and quantified (D, F) (mean ± SEM, n=3). *, p<0.01, one-way ANOVA. (G) Immunoblotting analysis with anti-acetylated antibodies against H2AK5, H2BK5, H3K9, or H4K8 of HT29 or P1 TCs after 48 h Asp treatment. (H) In vitro ChIP assay using antibodies against Ac-H3K9 followed by qPCR to detect a fragment on the promoter of the FasL gene of HT29 TCs or ACs treated with 5 mM Asp (mean ± SEM, n=3, normalized to the expression level of 10% input). *, p<0.01, unpaired t-tests. (I) Immunoblotting against Mono-Me-H3K9 of ACs or TCs from HT29 and P1 cell lines. Immunoblotting against Ac-H3K9. (J) FCM assay of FasL expression in HT29 ACs in the presence of Asp (5 mM), 5-AZA (5 μM), or both Asp (5 mM) and 5-AZA (5 μM). The average percentages of FasL immunoreactive cells in the total cell counts were quantified (mean ± SEM, n=3). *, p<0.01, one-way ANOVA. (See also Figure S4)

Figure 5

Locus-specific analysis of aspirin-induced acetylation of histone H3K9. (A) Chemical structures of Biotin-Asp and Biotin-NH₂. (B) Proliferation analysis of the survival of HT29 or P1 TCs following 48 h exposure to 5 mM Asp, Biotin-NH₂, or Biotin-Asp (mean ±SEM, n=3). *, p<0.01, one-way ANOVA. (C) Immunocytochemical fluorescence staining with antibodies against biotin in HT29 TCs treated with 5 mM Biotin-NH₂ or Biotin-Asp for 48 h. Scale bar indicates 10 μm. (D) In vitro IP assay. After 48 h treatment with 5 mM Biotin-NH₂ or Biotin-Asp, HT29 TCs lysates were immunoprecipitated with anti-Biotin antibodies and immunoblotted with antibodies against Ac-H3K9 or GAPDH. (E) Schematics of locus-specific histone modification based on the CRISPR/dCas9 system (upper) and the associated gRNA-SID4X-dCas9 vector design (lower). (F) qPCR results of FasL mRNA from HT29 TCs transfected with the indicated plasmids (mean ±SEM, n=3, normalized to Gapdh mRNA expression) following 5 mM Asp treatment. *, p<0.01, one-way ANOVA. (G) Schematics of light-inducible locus-specific histone modification (upper) and the associated vectors (lower). (H) qPCR results of FasL mRNA from HT29 TCs transfected with the indicated plasmids (mean ± SEM, n=3, normalized to Gapdh mRNA expression); 5 mM Asp and 5 mW/cm² 466 nm blue light were applied as indicated. *, p<0.01, unpaired t-test. (See also Figure S5)

Figure 6

p300 is required for aspirin-induced acetylation of histone H3. (A-D) FCM assay of FasL expression in HT29 TCs treated with 5 mM Asp in the presence of siRNA against p300 or its inhibitor C646 (10 μM) (A), or siRNAs against CBP (B), PCAF (C), or GCN5L2 (D) (mean ± SEM, n=3). *, p<0.01, one-way ANOVA. (E) In vitro ChIP assay with antibodies against p300 followed by qPCR to detect fragments on the promoter of the FasL gene in HT29 TCs treated with 5 mM Asp, siRNA against p300, or C646 (mean ± SEM, n=3, normalized to the expression level of 10% input). *, p<0.01, one-way ANOVA. (F) In vitro IP assay. After 48 h treatment with 5 mM Biotin-NH₂ or Biotin-Asp, HT29 TC lysates were immunoprecipitated with anti-Biotin antibodies and immunoblotted with antibodies against p300 or GAPDH. (G) Diagram summarizing the difference between aspirin-induced effects in CSCs (upper) and non-CSCs (lower). (See also Figure S6)

Figure 7

Aspirin suppresses oxaliplatin-enriched CSCs. (A-B) Proliferation analysis of the survival of cells following 48 h exposure to Oxa at the indicated concentrations. The results of ACs, TCs, or TCs with Asp and Oxa co-treatment (5 mM) from the HT29 (A) or P1 (B) cell lines are presented (mean ±SEM, n=3). (C) Proliferation analysis of the survived cells following 48 h treatment with saline (Control group), Oxa (20 μg/mL), Asp (5 mM), or both Oxa and Asp. ACs from both HT29 and P1 cell lines were used. The data are presented as the mean ±SEM. * p<0.01, one-way ANOVA. (D) Tumor growth curve from xenograft assays following subcutaneous injection of 1×10⁵ HT29 TCs into 6-week-old female nude mice. The data are presented as the mean ± SEM and n=6 for each group. In vivo effects of Asp were analyzed by i.p. injection of both Asp (daily, 20 mg/kg body weight) and Oxa (once a week, 2 mg/kg/body weight) 10 days after seeding of TCs. The data are presented as the mean ± SEM. *, p<0.01 two-way ANOVA. (E) Representative images of TUNEL assays for detecting apoptotic cells from tumor tissues derived from HT29 TCs after Oxa treatment or both Oxa and Asp co-treatment. Scale bar indicates 20 μm. (F) Representative results of H&E staining and immunohistochemistry against ALDH1, FasL and ac-H3K9 on xenograft tumor sections from HT29 TCs mice treated with Oxa (once a week, 2 mg/kg/body weight) or Oxa combined with Asp (20 mg/kg body weight). Scale bar indicates 50 μm. (G) Quantification of immunohistochemistry on FasL or Ac-H3K9 positive cells in the tumors. Five visual fields were randomly selected and 100 cells in each visual field were counted. Data are shown as the percentage of the total number of positive cells. Values are presented as means ± SD, n=4. *P<0.05.