An N6-methyladenosine at the transited codon 273 of p53 pre-mRNA promotes the expression of R273H mutant protein and drug resistance of cancer cells

Abstract

Mutant p53 proteins that promote cancer cell invasive growth, metastasis and drug resistance emerge as therapeutic targets. Previously, we reported that suppression of ceramide glycosylation restored wild-type p53 protein and tumor suppressing function in cancer cells heterozygously carrying p53 R273H, a hot-spot missense mutation; however, the mechanisms underlying the control of mutant protein expression remain elusive. Herein, we report that an N⁶-methyladenosine (m⁶A) at the point-mutated codon 273 (G > A) of p53 pre-mRNA determines the mutant protein expression. Methylation of the transited adenosine was catalyzed by methyltransferase like 3 (METTL3), and this m⁶A-RNA promoted a preferential pre-mRNA splicing; consequently, the produced p53 R273H mutant protein resulted in acquired multidrug resistance in colon cancer cells. Furthermore, glycosphingolipids (particularly globotriaosylceramide) generated from serial ceramide glycosylation were seen to activate cSrc and β-catenin signaling so as to upregulate METTL3 expression, in turn promoting expression of p53 R273H mutant protein, with consequent drug resistance. Conversely, either silencing METTL3 expression by using small interfering RNA (siRNA) or inhibiting RNA methylation with neplanocin A suppressed m⁶A formation in p53 pre-mRNA, and substantially increased the level of phosphorylated p53 protein (Ser15) and its function in cells heterozygously carrying the R273H mutation, thereby re-sensitizing these cells to anticancer drugs. Concordantly, suppression of ceramide glycosylation repressed METTL3 expression and m⁶A formation in p53 pre-mRNA, thus sensitizing cells carrying R273H to anticancer drugs. This study uncovers a novel function of pre-mRNA m⁶A as a determinant of mutant protein expression in cancer cells heterozygously carrying the TP53 R273H mutation. Suppressing both RNA methylation and ceramide glycosylation might constitute an efficacious and specific approach for targeting TP53 missense mutations coding for a G > A transition, thereby improving cancer treatments.

Keywords: Drug resistance; Glucosylceramide synthase; Missense mutation; N(6)-methyladenosine; RNA methylation; Tumor suppressor p53.

Fig. 1.. m⁶A RNA methylation correlated with doxorubicin resistance in colon cancer cells harboring p53 R273H mutations.

A, R273H mutation in p53. Codon 273 of p53 RNA was CAU, rather CGU, in TP53-Dox cells. B, Cell responses to doxorubicin. After pretreatments with siMETTL3 or NPC for 3 days, SW48-Dox (wild-type p53) and TP53-Dox (p53/R273H⁺) cells were treated with doxorubicin (Dox) combined with these agents for 72 hr. siMETTL3, 100 nM siRNA against human METTL3; siRNA-SC, 100 nM siRNA scrambled control; NPC, 20 nM neplanocin A. *, p<0.001, compared to SW48-Dox cells; ** or #, p<0.001 compared to TP53-Dox cells treated with vehicle or siRNA-SC (not depicted in this graph). C, IC₅₀ values for Dox in cells treated with combinations of NPC, siMETTL3 or PDMP. *, p<0.001, compared to SW48-Dox cells; ** or #, p<0.001 compared to TP53-Dox cells treated with vehicle or siRNA-SC.

Fig. 2.. Inhibition of m⁶A methylation increased Dox-induced apoptosis in TP53-Dox cancer cells.

SW48-Dox and TP53-Dox cells were treated with neplanocin A (NPC, 20 nM) or vehicle for 3 days, and then exposed to Dox (100 nM, for 48 h) with these treatments to induce apoptosis. Imaging flow cytometry was accomplished with an ImageStream X Mark II system, and the data analyzed with IDEAS® Software. A, Apoptosis assay with imaging flow cytometry. BrdU, bromodeoxyuridine-FITC; BF, bright field; PI, propidium iodide. Top-panel, images of apoptotic and nonapoptotic cells of TP53-Dox cells identified using imaging flow cytometry for analysis. Apoptotic cells labelled with BrdU-FITC (green) are presented as percentages of total cells detected in each sample. B, NPC effects on apoptosis. **, p<0.001 compared to SW48-Dox cells; *, p<0.001 compared to cells treated with vehicle. C, Representative Western blot of PARP protein. After above treatments, equal amounts of detergent-soluble proteins extracted (50 μg/lane) were resolved using 4–20% gradient SDS-PAGE, and then immunoblotted with PARP antibody. Cleaved-P, cleaved PARP; GAPDH served as internal control. Protein levels (bottom panel) are represented as means of their optical densities normalized against GAPDH from three settings of blots. *, p<0.01 compared to corresponding cells treated with vehicle.

Fig. 3.. m⁶A methylation determined missense protein expression in cells carrying p53 R273H mutation.

A, Protein expression of p53 and p53 target genes in cells exposed to Dox. Cells were treated with NPC (20 nM), PDMP (5 μM) or vehicle for 5 days, with exposed to Dox (100 nM) during the last 48 h of treatments. Equal amounts of detergent-soluble proteins extracted (50 μg/lane) were resolved using 4–20% gradient SDS-PAGE, and then immunoblotted with corresponding antibodies. Top panel, representative Western blots. METTL3, methyltransferase like 3; pp53, phosphorylated p53 (Ser15). Protein levels (bottom panel) are represented as mean ± SD of their optical densities normalized against GAPDH from three settings of blots. *, p<0.001 compared to SW48-Dox cells; #, p<0.001 compared to TP53-Dox cells treated with vehicle. B, Protein Expression of p53 and p53 target genes in cells exposed to vehicle. Cells were treated with NPC (20 nM), PDMP (5 μM) or vehicle for 5 days, with exposed to vehicle during the last 48 h of treatments. C, Immunostaining for METTL3 and pp53. Scale bar represents 25 μM in photomicrographs (100× magnification). Red, METTL3-Qdot 605; green, pp53-Alexa Fluor 488. D, Western blotting of p53 and pp53 after immunoprecipitation (IP). Cells were treated with PDMP (5 μM, 6 days) or vehicle, with exposed to Dox (100 nM) during the last 48 h of treatments. Equal amounts of detergent-soluble proteins (500 μg) were precipitated with indicated antibody and then immunoblotted after non-denaturing 12% PAGE resolution. Relative levels of p53 or pp53 are represented as means of their optical densities normalized against SW48-Dox cells with vehicle. *, p<0.001 compared to SW48-Dox with vehicle; **, p<0.001 compared to TP53-Dox with vehicle. E, MS/MS proteomics analysis of p53. Equal amounts of proteins were precipitated with pp53 antibody, resolved and assessed via mass spectroscopic (MS/MS) analysis.

Fig. 3.. m⁶A methylation determined missense protein expression in cells carrying p53 R273H mutation.

A, Protein expression of p53 and p53 target genes in cells exposed to Dox. Cells were treated with NPC (20 nM), PDMP (5 μM) or vehicle for 5 days, with exposed to Dox (100 nM) during the last 48 h of treatments. Equal amounts of detergent-soluble proteins extracted (50 μg/lane) were resolved using 4–20% gradient SDS-PAGE, and then immunoblotted with corresponding antibodies. Top panel, representative Western blots. METTL3, methyltransferase like 3; pp53, phosphorylated p53 (Ser15). Protein levels (bottom panel) are represented as mean ± SD of their optical densities normalized against GAPDH from three settings of blots. *, p<0.001 compared to SW48-Dox cells; #, p<0.001 compared to TP53-Dox cells treated with vehicle. B, Protein Expression of p53 and p53 target genes in cells exposed to vehicle. Cells were treated with NPC (20 nM), PDMP (5 μM) or vehicle for 5 days, with exposed to vehicle during the last 48 h of treatments. C, Immunostaining for METTL3 and pp53. Scale bar represents 25 μM in photomicrographs (100× magnification). Red, METTL3-Qdot 605; green, pp53-Alexa Fluor 488. D, Western blotting of p53 and pp53 after immunoprecipitation (IP). Cells were treated with PDMP (5 μM, 6 days) or vehicle, with exposed to Dox (100 nM) during the last 48 h of treatments. Equal amounts of detergent-soluble proteins (500 μg) were precipitated with indicated antibody and then immunoblotted after non-denaturing 12% PAGE resolution. Relative levels of p53 or pp53 are represented as means of their optical densities normalized against SW48-Dox cells with vehicle. *, p<0.001 compared to SW48-Dox with vehicle; **, p<0.001 compared to TP53-Dox with vehicle. E, MS/MS proteomics analysis of p53. Equal amounts of proteins were precipitated with pp53 antibody, resolved and assessed via mass spectroscopic (MS/MS) analysis.

Fig. 4.. m⁶A methylation in the mRNA of p53 R273H.

A, Silencing METTL3 enhanced pp53 levels in TP53-Dox cells. TP53-Dox cells were treated with siRNA against METTL3 (100 nM, 6 days) and PDMP (5 μM, 6 days). Equal amounts of detergent-soluble proteins extracted (50 μg/lane) were resolved using 4–20% gradient SDS-PAGE, and then immunoblotted with corresponding antibodies. pp53, phosphorylated p53 (Ser15); METTL3, methyltransferase-like 3. Protein levels are represented below bands, as their OD values normalized against GAPDH from three settings of blots and compared to vehicle treatment. *, p<0.001 compared to vehicle. B, Methylation to m⁶A catalyzed by METTL3. C, RNA m⁶A methylation in cells. Equal amounts of total RNA extracted from cells of SW48-Dox (wt p53), TP53-Dox (TP53; R273H/⁺), WiDr (R273H⁺/⁺) and TP53-Dox treated with neplanocin A (NPC, 20 nM, 6 days) or PDMP (5 μM, 6 days) were assessed with MeRIP RT-qPCR. *, p<0.001, compared to SW48-Dox cells; **, p<0.001, compared to vehicle control in TP53-Dox cells. D, Real-time PCR chart, as described above.

Fig. 5.. Globotriaosylceramide modulates METTL3 and m⁶A methylation on p53 R273H mutant.

A, Representative Western blot. TP53-Dox cells were treated with PDMP (5 μM), FB1 (fumonisin B1, 25 μM), siGb3S (siRNA against Gb3 synthase, 100 nM), STxB (Shiga toxin 1 B subunit, 100 nM), PP2 (Src kinase inhibitor, 500 nM) or FH535 (β-catenin/Tcf inhibitor, 5 μM) for 6 days, with exposure to doxorubicin (100 nM) during the last 48 h of treatments. Equal amounts of detergent-soluble proteins (50 μg/lane) were resolved and then immunoblotted with corresponding antibodies. B, Relative protein levels in TP53-Dox cells with treatments. Protein levels are represented as mean ± SD of their OD values normalized against GAPDH from three settings of blots. *, p<0.001 compared to vehicle. #, p<0.001 compared to PDMP treatment. C, IC₅₀ values for Dox in cells after treatments. TP53-Dox cells were pretreated with PDMP (5 μM), FB1 (25 μM), siGb3S (100 nM), STxB (100 nM), PP2 (10 μM) or FH535 (5 μM) for 48 h, and then co-treated with these agents and various concentrations of Dox for an additional 72 h. *, p<0.001 compared to vehicle control. #, p<0.001 compared to PDMP treatments.