Increased short/long ratio (SLR) of AURKA APA isoforms in TNBC

(A) AURKA transcript isoforms (USCS Genome Browser). AURKA gene is located on (-) strand. (B) Median and range of SLR values for AURKA 3’UTR obtained using APADetect. Mann-Whitney test; ****, p<0.0001. (C),(D) RT-qPCR analysis of SLR of AURKA 3’UTR in TNBC cell lines (C) and patient samples (D). SDHA used as reference gene. TN, tissue number. (E) Relapse-free survival rates of TNBC patients with high (highest 25%) or low (lowest 25%) AURKA SLRs. p value determined by log-rank test.

AURKA shows 3’UTR isoform-dependent protein expression

(A) Top. UTR-dependent protein expression reporters. Venus CDS is flanked by AURKA 5’UTR and 3’UTR, WT or PAS-mutated. Bottom. Representative snapshots of transfected U2OS. Scale bar 50μm. (B) Mean and s.e.m. of median Venus/mCherry MFI ratios from transfected U2OS from three biological replicates. n ≥ 129 cells per condition. Ordinary one-way ANOVA with Tukey’s multiple comparisons test. (C) 3’RACE of endogenous AURKA APA isoforms. (D) RT-qPCR of endogenous AURKA SLR in U2OS. Long isoform abundance plotted as fold change over total AURKA mRNA. 18S rRNA used as reference target. (E) Same as (B) but in MCF10A (left) and RPE1 (right) cells. n ≥ 55 cells per condition. Unpaired t-test. ns, not significant; *, p<0.05; **, p<0.01; ***, p<0.001.

AURKA APA isoforms are translated with different efficiency

(A),(B) RT-qPCR of reporter mRNAs abundance (A) and decay rate (B) from transfected U2OS. mCherry mRNA used as reference target. Ordinary one-way ANOVA with Tukey’s multiple comparisons test; ns, not significant. (C) Decay rate of endogenous AURKA mRNA as in (B). 18S rRNA used as reference target. Abundance of long isoform plotted as fold change over total AURKA mRNA. (D) Design of the Nascent Chain Immunoprecipitation (NC IP) reporters and assay. (E) Mean and s.e.m. of median FlagVenus/mCherry MFI ratios from transfected U2OS from two biological replicates. n ≥ 160 cells per condition. Unpaired t-test; **, p<0.005. (F), (G) Immunoblots of NC IP fractions using Δ (F) or Flag-Δ (G) reporter. mCherry used as negative control. (H) RT-qPCR of eluted reporter mRNAs. Results representative of three biological replicates.

Translation rate of AURKA APA isoforms follows different cell cycle periodicity

(A) Design of the TRIPS reporters and assay (left) and CDK2 activity sensor (right). (B) Mean and s.e.m. of median sfGFP/mCherry MFI ratios from U2OS transfected with TRIPS-Δ and imaged at 0h and 2h of 50μM TMP treatment, with or without 0.1mg/ml CHX, from three biological replicates. Baseline at 0h. n ≥ 55 cells per condition. (C) RT-qPCR of sfGFP mRNA from U2OS transfected with TRIPS-Δ, at 0h and 2h of 50μM TMP treatment. mCherry mRNA used as reference target. RNA extracts at 0h of treatment used as reference sample. (D) TRIPS (left) and C-TRIPS (right) assays in transfected U2OSCDK2. n ≥ 200 cells per condition. Left. Mean and s.e.m. from three biological replicates. Right. Median and 95% CI of pooled data from left. Kruskal-Wallis with Dunnett’s multiple comparisons test. (E) RT-qPCR of endogenous AURKA mRNA in U2OS. 18S rRNA used as reference target. Baseline at G1/S. (F) Endogenous AURKA long isoform as in (E) plotted as percentage of total AURKA mRNA. (D) Left, (E), (F) Ordinary one-way ANOVA with Tukey’s multiple comparisons test. ns, not significant; *, p<0.05; **, p<0.01; ***, p<0.0005; ****, p<0.0001.

Translational periodicity of long 3’UTR isoform is regulated by hsa-let-7a miRNA

(A) Complementarity of hsa-let-7a binding to AURKA 3’UTR. (B) Mean and s.e.m. of median Venus/mCherry MFI ratios from U2OS co-transfected with 250nM hsa-let-7a or a negative control (NC) miRNA from three biological replicates. n ≥ 182 cells per condition. Unpaired t-test. (C) Same as (B) but co-transfecting 300nM anti-let-7a or NC. n ≥ 94 cells per condition. Unpaired t-test. (D) RT-qPCR of reporter mRNAs abundance from U2OS cells transfected as (B), at 8h of 10μg/ml ActD. mCherry mRNA used as reference target. Ordinary one-way ANOVA with Tukey’s multiple comparisons test. (E) TRIPS (left) and C-TRIPS (right) assays in transfected U2OSCDK2. n ≥ 162 cells per condition Left. Mean and s.e.m. from three biological replicates. Ordinary one-way ANOVA with Dunnett’s multiple comparisons test vs NC. Right. Median and 95% CI of pooled data from left. Kruskal-Wallis with Dunnett’s multiple comparisons test vs NC of the respective phase. ns, not significant; *, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001.

Increased AURKA short/long ratio is sufficient to disrupt cell behavior

(A) Design of CRISPR editing. Nucleotide substitutions in red. (B) 3’RACE of endogenous AURKA APA isoforms. (C),(D) Western blot after G1/S enrichment (C) and with or without 6h treatment with 0.1mg/ml CHX (D). Blots representative of three biological replicates. (E) CCK8 assay. A non-parental WT U2OS cell line used as negative control. Ordinary one-way ANOVA with Dunnett’s multiple comparisons test vs WT; **, p<0.005; ***, p<0.001. (F) Left. Representative images of cells grown in soft agar. Right. Mean number of clones and s.e.m. of three biological replicates. (G) Measurement of migration rate. (F), (G) Ordinary one-way ANOVA with Dunnett’s multiple comparisons test vs WT; *, p<0.05; **, p<0.01.

(A) R2 (coefficient of determination) values to indicate goodness of fit of a simple linear regression between mCherry and Venus single-cell MFI values. Data relative to Figure 2B. (B) RT-qPCR of mCherry and Venus mRNAs from three RNA extracts of U2OS cells transfected with Δ reporter. Ct values are mean between three technical replicates of the amplification reaction. (C) Quantification of mCherry and Venus MFI from U2OS cells transfected with the Δ reporter and imaged at the indicated time points. Mean and s.e.m. (n=8 cells) shown at each time point. (D) Single-cell Venus/mCherry MFI values from U2OSCDK2 cells transfected with Δ reporter grouped by intervals of CDK2 activity (Figure 4A, right) and plotted as median and 95% CI. n = 43 cells analysed. Kruskal-Wallis with Dunnett’s multiple comparisons test; ns, not significant. (E) Mean and s.e.m. of median Venus/mCherry MFI ratios from U2OS cells transfected with constructs containing long or short CDC6 3’UTR from three biological replicates. n ≥ 119 cells per condition. Unpaired t-test; ***, p=0.0005.

(A) Left. Representative U2OS cells transfected with TRIPS-S and imaged at 0h and 2h of 50μM TMP treatment. Right. Representative U2OSCDK2 cells. (B) TRIPS assay performed in U2OS cells transfected with TRIPS-Δ and treated with DMSO for 2h. Mean and s.e.m. of median sfGFP/mCherry ratios from two biological replicates, with baseline at 0h. n = 95 cells analysed. (C) Graph using data from Figure 4D showing MFI of mCherry and sfGFP at 0h and 2h of TMP treatment. Mean and s.e.m. of median MFI values from three biological replicates. (D) TRIPS assay performed in transfected U2OS cells and imaged at 0h and 2h of 50μM TMP treatment. Ratios between the median of single-cell sfGFP/mCherry ratios at 2h and that at 0h of TMP treatment were calculated for two biological replicates and are shown as mean and s.e.m.. n ≥ 343 cells analysed per condition. (B), (C), (D) Unpaired t-test. ns, not significant; *, p<0.05.

(A) Sanger sequencing validation of the mutated dPAS element in two cell lines. (B) Western blot showing similar AURKA protein expression in unsynchronized WT or U2OS cell lines that were not mutated following CRISPR editing (#1, #2).