Figure 4

eIF4A regulates the TE of a discrete mRNA subset with common features. (a) 5′ UTR sequences from Silvestrol-responsive genes were analyzed for secondary structure, variant 5′ UTRs and 5′ TOP sequences. Complex 5′ UTRs are defined as those with significant secondary structure (ΔG < -104 kcal/mol) or annotated 5′ UTR variants. Genes with decreased TE: 37.3% structured, 38.2% variant, 2.3% 5′ TOP or 5′ TOP-like, 24.7% non-TOP and non-complex. Insensitive TE genes: 8.3% structured, 0.3% variant, 3.7% 5′ TOP or 5′ TOP-like, 88% non-TOP and non-complex. Genes with increased TE: 10.6% structured, 0% variant, 2.1% 5′ TOP or 5′ TOP-like, 87.2% non-TOP and non-complex. (b,c) PC-3 (b) or MDA-MB-231 (c) cells were treated with increasing concentrations of Silvestrol in combination with INK128 at fixed doses: 15.6 nM, 31.25 nM and 62.5 nM. Cell viability was measured by CellTiter-Glo after 3 days. (d) Distribution of changes in TE upon INK128 treatment of MDA-MB-231 cells. INK128-sensitive genes with decreased TE (z-score < -1.5) are indicated. Population mean indicated by a solid vertical line; dotted vertical lines indicate σ values above and below the mean. (e) z-scores for INK128-dependent changes in TE for known 5′ TOP mRNAs. The dotted line is drawn at z-score = -1.5. (f) 5′ UTR characteristics for genes with INK128-dependent reduction in TE; 51.2% 5′ TOP or 5′ TOP-like, 37.3% variant, 7.7% structured and 6.6% non-TOP and non-complex. (g) z-scores of transcripts in INK128-treated (left panel) or Silvestrol-treated (right panel) MDA-MB-231 cells. Each point represents a single gene; genes in orange had decreased TE in INK128 while those in blue had decreased TE in Silvestrol. Dotted lines are drawn at z-score = -1.5. (h) A model proposing that the subset of mRNAs most sensitive to eIF4A inhibition is distinct from the pool of mRNAs regulated via mTOR-mediated assembly of eIF4E and eIF4G.