Fig. 6

DHX36 and DHX9 mediate translation of selected cancer genes. a Gene ontology classification for genes, which TE decreases (P < 0.05) upon depletion of DHX36 (red) and DHX9 (green). b Immunoblots of lysates from HeLa cells depleted in DHX36 and DHX9 and probed as indicated. Immunoblots were performed 96 h after siRNA transfection. c Diagram showing a DHX36 / DHX9-dependent mechanism of translational control. (1) Scanning 43S PICs that translate unstructured 5′-UTRs or rG4-containning 5′-UTRs, that are maintained in their unfolded state by the DHX36 and DHX9 helicases, initiate translation at the main ORF (mORF). (2) A fraction of scanning PICs may initiate translation at upstream start codons, present within 5′-UTR in a suboptimal context, affecting the efficiency of the mORF translation. (3) In the absence of the rG4 processing helicases, rG4 motifs folding may slow down PIC scanning, thereby providing more time for the recognition of the upstream start codon and stimulating the translation of the upstream open reading frame (uORF). (4) 80S ribosomes may either dissociate from the mRNA after termination or stall during elongation or termination by the uORF-encoded attenuator peptide, preventing the translation of the mORF. d Schematic of bicistronic reporter genes containing within their 5′-UTRs either an rG4 motif (∆ uORF + rG4), a mutated rG4 (∆ uORF + rG4 mut), an rG4-containing uORF (uORF + rG4) or an rG4-mutated uORF (uORF + rG4 mut). e Relative translation of the different expression vectors showing that an rG4 enhances the repressive effect of a short uORF. f Effect of DHX36 (red) and DHX9 (green) depletion on the relative translation of reporter genes containing within their 5′-UTRs either an rG4-containing uORF or an rG4-mutated uORF as compared to control (non-targeting siRNAs, gray). Data represent the mean and s.d., n = 3 biological replicates. P-values were calculated using an unpaired student’s t-test. ns non-significant, **P < 0.01, ***P < 0.001. Representative flow cytometry profiles are reported in Additional file 1: Figure S15