Fig. 1
From: Roquin targets mRNAs in a 3′-UTR-specific manner by different modes of regulation
Roquin PAR-CLIP identifies a linear binding element. a Number of Roquin-bound clusters per nucleotide identified within the 5′-UTR, CDS and 3′-UTR regions of the target mRNAs. Each bar represents a region covering one-tenth of the respective mRNA region. b Multiple sequence alignments along with the consensus sequence-structure RNA motif are shown for the two most over-represented motifs, both characterized by a stem–loop (SL) secondary structure with a U-rich loop, derived from the top 100 Roquin-bound clusters. The numbers within parentheses represent the fraction of Roquin-bound clusters with similar structural motif. The different colors provide information about the number of distinct base pairs occurring (up to six types: C:G, G:C, A:U, U:A, G:U, and U:G) whereas the shading indicates how many sequences/structures in the alignment do not form a particular base pair. c Consensus sequence generated by MEME/PhyloGibbs analysis of all the 974 Roquin-bound cluster sequences. d Density of T to C mutations in the regions surrounding the linear binding element (LBE) revealed an enrichment of mutations in the flanking regions of the Py–Pu–Py core. Solid line and shadow represent the mean ± SEM. e Imino-1H-1D spectra of the Ox40 ADE-like and CDE-like SLs compared to the LBE (UUUUAACAUUAUUUU) of the 3′-UTR of Nfkbiz in the absence or presence of ROQ or zinc finger (Znf) domains of Roquin-1 and measured at RT or at temperatures as indicated. f Overlay of 1H-15N HSQC spectra of the Roquin ROQ domain (aa171–326) either without (black) or with different stoichiometric amounts of the 15-mer LBE RNA from the Nfkbiz 3′-UTR (in color). g ITC isotherm characterizing the binding of the ROQ domain of Roquin to the LBE RNA. The data were fitted to a one-site binding model, Kd and stoichiometry are indicated. h The same as in (g) but with a mutated form of LBE (UUUUAAGCGUAUUUU). i Percentage of Roquin clusters containing the different sequence and structural motifs as compared to those found in the same Roquin-binding sequences shuffled (background model). j Frequency of mRNAs containing different number of Roquin-binding sites. Data are representative of three (a–d, i, j) and four (g, h) independent experiments
