Fig. 3: HDLBP specificity for membrane-bound mRNAs.
a Frequency of top ten HDLBP crosslinked seven-mers located either in 3′ UTR or CDS of membrane-bound and cytosolic mRNAs. b Sequence logo of top five HDLBP crosslinked seven-mers ranked according to their frequency among all detected crosslinked seven-mers. c Distribution of z-scores calculated from differences in the frequency of all possible k-mers within membrane-bound and cytosolic CDS and 3′ UTR sequences. For each k-mer length, this analysis was performed for the group of top 40 HDLBP crosslinked k-mers (left) and all other k-mers (right). d p values of pairwise Wilcoxon rank-sum test between z-scores obtained for top 40 bound HDLBP k-mers as described in (c). e HDLBP multivalency analysis in +40/−40 nt regions around cross-linking sites. To evaluate the T-C binding affinity as a function of multivalency of HDLBP binding sites, we binned the multivalency scores for the top ten enriched four-mers within the 40-nt regions into five categories (group sized from highest to lowest score groups, n = 994, n = 973, n = 673, n = 1036, n = 1308). The total normalized T-C transition signal over the +40/−40 nt regions was then plotted for all five categories. Lower and upper hinges of box plots correspond to the 25th and 75th percentiles, respectively. Upper and lower whiskers extend from the hinge to the largest or smallest value no further than the 1.5× interquartile range from the hinge, respectively. Center lines of box plots depict the median values. f Analysis of the percentage of total T-C transitions for every nucleotide position within the +40/−40 nt region for each multivalency bin. The two bins with the highest multivalency scores are shown and correspond to (e). g Comparison of mean multivalency scores between differentially localized mRNAs in their CDS. A positive set (a four-mer group consisting of top ten HDLBP crosslinked four-mers, UUCU) and a negative set (AAGU) with no HDLBP enrichment. Occurrence of these four-mer groups were counted in 30-nt sliding windows and the mean score per transcript was computed. Mean distribution was then compared between different localized CDS by Wilcoxon rank-sum tests. h Apparent dissociation constants of recombinant GST-HDLBP fragments (constructs A though D) and full-length protein (FL), schematically shown, for different RNA oligonucleotides as determined by fluorescence anisotropy binding assays. Dissociation constants that were measured but could not be determined are indicated with “n.d.” and interaction that were not measured by “—“. i Fluorescence anisotropy binding assays for RNA oligonucleotides with different number of HDLBP binding four-mers (left, H40-44). GST-HDLBP construct B (KH5-9) was incubated with FAM-labeled RNA oligonucleotides, anisotropy measured and KD determined from the binding curves (middle panel). For each oligonucleotide, three independent Kd values were determined. Significant differences in Kd values were evaluated using two-sided t-test and are indicated with asterisks (*P < 0.01, *P < 0.05). Data were presented as mean values ± SD. a–i Source data are provided as a Source Data file.
