A fractionation-based approach to determine transcriptome-wide poly(A)-tail lengths.

<p>(A) Bulk poly(A)-tail analysis of the input and the unbound, short-tail and long-tail fractions from wild-type seedlings. RNA was 3’ labeled with ³²P-cordycepin and yeast poly(A) polymerase, before digesting all non-poly(A) RNA using RNAse A and RNAse T1, which specifically hydrolyze RNA at C and U residues (RNAse A) and at G residues (RNAse T1). The remaining ³²P-labeled poly(A) tails were separated by urea-PAGE before Phosphorimager analysis. Fragment sizes in nucleotides are indicated on the left based on a labeled RNA-size marker. Values indicated on top of the lanes are ratios of radioactive signal above 50 nucleotides to radioactive signal below 50 nucleotides in the respective lane. A replicate experiment is shown in <a href="http://www.plosgenetics.org/article/info:doi/10.1371/journal.pgen.1005474#pgen.1005474.s001" target="_blank">S1A Fig</a>. (B) Abundance estimates of the three spike-in control RNAs from RNA-seq of the short- and long-tail fractions in L<i>er</i> wild-type and <i>paps1-1</i> mutants. Box-plots show the distribution of values from four biological replicates each. Filled black dots indicate the median values. (C) Correlation of poly(A)-tail length estimates of wild-type plants obtained here with those from a previous study based on PAL-seq [<a href="http://www.plosgenetics.org/article/info:doi/10.1371/journal.pgen.1005474#pgen.1005474.ref017" target="_blank">17</a>].