σB-dependent mRNA isoforms have elevated TE. (A) Estimation of the isoform-specific TE for the short, σB-dependent and long, σA-dependent isoforms of ctc and yvrE. Each point is an experimental condition which has a different short isoform fraction and correspondingly different apparent TE (conditions shown in Fig. 2 are distinctly marked by a triangle and a square for σB inactive and active, respectively). Error bars correspond to standard deviations from subsampling bootstraps. The gray lines are linear regressions, whereas the dashed lines indicate estimates of isoform-specific TE calculated from the fits (Materials and Methods). Estimated isoform-specific TEs and errors (standard deviations) from a bootstrapped linear fit (Materials and Methods) are shown. (B) Distribution (beeswarm and boxplot, whiskers corresponding to 10th and 90th percentile) of apparent TE in σB inactive conditions. Translationally activated σB regulon genes (subset from Fig. 1 for which isoform-specific TE could be estimated, Materials and Methods) are marked (red). (C) Isoform-specific TE values inferred, with error bars as in A. (D) Fluorescent reporter assay for validating differential TE between isoforms. Protein expression (from fluorescence) and mRNA levels (from RT-qPCR) were measured for synthetic constructs (left) representing σA-dependent (L) and σB-dependent (S) isoforms. Relative (to S reporters) isoform-specific TE (right) was calculated by dividing relative protein expression by relative mRNA levels. Errors bars represent the standard deviation for technical replicates (n = 3 for fluorescence, n = 4 for RT-qPCR). See also Supplemental Figure S3.
