A synthetic RNA editing factor edits its target site in chloroplasts and bacteria

Abstract

Members of the pentatricopeptide repeat (PPR) protein family act as specificity factors in C-to-U RNA editing. The expansion of the PPR superfamily in plants provides the sequence variation required for design of consensus-based RNA-binding proteins. We used this approach to design a synthetic RNA editing factor to target one of the sites in the Arabidopsis chloroplast transcriptome recognised by the natural editing factor CHLOROPLAST BIOGENESIS 19 (CLB19). We show that our synthetic editing factor specifically recognises the target sequence in in vitro binding assays. The designed factor is equally specific for the target rpoA site when expressed in chloroplasts and in the bacterium E. coli. This study serves as a successful pilot into the design and application of programmable RNA editing factors based on plant PPR proteins.

Fig. 1. Schematic representation of CLB19 and dsnPLS-DYW bound to their target sites.

The proteins are represented by ovals for each PPR motif including the 5th and last specificity-determining amino acids. The target sites are coloured according to the predicted favourability of the alignment at each position, from favoured (green) to neutral (white) to disfavoured (purple). These values are taken from ref. . The C at the editing site is shaded in blue.

Fig. 2. MORF-enhanced sequence-specific binding of Dsn3PLS proteins torpoA-789691target RNA.

Binding assays were performed with dsn3PLS (a, b) or dsn3PLS-DYW (c, d) in comparison to, or together with, MORF9 in the presence of 1 nM rpoA-78691 target RNA (a, c) or clpP1-69942 (b, d).

Fig. 3. Visible and molecular phenotypes ofclb19transformants.

a Seedlings were grown under a 16 h photoperiod with a light intensity of either 100 µmol m⁻² s⁻² (left panel) or 40 µmol m⁻² s⁻² (right panel). The construct used to transform the clb19-3 mutant line is indicated on the left. The scale bar is 1 cm. b Chloroplast transcript abundances are shown as log₂ fold-difference compared to their levels in clb19. The RNA was extracted from plants initially grown under low light. c RNA editing at the rpoA-78691 and clpP1-69942 sites in the four genotypes.

Fig. 4. Putative off-target editing by dsn3PLS-DYW and CLB19.

a Potential editing sites plotted by their predicted binding to the expressed editing factor (x-axis) against the ratio of the editing activity compared to a negative control (expressed as the log of the odds ratio). Predicted binding scores were calculated with PPRmatcher (https://github.com/ian-small/PPRmatcher). For dsn3PLS(E70A) the negative control was clb19, for dsn3PLS-DYW and CLB19, the negative control was dsn3PLS(E70A). All sites where the difference in editing is statistically significant (Fisher exact test corrected for multiple testing) and the log(odds ratio) is greater than 2 are highlighted in red (or orange for two potential ‘false positives’ discussed in the text). The ~32,000 sites below this threshold are indicated by the density contours (grey-black). b Alignment of the highlighted sites from a with the corresponding editing factor. The rpoA-78691 site is highlighted in red. The target sites are coloured according to the PPRmatcher score at each position, from favoured (green) to neutral (white) to disfavoured (purple). The percentage of edited transcripts at each site is indicated on the right. The sequence logos (constructed by Skylign) indicate the nucleotide biases at each position.

Fig. 5. RNA editing in bacteria.

A RNA editing quantified by RNAseq at the intended target site in three independent samples of bacteria expressing dsn3PLS-DYW, dsn3PLS-DYW together with MORF2 or dsn3PLS-DYW(E70A), together with MORF2. B Potential editing sites plotted by their predicted binding to dsn3PLS-DYW (x-axis) against the ratio of editing activity compared to the negative control (dsn3PLS-DYW(E70A) together with MORF2), expressed as the log of the odds ratio. Predicted binding scores were calculated with PPRmatcher (https://github.com/ian-small/PPRmatcher). The only significant site where the log(odds ratio) exceeds 2 is the intended target site, rpoA-78691, highlighted in red. The ~492,000 sites below this threshold are indicated by the density contours (grey-black). C Additional tests of editing in E. coli. From left to right: dsn3PLS-DYW(E70A) + MORF2 as a negative control; MORF2 and MORF9 enhance editing by dsn3PLS-DYW by similar amounts; a U at the rpoA-78691 editing position is not detectably reverse-edited to C; the clpP1-69942 site is not detectably edited with or without MORF2. All experiments were repeated in triplicate, as shown.