A single alteration 20 nt 5' to an editing target inhibits chloroplast RNA editing in vivo

Abstract

Transcripts of typical dicot plant plastid genes undergo C-->U RNA editing at approximately 30 locations, but there is no consensus sequence surrounding the C targets of editing. The cis-acting elements required for editing of the C located at tobacco rpoB editing site II were investigated by introducing translatable chimeric minigenes containing sequence -20 to +6 surrounding the C target of editing. When the -20 to +6 sequence specified by the homologous region present in the black pine chloroplast genome was incorporated, virtually no editing of the transcripts occurred in transgenic tobacco plastids. Nucleotides that differ between the black pine and tobacco sequence were tested for their role in C-->U editing by designing chimeric genes containing one or more of these divergent nucleotides. Surprisingly, the divergent nucleotide that had the strongest negative effect on editing of the minigene transcript was located -20 nt 5' to the C target of editing. Expression of transgene transcripts carrying the 27 nt sequence did not affect the editing extent of the endogenous rpoB transcripts, even though the chimeric transcripts were much more abundant than those of the endogenous gene. In plants carrying a 93 nt rpoB editing site sequence, transgene transcripts accumulated to a level three times greater than transgene transcripts in the plants carrying the 27 nt rpoB editing sites and resulted in editing of the endogenous transcripts from 100 to 50%. Both a lower affinity of the 27 nt site for a trans-acting factor and lower abundance of the transcript could explain why expression of minigene transcripts containing the 27 nt sequence did not affect endogenous editing.

Figure 1

Sequences surrounding rpoB editing site II. (A) Nucleotide sequence from tobacco and black pine located –20 to +6 relative to rpoB editing site II. The C→U editing site is indicated by an arrow. Differences between the two sequences are shown as bold in the black pine sequence. (B) Plasmid constructs containing test sequences were named as indicated. The bold nucleotides denote differences between the endogenous tobacco sequence (MR411) and the test sequence.

Figure 2

Analysis of DNA from transplastomic tobacco plants. (A) Schematic representation of the wild-type plastid genome and the inserted sequences in the transformed genome. –20/+6 sequence represents the region in which potentially edited test sequences were cloned. See Figure 1 for exact nucleotide sequence. (B) DNA was isolated from leaves of transformed and control plants (WT) and digested with BamHI. DNA gel blots were probed with 350 nt of plastid genome sequence surrounding the plasmid integration site. Sizes of λ DNA digested with HindIII are indicated on the left, in kb.

Figure 3

Restriction endonuclease assay to detect editing of the transgene transcripts. (A) Sau3AI (S) restriction map of the edited or non-edited RT–PCR products of the MR413 transgene with primers Trps16α3.1 and PPrrn2 (130 bp product). The size in bp is written under each restriction fragment. (B) NuSieve agarose gel (4%) of a Sau3AI restriction digest of 130 bp PCR product. PCR was performed on cloned and sequenced, edited (MR413T) or non-edited (MR413C) cDNA. Lanes 1–6 represent independent cloned RT–PCR products from a MR413 plant. Lane 5 corresponds to an edited RT–PCR product.

Figure 4

Poisoned primer extension assay to analyze endogenous rpoB transcript editing. (A) Portion of the sequence of the RT–PCR product (primers oNP91, oNP92, 510 bp amplicon) on which the poisoned primer extension with primer oNP90 (25 bp long) is performed. The primer extension is poisoned by addition of ddGTP. When the RNA is not edited, oNP90 is extended only to the first G (arrow); when the RNA is edited, the corresponding cDNA does not contain the first G, and oNP90 is extended to the second G (arrow). When extension is stopped at the first G, the length of the extended primer is 41 bases; if the extension goes to the second G, the length is 59 bases. (B) Autoradiograph of an acrylamide gel of primer extension products in which radiolabeled oNP90 was used with different RT–PCR templates. Lanes 1–6 consist of serial mixtures of PCR product (oNP91, oNP92) from a cloned edited and non-edited rpoB cDNA. Lane 1, 100 ng edited; lane 2, 80 ng edited plus 20 ng non-edited; lane 3, 60 ng edited plus 40 ng non-edited; lane 4, 40 ng edited plus 60 ng non-edited; lane 5, 20 ng edited plus 80 ng non-edited; and lane 6, 100 ng non-edited. Lane 7 is a control without template. In the next lanes, 80–100 ng of RT–PCR product (oNP91, oNP92) was used as template. The relative intensity of the 41 and 59 bases bands produced from total RNA isolated from transformed or control (WT) plants is an indication of the level of editing of the tobacco rpoB site II in the different plants tested.

Figure 5

Analysis of RNA from transplastomic plants. Total RNA isolated from leaves of controls (WT) and transformed plants was probed with an oligonucleotide antisense to the 27 nt editing site sequence in MR411. Approximately 10 µg of each RNA sample was run per lane. Endogenous rpoB bands are under exposed in order to visualize the transgene. MR210 is transformed with the 93 nt rpoB editing site; all other transformants carry the 27 nt editing site. Sizes of RNA markers (Promega) are indicated on the left, in kb.