MEF9, an E-subclass pentatricopeptide repeat protein, is required for an RNA editing event in the nad7 transcript in mitochondria of Arabidopsis

Abstract

RNA editing in plants alters specific nucleotides from C to U in mRNAs in plastids and in mitochondria. I here characterize the nuclear gene MITOCHONDRIAL EDITING FACTOR9 (MEF9) that is required for RNA editing of the site nad7-200 in the nad7 mitochondrial mRNA in Arabidopsis (Arabidopsis thaliana). The MEF9 protein belongs to the E subfamily of pentatricopeptide repeat proteins and unlike the three previously identified mitochondrial editing factors MEF1 and MEF11 in Arabidopsis and OGR1 in rice (Oryza sativa) does not contain a DYW C-terminal domain. In addition, the E domain is incomplete, but seems to be functionally required, since one of the two independent EMS mutants encodes a MEF9 protein truncated by a stop codon at the beginning of the E domain. In both mutant plants premature stop codons in MEF9 inactivate RNA editing at site nad7-200. The homozygous mutant plants are viable and develop rather normally. The lack of RNA editing at site nad7-200 thus seems to be tolerated although this editing event is conserved in most plant species or the genomic sequence already codes for a T at this position, resulting in a generally conserved amino acid codon.

Figure 1.

RNA editing is not detectable at the nad7-200 editing site in mitochondria of mef9-1 and mef9-2 mutant plants. Comparison of the cDNA sequence analysis of two RNA editing sites (boxed) in the mitochondrial nad7 mRNA between wild-type Arabidopsis (wt) and the mef9-1 and mef9-2 mutant plants shows that both mutants have lost the ability of C to U editing at site nad7-200, but not at another site in the same mRNA, site nad7-137. This latter site is correctly edited in wild type and in both mutant plants changing Ser to Leu codons. The absence of RNA editing event nad7-200 results in incorporation of the genomically encoded Ser rather than the Phe specified by the edited codon. In the cDNA strand analyzed, the detected T nucleotide (red trace) corresponds to the edited U, the observed C (blue trace) is derived from an unedited C.

Figure 2.

Phenotype and pollen analysis of the mef9-1 and mef9-2 mutant plants. A, The growth habitus of the mef9-1 mutant plants is indistinguishable from the wild-type (wt) Col plant and shows flowers and beginning seed set after 6 weeks. B, Anther and pollen staining shows that the mef9-1 and mef9-2 mutant plants are not male sterile and differentiate viable pollen comparable to the wild-type plants. Staining was done according to Alexander (1969). [See online article for color version of this figure.]

Figure 3.

Schematic structure of the MEF9 PPR protein encoded by locus At1g62260. A, The mef9-1 and mef9-2 mutant plants each have a stop codon truncating the MEF9 protein in different positions (denoted by stars). Both are C to T transitions in CGA codons resulting in TGA translational stops. The amino acid sequences in the boxes marked by black rectangles show little structural similarity and are not predicted to form further S or P repeats. The MEF9 protein contains an E domain at the C terminus, but is lacking a DYW region. B, Amino acid alignment of the C-terminal regions of the mitochondrial editing factor MEF9 with two chloroplast editing factors of the E class, CRR4 and CRR21. MEF9 is the shortest PPR protein identified so far to be involved in specific RNA editing events in either mitochondria or plastids. In the mef9-2 mutant, the premature stop codon removes the entire E domain. Amino acids in inverse shading are identical between at least two of three aligned sequences, similar amino acids are underlayered with gray. C, The amino acid alignment of the S and P domains in MEF9 reveals the varying lengths of the so-called 35er repeats and shows that only few amino acids are actually shared by most of these repeat regions. Amino acids in inverse shading are identical between at least eight of the 12 aligned sequences, those highlighted against a gray background are similar in eight or more of the 12 repeats.

Figure 4.

The Col MEF9 gene sequence restores the ability for RNA editing in protoplasts of the EMS mutant lines mef9-1 and mef9-2. The cDNA sequence tracings compare the effects of the protoplast complementation assays of the mef9-1 and mef9-2 mutants with the Col MEF9 gene (35S:MEF9) on RNA editing at site nad7-200 (boxed nucleotide) at 48 h after transfection. The regained ability to edit this site is detected by the increased T-signal trace (red). For control, transfection with the MEF1 gene is assayed. This gene is involved in editing at other mitochondrial editing sites (Zehrmann et al., 2009) and does not complement the mef9-1 or mef9-2 mutant protoplasts. Untransfected protoplasts do not recover any editing at the site monitored. On the very right-hand side the cDNA analysis of a mef9-1 mutant stably transformed with the wild-type MEF9 gene shows that RNA editing at site nad7-200 is fully recovered in vivo and that no trace of an unedited C is detectable.

Figure 5.

Northern analysis of the nad7 transcript pattern in the mef9-1 and mef9-2 mutant plants. The top section shows the nad7 transcripts visualized by autoradiography of the radioactive nad7 probe in wild-type (wt) plants, in the mef9-1 and mef9-2 mutant plants, and in the mef9-1 mutant stably complemented with a wild-type MEF9 gene. Besides the mature mRNA of 1,612 nucleotides low-abundant precursor transcripts can be detected. The bottom section shows the large rRNA in a methylene blue stain of the membrane to confirm the intactness of the RNA in these preparations. Five micrograms of total RNA were loaded into each slot, the blot was exposed overnight. [See online article for color version of this figure.]

Figure 6.

Comparison of nucleotide identities in the cis-recognition sequence around the nad7-200 editing site with other editing sites. The presumed recognition sequence between nucleotides −30 and +5 of the MEF9 target site nad7-200 was used to scan all other RNA editing sites in the mitochondrial transcriptome of Arabidopsis for sites with shared nucleotide identities. Sites cox2-581 and nad2-842 share with 18 and 17 nucleotides, respectively (plus the edited C), the most nucleotides. Nevertheless, both of these sites are edited normally in the mef9-1 or mef9-2 mutants.

Figure 7.

Nucleotide and amino acid sequences at and around the nad7-200 editing site are conserved in flowering plants. The top part shows the nucleotide sequence alignment around the nad7-200 editing site in several plant species. The editing site here at nucleotide number 0 is conserved only between Arabidopsis and Brassica napus. The other plants compared encode a genomic T at this position. In tobacco, a new C residue one nucleotide downstream is now edited to the U present in the other monocot and dicot plants. In the bottom part, amino acids of the NAD7 proteins around the site of editing are aligned from several plant species. The nad7-200 editing event alters amino acid 67 in the protein, another editing event changes the codon, but not the identity of amino acid 70. Nucleotides and amino acids at RNA editing sites are highlighted in bold. The Ser residue encoded in the unedited mef9 mutant mRNAs is inversely shaded.