The MTL1 Pentatricopeptide Repeat Protein Is Required for Both … · The MTL1 Pentatricopeptide Repeat Protein Is Required for Both Translation and Splicing of the Mitochondrial

The MTL1 Pentatricopeptide Repeat Protein IsRequired for Both Translation and Splicing of theMitochondrial NADH DEHYDROGENASESUBUNIT7 mRNA in Arabidopsis1

Nawel Haïli, Noelya Planchard, Nadège Arnal2, Martine Quadrado, Nathalie Vrielynck, Jennifer Dahan3,Catherine Colas des Francs-Small, and Hakim Mireau*

Institut Jean-Pierre Bourgin, Institut National de la Recherche Agronomique, AgroParisTech, Centre Nationalde la Recherche Scientifique, Université Paris-Saclay, 78026 Versailles cedex, France (N.H., N.P., N.A., M.Q.,N.V., J.D., H.M.); Université Paris-Sud, Université Paris-Saclay, 91405 Orsay cedex, France (N.H., N.P.); andAustralian Research Council Centre of Excellence in Plant Energy Biology, University of Western Australia,Crawley, Western Australia 6009, Australia (C.C.d.F.-S.)

ORCID IDs: 0000-0002-6386-5672 (C.C.d.F.-S.); 0000-0002-2299-5139 (H.M.).

Mitochondrial translation involves a complex interplay of ancient bacteria-like features and host-derived functionalities.Although the basic components of the mitochondrial translation apparatus have been recognized, very few protein factorsaiding in recruiting ribosomes on mitochondria-encoded messenger RNA (mRNAs) have been identified in higher plants. In thisstudy, we describe the identification of the Arabidopsis (Arabidopsis thaliana) MITOCHONDRIAL TRANSLATION FACTOR1(MTL1) protein, a new member of the Pentatricopeptide Repeat family, and show that it is essential for the translation of themitochondrial NADH dehydrogenase subunit7 (nad7) mRNA. We demonstrate that mtl1mutant plants fail to accumulate the Nad7protein, even though the nad7 mature mRNA is produced and bears the same 59 and 39 extremities as in wild-type plants. Wenext observed that polysome association of nad7 mature mRNA is specifically disrupted in mtl1 mutants, indicating that theabsence of Nad7 results from a lack of translation of nad7 mRNA. These findings illustrate that mitochondrial translationrequires the intervention of gene-specific nucleus-encoded PPR trans-factors and that their action does not necessarily involvethe 59 processing of their target mRNA, as observed previously. Interestingly, a partial decrease in nad7 intron 2 splicing was alsodetected in mtl1 mutants, suggesting that MTL1 is also involved in group II intron splicing. However, this second functionappears to be less essential for nad7 expression than its role in translation. MTL1 will be instrumental to understand themultifunctionality of PPR proteins and the mechanisms governing mRNA translation and intron splicing in plant mitochondria.

Translation is the fundamental process decoding thegenetic message present on mRNAs into proteins. Inplant cells, mRNA translation occurs in the cytoplasm

but also in two organelles, mitochondria and plastids.Because of their prokaryotic origin, the translationmachineries operating in these two organelles sharemany characteristics with the bacterial translation ap-paratus (Bonen, 2004; Barkan, 2011). However, mostof these bacteria-like features have been modifiedthroughout evolution, and current organellar translationsystems cooperate with numerous nucleus-encodedeukaryotic trans-factors. The divergence from bacteriais particularly obvious in plant mitochondria, notablybecause mitochondrial mRNAs lack the typical Shineand Dalgarno (SD) motif in their 59 leaders and alter-native start codons other than AUG are often used toinitiate translation (Bonen, 2004). Proteomic and bio-informatic analyses allowed the identification of mostproteins and RNA factors forming the core of the plantmitochondrial translation machinery, including transla-tion initiation and elongation factors aswell as ribosomalproteins (Bonen, 2004; Bonen and Calixte, 2006). How-ever, the dynamics of this machinery remains largelyobscure. In particular, nothing is known about the re-cruitment of mitochondrial ribosomes on 59 untranslated

1 This work was supported by the LabEx Saclay Plant Sciences(grant no. ANR–10–LABX–0040–SPS).

2 Present address: Institut National de la Recherche Agronomique,Centre National de Ressources Génomiques Végétales, 31326 Casta-net Tolosan, France.

3 Present address: Department of Plant, Soil, and EntomologicalSciences, University of Idaho, Moscow, ID 83844.

* Address correspondence to [email protected] author responsible for distribution of materials integral to the

findings presented in this article in accordance with the policy de-scribed in the Instructions for Authors (www.plantphysiol.org) is:Hakim Mireau ([email protected]).

N.H. performed most of the experiments; N.P. performed part ofthe experiments; J.D. performed part of the experiments; C.C.d.F.-S.performed part of the experiments; N.A. provided technical assis-tance to N.H. and J.D.; M.Q. provided technical assistance to N.H.,J.D., and N.P.; H.M. conceived the project, supervised the experi-ments, and wrote the article with contributions of all the authors.

www.plantphysiol.org/cgi/doi/10.1104/pp.15.01591

354 Plant Physiology�, January 2016, Vol. 170, pp. 354–366, www.plantphysiol.org � 2016 American Society of Plant Biologists. All Rights Reserved. www.plantphysiol.orgon April 2, 2020 - Published by Downloaded from

Copyright © 2016 American Society of Plant Biologists. All rights reserved.

http://orcid.org/0000-0002-6386-5672

http://orcid.org/0000-0002-2299-5139

mailto:[email protected]

http://www.plantphysiol.org

mailto:[email protected]

http://www.plantphysiol.org/cgi/doi/10.1104/pp.15.01591


regions in the absence of the SD motif and about therecognition of the correct translation initiation codonby the small ribosomal subunit. The high degree of se-quence divergence among 59 leaders of mitochondrialgenes suggests a ribosome recruitment mechanism in-volving gene-specific cis-sequences and trans-factors(Hazle and Bonen, 2007; Choi et al., 2012). Up to now,only two proteins belonging to the PentatricopeptideRepeat (PPR) family have been found to promote mi-tochondrial translation in higher plants (Uyttewaalet al., 2008b; Manavski et al., 2012). How they facilitatetranslation is still unclear, as for the few characterizedPPR proteins shown to participate in plastid translation(Fisk et al., 1999; Schmitz-Linneweber et al., 2005; Caiet al., 2011; Zoschke et al., 2012, 2013). The plastidPENTATRICOPEPTIDE REPEAT PROTEIN10 (PPR10)protein of maize (Zea mays) is the only one for whichthe function has been elucidated at the molecular level.It was shown that, upon binding, PPR10 impedes theformation of a stem-loop structure in the 59 leader oftheATP synthase subunit c (atpH) mRNA, permitting therecruitment of ribosomes through the liberation of anSD motif (Prikryl et al., 2011).PPR proteins represent a large family of RNA-

binding proteins that has massively expanded in terres-trial plants (Barkan and Small, 2014). Most eukaryotesencode a handful of these proteins, whereas plant nu-clear genomes express over 400 PPR proteins that arealmost exclusively predicted to target mitochondriaand/or plastids (Lurin et al., 2004; O’Toole et al., 2008).This family of proteins is characterized by the succes-sion of tandem degenerate motifs of approximately 35amino acids (Small and Peeters, 2000; Lurin et al., 2004).Based on the length of these repeats, the PPR family hasbeen divided into two groups of roughly equal sizein higher plants. P-type PPR proteins contain onlysuccessions of canonical 35-amino acid repeats (P),whereas PLS PPR proteins are composed of sequentialrepeats of P, short (S), and long (L) PPR motifs. P-typePPR proteins were shown to participate in various as-pects of organellar RNA processing, whereas PLS PPRproteins have been almost exclusively associated withC-to-U RNA editing (for review, see Barkan and Small,2014; Hammani and Giegé, 2014). Recent crystal struc-tures showed that PPRmotifs adopt an antiparallel helix-turn-helix fold whose repetition forms a solenoid-likestructure (Ringel et al., 2011; Howard et al., 2012; Banet al., 2013; Yin et al., 2013; Coquille et al., 2014; Gullyet al., 2015). PPR tracks organize highly specific inter-action domains that were shown to associate withsingle-stranded RNAs (Schmitz-Linneweber et al., 2005;Beick et al., 2008; Uyttewaal et al., 2008a; Williams-Carrier et al., 2008; Pfalz et al., 2009; Cai et al., 2011;Hammani et al., 2011; Prikryl et al., 2011; Khrouchtchovaet al., 2012; Manavski et al., 2012; Zhelyazkova et al.,2012; Ke et al., 2013; Yin et al., 2013). The mechanism ofsequence-specific RNA recognition by PPR proteins wasrecently uncovered, and combinations involving aminoacid 6 of one motif and amino acid 1 of the subsequentmotif correlate strongly with the identity of the RNA

base to be bound (Barkan et al., 2012; Takenaka et al.,2013; Yagi et al., 2013).

Besides those involved in RNA editing, fewmitochondria-targeted PPR proteins have been char-acterized to date. Thus, our knowledge of the mecha-nisms governing the production and the expressionof mitochondrial RNAs in higher plants is very limited.In this analysis, we describe the function of a novelmitochondria-targeted PPR protein of Arabidopsis(Arabidopsis thaliana) called MITOCHONDRIALTRANSLATION FACTOR1 (MTL1). Genetic and bio-chemical analyses indicate thatMTL1 is essential for thetranslation of the mitochondrial NADH dehydrogenasesubunit7 (nad7) mRNA. Effectively, the Nad7 proteindoes not accumulate to detectable levels inmtl1mutants,and this absence correlates with a lack of associationof nad7 mature mRNA with mitochondrial polysomes.Interestingly, a partial but significant decrease in nad7intron 2 splicing was also detected in mtl1 mutants, sug-gesting that the MTL1 protein is also involved in group IIintron splicing. Since the decrease in splicing was onlypartial, this second function of MTL1 appears less es-sential for nad7 expression than its role in translation.

RESULTS

Arabidopsis mtl1 Mutants Display a Slow-Growth Phenotype

In an effort to better understand gene expressionin higher plant mitochondria, a series of Arabidopsismutants bearing transfer DNA (T-DNA) insertionspredicted to affect mitochondrially targeted P-type PPRproteins was collected. The search for interesting mu-tants revealed the mtl1-1 line for which homozygousmutant plants displayed significantly retarded growthon soil compared with the wild type (Fig. 1A). Theaffected PPR gene in this line corresponded to theAt5g64320 gene and encoded an 82-kD protein com-prising 16 PPR repeats according to predictions (Fig. 1B;Lurin et al., 2004). A second T-DNA insertion line af-fecting the same gene was subsequently identified. Thissecond allelic mutant, named mtl1-2, displayed thesame growth alterations as mtl1-1, strongly suggestingthat the developmental phenotype observed in theselines was effectively associated with inactivation of theAt5g64320 gene (Fig. 1, A and B). Reverse transcription(RT)-PCR analysis indicated that no detectable full-length mRNA derived from the At5g64320 gene accu-mulates in both mtl1-1 and mtl1-2 mutant plants,supporting that both identified mutant lines representednull mutants (Fig. 1C). mtl1 mutant plants showedvarious developmental abnormalities compared withwild-type plants. Both mutant lines grew rather slowlycompared with the wild type but reached about 80%of the size of Col-0 plants when cultured on soil for2.5 months (Supplemental Fig. S1A). Additionally,mtl1plants generally bear deformed and dark green rosetteleaves (Fig. 1A). They also needed nearly twice as muchtime to flower compared with the wild type, but they are

Plant Physiol. Vol. 170, 2016 355

A PPR Protein Aiding in Translation and Splicing

www.plantphysiol.orgon April 2, 2020 - Published by Downloaded from Copyright © 2016 American Society of Plant Biologists. All rights reserved.

http://www.plantphysiol.org/cgi/content/full/pp.15.01591/DC1


fertile (Supplemental Fig. S1B). Seeds produced by ho-mozygous mtl1 mutant plants were darker than normalArabidopsis seeds, but they germinate with around 80%efficiency on soil or in vitro (Supplemental Fig. S1C).

The Arabidopsis MTL1 Gene Encodes a Mitochondrion-Targeted PPR Protein

The cellular distribution of the MTL1 protein wasverified by expressing a GFP translational fusion com-prising the first 109 amino acids of MTL1 in transgenicArabidopsis plants. Roots of transformed plants wereobserved with a confocal microscope, and the GFPfluorescence appeared as a punctuated signal distrib-uted throughout the cytoplasm. The use of MitoTrackerRed indicated that these signals corresponded to mi-tochondria (Fig. 2A). To confirm these results and togain insight into the submitochondrial distribution ofMTL1, a C-terminal fusion comprising the full-lengthMTL1 protein and 10 copies of the MYC epitope tag

was expressed in Arabidopsis. This strategy was fa-vored, as rabbit polyclonal antibodies produced againstfragments of MTL1 did not succeed in detecting theprotein from Arabidopsis protein extracts. Expressionof the MTL1-MYC protein in the mtl1-1 mutant linerescued the mutant phenotypes, strongly supportingthe functionality of the fusion protein (SupplementalFig. S2, A and B). The distribution of the tagged MTL1was then analyzed by probing total, plastid, and mi-tochondrial proteins prepared from transgenic plantswith an anti-MYC monoclonal antibody. The MTL1-MYC fusion appeared to be highly enriched in themitochondrial protein preparation, as attested by theanti-Nad9 and anti-ATPC control antibodies (Fig. 2B).The submitochondrial localization of MTL1-MYC wasdetermined by probing soluble and membrane-boundmitochondrial proteins for the presence of MTL1-MYC.The tagged version of MTL1 was detected at similarlevels in both soluble and membrane-associated pro-tein preparations (Fig. 2C). The use of anti-Nad9 and

Figure 1. Arabidopsis mtl1 mutants are delayed intheir development. A, Comparative vegetative phe-notypes of mtl1 and Columbia-0 (Col-0) plants.Homozygous mtl1 mutants grow much slower thanwild-type plants on soil and produce plants withtwisted rosette leaves. The photograph was taken after10 weeks of culture in long-day conditions. B, Sche-matic diagram showing the predicted domain struc-ture of the Arabidopsis MTL1 protein. Thick arrowsalong the protein indicate PPR-P repeats, and theblack rectangle at the N terminus depicts the mito-chondrial targeting sequence. Thin arrows above andbelow the protein indicate the locations of T-DNAinsertions harbored in mtl1-1 and mtl1-2 mutants. C,RT-PCR analysis ofMTL1 transcripts in Col-0 andmtl1mutants. Five hundred nanograms of total RNA fromplants of the indicated genotypes was reverse tran-scribed with random primers. Resulting comple-mentary DNAs (cDNAs) were PCR amplified witholigonucleotides bracketing T-DNA insertion sites,and the resulting amplification products were sizefractionated by agarose gel electrophoresis. The BIO2cDNA was separately amplified to control the effi-ciency of the RT reaction. Reverse transcriptase waseither added (RT+) or omitted (RT2) from the initialRT reaction.

356 Plant Physiol. Vol. 170, 2016

Haïli et al.







anti-FDH control antibodies confirmed the proper sep-aration of both mitochondrial protein fractions.

mtl1 Mutants Are Impaired in the Synthesis of RespiratoryComplex I

The mitochondrial localization of MTL1 led us toconsider that a defective respiratory activity may be atthe origin of mtl1 mutant phenotypes. To clarify the ori-gin of a potential respiratory perturbation in these plants,we analyzed the steady-state levels of the different res-piratory chain complexes in comparison with the wildtype. Mitochondria prepared from mtl1 and wild-typeplants were lysed in the presence of digitonin, and thesolubilized respiratory complexes were separated onblue-native gels. Coomassie Blue staining showed thatmost respiratory complexes accumulated to similar levelsin wild-type and mutant plants, except for complex I andthe supercomplex I + III (Fig. 3A, left). An in-gel NADHdehydrogenase activity test confirmed the lack of bothforms of complex I in mtl1 mutants (Fig. 3A, right).To further analyze the impact of mtl1 mutations

on Arabidopsis respiration, oxygen uptake in the darkwas determined on detached leaves in both mutant

and wild-type plants. This led us to show that mtl1plants consumed nearly twice as much oxygen as Col-0plants (Supplemental Fig. S3). Since this increase ofoxygen consumption could be suggestive of a stronginduction of the alternative respiratory pathways inthe mutant plants, the expression levels of alternativeNADH dehydrogenases (NDA and NDB) and alter-native oxidase (AOX) genes were measured. Quanti-tative RT-PCR indicated that steady-state levels ofAOX1A, NDA1, and NDB4 transcripts overaccumulatefrom 4- to 8-fold in mtl1 plants (Fig. 3B). The strongincrease in AOX expression was further estimatedby immunoblot analysis of mitochondrial extracts withan antibody to AOX. The results showed a highly sig-nificant increase in the AOX signal in bothmtl1mutants(Fig. 3C). Our results indicated that mtl1 plants are com-plex I respiratory mutants and that the alternative res-piratory pathway is strongly activated in these plants.

mtl1 Mutants Do Not Accumulate the MitochondrialNad7 Protein

To better understand the lack of assembled complex Iin mtl1 plants, we next analyzed the expression of all

Figure 2. Subcellular and submitochondrial localization of the MTL1 protein. A, Confocal microscope images showing thecellular distribution of an MTL1-GFP fusion in Arabidopsis root cells. The MTL1-GFP fusion comprising the first 109 amino acidsof MTL1 and the GFP was stably transformed into Arabidopsis plants. Prior to observation, roots were briefly soaked in Mito-Tracker Red to label mitochondria. The green channel on the left shows the GFP fluorescence; the center shows mitochondrialabeledwith theMitoTracker Red dye; the right presents themerged signals. B, Immunoblot analysis of total (Tot), chloroplast (Cp),and mitochondrial (Mt) protein fractions prepared from Arabidopsis plants expressing an MTL1-MYC fusion. The protein prep-arations were probed with an anti-MYC (a MYC) monoclonal antibody to verify the mitochondrial localization of MTL1. Thepurity of the different fractions was tested with antibodies directed against the mitochondrial Nad9 (a Nad9) and the chloroplastATPC (aATPC) proteins. C, Assessment of proteins present in total mitochondrial protein preparation (TotM) and in soluble (Sol) aswell as membrane-derived (Mb) protein fractions prepared from plants expressing an MTL1-MYC fusion. Protein gel blots wereprobed with an anti-MYC monoclonal antibody to detect the MTL1-MYC fusion, anti-Nad9 antibodies (corresponding to anextrinsic protein of the inner membrane of mitochondria), and anti-formate dehydrogenase (a FDH) antibodies (a known solublematrix protein). Each lane was loaded with about 50 mg of protein preparation.






mitochondria-encoded complex I subunits in the mu-tants. In a first approach, steady-state levels and pro-cessing efficiency of mitochondria-encoded complex ImRNAs were analyzed by RNA gel-blot analysis inboth mtl1 and wild-type plants. This approach showed

that all nine mitochondrial mRNAs encoding complex Isubunits accumulate to detectable levels in mtl1 plants(Fig. 4A; Supplemental Fig. S4). Most of the corre-sponding mature transcripts are produced in amountsclose to wild-type levels, with a slight increase notably

Figure 3. mtl1 mutants are complex Irespiratory mutants. A, Blue Native gelanalysis of mitochondrial respiratorycomplexes extracted fromwild-type andmtl1mutant plants. The gel presented onthe left was stained with CoomassieBlue, whereas the gel on the right wasstained to reveal the NADH dehydro-genase activity of complex I. About200 mg of protein treated with digitoninwas loaded per lane. I, III, and V indicatethe positions of the respective respira-tory chain complexes. I+III correspondsto the I + III supercomplex. The locationsof native molecular markers are indi-cated. NBT, Nitroblue tetrazolium. B,Quantitative RT-PCR measuring the rel-ative steady-state levels of NDA, NDB,and AOX transcripts inmtl1-1 andmtl1-2mutants. C, Analysis of the accumulationlevel of the mitochondrial AOX in bothwild-type and mtl1 plants. Total mito-chondrial protein preparations obtainedfrom plants of the indicated genotypeswere probed with anti-AOX antibodies(a AOX). Fifty micrograms of proteinwas loaded in each lane. CoomassieBlue staining of part of the gel is shownbelow the blots to attest for equivalentloading of the different samples.


Haïli et al.




for nad1, nad4, nad5, and nad6 mRNAs. The only ex-ceptions are the nad7 and nad9 mature transcripts,which are much less abundant inmtl1 plants comparedwith Col-0 (Fig. 4A). In the case of nad7, this decreasewas associated with the overaccumulation of anunspliced mRNA precursor, indicating that the loss ofMTL1 caused, directly or indirectly, a splicing defect

of nad7 mRNA (Fig. 4A). Quantitative RT-PCR analy-sis measuring the splicing efficiency of all intron-containing mitochondrial transcripts allowed us todetermine that this splicing defect concerns the secondintron of nad7 (Supplemental Fig. S5). A moderate2-fold decrease of nad2 intron 1 and 2 splicing efficiencywas also detected inmtl1mutants with this analysis. Toanalyze the processing of nad transcripts in more detail,the 59 and 39 extremities of all mitochondria-encodednadmRNAs were amplified by circular RT-PCR in bothwild-type and mtl1 plants. No obvious processing dif-ferences for any of these mRNAs could be identified inthe mutants. Cloning and sequencing of the obtainedamplification products further confirmed that all com-plex I mature transcripts in mtl1 plants bear 59 and 39extremities that are identical to the ones found in thewild type, including nad7 (Supplemental Fig. S6).Shorter circular RT-PCR products for nad7 were moreapparent in mtl1 plants compared with the wild type.Cloning and sequencing of these fragments in bothCol-0 and mtl1 mutants indicated that they derivedfrom 59 truncated nad7 transcripts that appear to accu-mulate to slightly higher levels in the mutants. Theirabundance is insufficient to be visible on RNA gel blots(Fig. 4A). We next reasoned that the partial destabili-zation of nad7 and nad9 mature transcripts could resultfrom difficulties in expressing the Nad7 and/or Nad9proteins. Consequently, mitochondrial protein extractswere prepared from both mutants andwild-type plantsand probed with anti-Nad7 and anti-Nad9 antibodies.Whereas the Nad9 protein was still detectable, al-though significantly reduced, no trace of Nad7 could befound in either mtl1 mutant line (Fig. 4B). This showedthat mtl1 mutants did not accumulate detectable levelsof Nad7 protein and that the lack of this mitochondrialprotein was very likely responsible for the differentmolecular and physiological perturbations observed inmtl1 plants.

The nad7 mRNA Does Not Associate with TranslatingMitochondrial Polysomes in mtl1 Mutants

We next wondered whether the lack of Nad7 proteinin mtl1 plants could simply result from insufficient ac-cumulation of the nad7 mature mRNA. To answer thisquestion, we used the Arabidopsis bso-insensitive roots6(bir6) pprmutant that was also described to accumulatelow levels of mature nad7 transcript in response to diffi-culties in splicing the first nad7 intron (SupplementalFig. S7A; Koprivova et al., 2010). Comparative RNAgel-blot analyses suggested that bir6 plants accumulateless nad7 mature transcript compared with mtl1 mu-tants (Supplemental Fig. S7B). Indeed, quantitative RT-PCR further showed that bir6 plants contain only 6%of wild-type levels of mature nad7 transcripts, whereasmtl1-1 plants contain about 20% (Supplemental Fig.S7C). Probing ofmitochondrial proteinswith anti-Nad7and anti-Nad9 antibodies showed that bir6 plantsaccumulate slightly reduced levels of the Nad9 protein

Figure 4. The Nad7mitochondrial protein does not accumulate inmtl1mutants. A, Analysis of nad7 and nad9 mRNA abundance in wild-typeand mtl1 plants. Fifteen micrograms of total RNAwas size fractionatedon denaturing agarose gels, blotted, and analyzed by hybridization toradiolabeled DNA probes corresponding to nad7 and nad9 mRNA.Signals corresponding to the expected mature transcripts are indicatedby M and precursors by P. Asterisks indicate unexplained hybridizationsignals that may not be directly related to the nad9mitochondrial locus.Ethidium bromide staining of ribosomal RNAs is shown below the blotsand serves as a loading control. B, Detection of both Nad7 and Nad9proteins in mitochondrial extracts of mtl1 mutants. Mitochondrialproteins prepared from plants of the indicated genotypes were analyzedby immunoblot assay. The blots were probed with antisera to Nad7 (aNad7) and Nad9 (a Nad9) mitochondrial complex I subunits and toPORIN (a PORIN) used as a loading control.












compared with the wild type and, unlike mtl1 mu-tants, very low but still detectable levels of Nad7(Supplemental Fig. S7D). Therefore, the comparisonbetween mtl1 and bir6 mutants supported that the lossof Nad7 inmtl1 plants was not directly linked to the lowlevels of nad7 mature transcript but likely from molec-ular events involved directly in the production of theNad7 protein. However, the lack of antibody againstthe othermitochondria-encoded complex I subunits didnot allow us to rule out the possibility that other Nadproteins were also missing in mtl1 mutants.

Therefore, the translation status of all mitochondria-encoded nad transcripts was analyzed by polysomesedimentation analysis in both wild-type and mtl1-1plants. Polysomes were isolated from Arabidopsis in-florescences and fractionated on continuous sucrosedensity gradients. Ten fractions were collected along

the gradients after centrifugation. Polysome integritywas shown by the distribution of ribosomal RNAs alongthe gradients in the presence of MgCl2 (SupplementalFig. S8A). The disruption of polysomes by the additionof EDTA in the gradients indicated that polysomal RNAmigrated toward the center and the bottom of thegradients, whereas free mRNAs accumulated in theupper fractions (Supplemental Fig. S8A). RNA wasextracted from each fraction and subjected to RNA gel-blot analysis using probes specifically recognizing eachmitochondria-encoded complex I transcript (Fig. 5). Thehybridized membranes were then exposed to a phos-phorimager screen, and the relative distribution ofeach hybridization signal was determined for eachmature mRNA (Supplemental Fig. S9). Most mRNAswere distributed similarly all along the gradients inCol-0 and mtl1 extracts, indicating that free and

Figure 5. The nad7 mature mRNA does notsediment with mitochondrial polysomes inthe mtl1-1 mutant. Flower bud RNA extractsprepared from Col-0 and mtl1-1 plants werefractionated in 15% to 55% sucrose densitygradients by ultracentrifugation and under con-ditions maintaining polysome integrity. Ninefractions of equal volume were collected(fractions 2–10, as fraction 1, devoid of RNA,was discarded). Equal aliquots of each fractionwere analyzed by RNA gel-blot analysis andhybridized to DNA probes detecting each oneof the mitochondria-encoded complex I genes(nad1–nad9). A representative blot is shown foreach mRNA and each genotype. M indicatesthe band corresponding to themature transcriptsize for each gene.


Haïli et al.








polysome-associated transcripts exist for all complex Itranscripts. In contrast, the distribution of nad7 mRNAwas drastically different between mtl1-1 and wild-typeplants. In the former, most of the nad7 signal was con-centrated in the top four fractions, suggesting that noribosomes are loaded on nad7 mature transcripts in theabsence of a functional copy of MTL1. No such differ-ence was found for the other complex I transcripts,supporting a specific role of MTL1 in the translationof nad7 mRNA. We could conclude from these resultsthat nad7 mRNA is not properly translated in the ab-sence of MTL1 and that, subsequently, Nad7 is indeednot produced in mtl1 mutants.To further support the role of MTL1 in mitochondrial

translation, we analyzed the distribution of the MTL1-MYC protein fusion on a polysome gradient preparedfrom mtl1-complemented plants (Supplemental Fig.S2A). Total proteins were prepared from the 10 col-lected fractions, and fractions 2 to 10 were subjectedto immunoblot analysis using anti-MYC antibodies.The result indicates that the MTL1-MYC fusion isdetected all along the gradient and does cosedimentwith heavy polysome fractions on sucrose densitygradients (Supplemental Fig. S8B).

DISCUSSION

MTL1 Is a P-Type PPR Protein Essential for nad7 MaturemRNA Translation

PPR proteins have been implicated in multiple as-pects of organelle gene expression, but very few of themhave been specifically associated with the translation ofmRNAs. In this study, we convincingly showed thatthe mitochondria-targeted MTL1 protein of Arabidopsisis required for the translation initiation of the nad7mRNA. To reach this conclusion, we observed thatthe Nad7 protein did not accumulate in mtl1 mutantsand that this deficiency likely resulted from the lack ofsynthesis of Nad7, as nad7 mature mRNA does notcosediment with translating mitochondrial polysomeson sucrose density gradients. We also showed that theabsence of Nad7 synthesis does not result from an in-correct processing of nad7 transcripts, as nad7 maturemRNAs bear the same 59 and 39 extremities in mtl1 andwild-type plants. Furthermore, comparison with an-other Arabidopsis ppr mutant (bir6; Koprivova et al.,2010) allowed us to show that the lack of Nad7 proteinaccumulation in mtl1 plants did not result from an in-sufficient accumulation of nad7mature mRNA. Finally,MTL1 was found to cosediment with translating poly-some fractions on sucrose density gradients. We alsoobserved a significant reduction of nad9 expression inmtl1plants. A role ofMTL1 in nad9 translation is unlikely,as nad9 transcripts still cosediment with mitochondrialpolysomes inmtl1-1 plants. It is more likely that the effectobserved on nad9 expression is secondary to the loss ofrespiratory complex I, as observed inmany other complexI-deficient mutants (Gutierres et al., 1997; Keren et al.,2009; Kühn et al., 2011; Hsu et al., 2014; Hsieh et al., 2015).

Very little is known about the molecular mechanismsgoverning the expression of mitochondrial genes inplants. Translation is by far the least understood RNAprocessing step in plant mitochondria. Due to its origin,the mitochondrial translation machinery exhibits cer-tain bacteria-type characteristics (Bonen, 2004). How-ever, in plants, the vast majority of mitochondrialmRNAs lack an SD motif upstream of their translationinitiation codon, certainly because the mitochondrial18S ribosomal RNAdoes not contain the correspondinganti-SD sequence (Hazle and Bonen, 2007). In bacteria,the ribosomal S1 protein was shown to promote ribo-some recruitment for mRNAs bearing leaders devoid ofthe SD motif (Sørensen et al., 1998). Such a mechanismis unlikely to permit ribosome association to mito-chondrial mRNAs in plants, as the S1 counterpart lacksthe necessary RNA-binding domain (Hazle and Bonen,2007). To compensate for the absence of a general ri-bosome recruitment system on mitochondrial mRNAsin plants, translation initiation likely relies on gene-specific trans-factors to allow for ribosome anchoringand guidance toward the correct AUG start codon.Besides MTL1, very few other mitochondrial transla-tion factors were identified in higher plants so far. Thefirst one was the Arabidopsis PENTATRICOPEPTIDEREPEAT PROTEIN336, which copurifies with mitochon-drial polysomes but whose action on mitochondrial trans-lation is currently not understood (Uyttewaal et al., 2008b).More recently, the MAIZE PENTATRICOPEPTIDEREPEAT PROTEIN6 (MPPR6) was demonstrated to beessential for the translation of the ribosomal protein s3(rps3) mitochondrial mRNA (Manavski et al., 2012).However, mppr6 mutants accumulate 59 extended rps3transcripts, suggesting that the primary function ofMPPR6 resides in the 59 processing of rps3 transcriptsto yield translatable mature mRNA. Finally, twofertility-restorer PPR proteins were shown to directlyor indirectly inhibit the translation of their cognatesterility-inducing mitochondrial gene. This concernsthe rice (Oryza sativa) RESTORER OF FERTILITY1A(RF1A) protein that upon RNA processing producesa nontranslatable open reading frame79 transcriptand the radish (Raphanus sativus) PPR-B protein thatis suspected to specifically impede the loading orthe progression of mitochondrial ribosomes on theorf138 mRNA (Kazama et al., 2008; Uyttewaal et al.,2008a).

Several PPR proteins have also been found to par-ticipate in plastid translation, and PPR10 is the only onefor which the molecular action has been elucidated(Fisk et al., 1999; Schmitz-Linneweber et al., 2005; Caiet al., 2011; Prikryl et al., 2011; Zoschke et al., 2012,2013). PPR10 function involves binding to the atpH 59mRNA leader to prevent the formation of a stem-loopstructure that otherwise would sequestrate an SD se-quence and thereby impede ribosome entry to the atpHmRNA (Prikryl et al., 2011). The example of PPR10 in-dicates that altering the structure of RNA sequencescould represent a major action of PPR proteins to dis-close RNA segments and allow the action of other








RNA processing enzymes, notably the ones enhancingmRNA translation. In the case of MTL1, we showedthat the 59 end processing of nad7 is not perturbed inmtl1mutants, indicating that MTL1 primary function isnot in nad7 59 processing. Thereby, this strongly sup-ports a direct influence of MTL1 on nad7mature mRNAtranslation through a mechanism that still needs tobe understood. According to the PPR code (Barkanet al., 2012), a rather good putative binding site(GAATCATCTCTTT) for MTL1 is present 35 nucleo-tides upstream of the nad7 AUG. However, we did notobserve obvious binding of recombinant MTL1 to thissequence by gel-shift assay. The 59 untranslated regionof nad7 transcript is 375 nucleotides long and, as inmostmitochondrial mRNAs, lacks any obvious SD sequence.MTL1 may be necessary to recruit or guide the mito-chondrial ribosomes toward the correct start codon ofnad7. To this end, MTL1 may influence short- or long-range RNA interactions within the nad7 59 leader toprevent the formation of a stem-loop structure, asproposed for the plastid PPR10 protein. Alternatively,MTL1 may interact with the mitochondrial ribosomesor may be necessary to recruit other cofactors essentialfor nad7 translation. Further molecular investigations ofthe mode of action of MTL1 are necessary to find whichone of these hypotheses is correct.

The MTL1 Protein Is Also Essential for Optimal Splicingof nad7 Intron 2

Our results showed that mtl1 plants were also af-fected in the splicing of several mitochondrial group IIintrons. This concerned introns present in nad2 andnad7 pre-mRNAs (Supplemental Fig. S5). The decreasemeasured for nad2 intron splicing was very modest andinduced no reduction in the accumulation of the cor-responding mature transcript (Supplemental Fig. S4).These very mild alterations could simply representpleiotropic effects of the loss of complex I, as observedin several other complex I mutants (Haïli et al., 2013;Colas des Francs-Small et al., 2014; Hsieh et al., 2015),and do not support a direct role of MTL1 in the splicingof these introns. Oppositely, the 16-fold decrease ofnad7 intron 2 splicing efficiency resulted in a measured43 reduction in the accumulation level of nad7 maturemRNA (Fig. 4; Supplemental Fig. S5). This effect issufficiently significant to support a role of MTL1 in thesplicing of this intron. However, the comparison madewith the bir6 mutant clearly showed that the reductionof nad7 splicing inmtl1mutants is insufficient to explainthe lack of complex I in these plants. This conclusionis further supported by the recent study of the slowgrowth3 (slo3) mutant, which displays a similar reduc-tion in nad7 intron 2 splicing but still accumulates de-tectable respiratory complex I (Hsieh et al., 2015). Acomparison of published Arabidopsis mitochondrialsplicing mutants indicates that splicing reductions of atleast 5003 compared with the wild type seems neces-sary to result in a nearly complete loss of complex I on

Blue Native gels (Koprivova et al., 2010; Kühn et al.,2011; Cohen et al., 2014; Colas des Francs-Small et al.,2014; Hsieh et al., 2015). The decrease in splicing effi-ciency can be much milder, though, when severalcomplex I introns are concomitantly affected in a samemutant (Keren et al., 2012; Zmudjak et al., 2013). Themanner by which PPR proteins assist group II intronsplicing is currently unknown at the molecular level.Mitochondrial, but also plastid, group II introns havelost their ability to self-splice and require the partici-pation of numerous nucleus-encoded general and spe-cific factors to achieve efficient splicing in vivo (Bonen,2008; Brown et al., 2014). Although poorly conserved atthe sequence level, group II introns arrange themselvesinto a phylogenetically conserved secondary structurecomprising six helical domains that radiate from acentral hub (Pyle et al., 2007). It has been speculated,based on their activity that is likely limited to a passivebinding to RNA, that P-type PPR proteins may facilitateintron splicing by stabilizing the structure of group IIintrons into their catalytically active forms. Severalnucleus-encoded splicing factors were reported previ-ously to assist the removal of nad7 intron 2, namely thetwo general splicing factors NUCLEAR-ENCODEDMATURASE2 and MITOCHONDRIAL CHLORO-PLAST RNA SPLICING ASSOCIATED FACTOR-like SPLICING FACTOR1 as well as the SLO3 PPRprotein, which is specific for this intron (Keren et al.,2009; Zmudjak et al., 2013; Hsieh et al., 2015). It isinteresting that none of the corresponding singlemutants showed a complete loss of nad7 intron 2splicing. This suggests that the loss of one of thesefactors at a time may be partially compensated by theothers and that, in the absence of MTL1, the nad7 in-tron 2 conserves a poorly active structure resulting ina suboptimal splicing reaction. In the future, it will beinteresting to explore the interfunctionality of thesefour splicing factors and understand how they mayfunctionally promote the removal of nad7 intron 2 inArabidopsis.

MTL1 Is One of the First Mitochondrial BifunctionalPPR-P Proteins Identified in Plants

Therefore, our results revealed that the ArabidopsisMTL1 protein carries two independent activities, bothof which are required for the proper expression of thenad7 transcript. It is currently difficult to tell with cer-tainty which of these two functions is the most essentialwith respect to nad7 expression. However, as indicated,we obtained good indications supporting that the de-fect in nad7 intron 2 splicing is insufficient to explain thelack of complex I in mtl1 mutants, which would favorthe translation as being the most essential role forMTL1. This hypothesis is hard to validate, since wehave no way to verify that nad7 mRNA translationwould still not occur if only fully spliced nad7 tran-scripts accumulated in mtl1 plants. Effectively, an in-crease in the production of mature nad7 mRNA may


Haïli et al.






result in a certain degree of translation of nad7 tran-script in the absence of MTL1. It is not uncommon toobserve multiple and apparently unrelated effects ofppr mutations on RNA metabolism in plant organelles(for review, see Barkan and Small, 2014). Some of theseeffects can result from secondary consequences of pprmutations, whereas others can be directly imputable tothe inactivation of multifunctional PPR proteins. Forinstance, PPR proteins involved in C-to-U RNA editingare often considered multifunctional in the sense thatthey frequently edit multiple RNA targets, notably inplant mitochondria (Hammani and Giegé, 2014). PPRproteins supporting completely different RNA pro-cessing activities have been foundmuch less frequently,and interestingly, known examples concern PPR pro-teins having a role in mRNA translation. The plastidPROTON GRADIENT REGULATION3 (PGR3) andPPR10 proteins were shown to be involved in bothstabilization and translation of their cognate mRNAtargets (Yamazaki et al., 2004; Cai et al., 2011; Prikrylet al., 2011).The analysis of PPR10 in maize provided a molecular

framework to explain how a PPR protein can achieveboth of these functions. As already mentioned, PPR10facilitates the translation of the atpH mRNA byuncovering a ribosome-binding site and, concomi-tantly, through the same binding site stabilizes the atpHtranscript by blocking the progression of degrading ri-bonucleases (Prikryl et al., 2011). PPR proteins involvedin organelle translation, like CRP1 and MPPR6 or therestorer of fertility RF1A of rice, have also been asso-ciated with RNA cleavage events (Fisk et al., 1999;Kazama et al., 2008; Manavski et al., 2012). Others, likePGR3, ATP4, and SUPPRESSOR OF VARIEGATION7but also CRP1, are required for the translation of dis-tinct mRNAs (Barkan et al., 1994; Yamazaki et al., 2004;Cai et al., 2011; Zoschke et al., 2012, 2013). Oppositely,PPR proteins engaged in intron splicing are rarelyfound to be involved in other mRNAprocessing events,and MTL1 stands as one of the first counterexamples inplants. A precedent was found, though, in plastids withthe maize PPR5 protein, which was shown to facilitateintron splicing of trnG-UCC pre-mRNA and to con-comitantly stabilize the unspliced precursor by mask-ing an RNase-sensitive site (Williams-Carrier et al.,2008). Another example concerns ATPF EDITINGFACTOR1/MITOCHONDRIAL PPR25, which is essen-tial for the splicing of atpF in plastids but which alsopromotes RNA editing of nad5 and atpF in mitochon-dria and plastids, respectively (Yap et al., 2015). Themultifunctionality of PPR proteins is facilitated by thefact that this class of RNA-binding proteins show a re-laxed recognition specificity to some extent, which al-lows them to recognize multiple RNA targets sharinglimited sequence homology (Barkan et al., 2012).However, the apparent restriction of PPR proteins en-gaged in splicing to a single function could reflect thenecessity for these proteins to fulfill their action withvery high specificity, preventing them from havingseveral RNA-binding targets. Therefore, the dual

activity of MTL1 in both nad7 mRNA translation andsplicing represents an interesting new situation for PPRproteins that very likely involves distinct RNA-binding sites. Additional investigations are requiredto understand how MTL1 can control these twoapparently disconnected RNA processing events.Therefore, MTL1 stands as a valuable model to un-derstand the multifunctionality of PPR proteins, andits molecular deciphering will provide new insightinto the mechanisms governing mRNA expression inplant mitochondria.

MATERIALS AND METHODS

Plant Material, Growth Conditions, andComplementation Analysis

The Arabidopsis (Arabidopsis thaliana) Col-0 plants were obtained from theArabidopsis Stock Centre of the Institut National de la Recherche Agronomiquein Versailles (http://dbsgap.versailles.inra.fr/portail/). The N539066 (mtl1-1),N567427 (mtl1-2), and bir6 (N500310) Arabidopsis T-DNA insertion lines wererecovered from the European Arabidopsis Stock Centre (http://arabidopsis.info/). Plants homozygous for the insertions were identified by PCR geno-typing. mtl1-1 and mtl1-2 mutants were genotyped using the MTL1-3 andMTL1-5 primers combined with the LBSALK2 primer. Plants were grown onsoil under long-day conditions (16 h of light and 8 h of dark). For the comple-mentation test with a MYC-tagged version of MTL1, the complete MTL1 genewas amplified by PCR from Arabidopsis total DNA using GWMTL1-1 andGWMTL1-11 primers, cloned into the pDONR207 vector by Gateway BP re-action (Invitrogen), and subsequently transferred into the pGWB20 expressionvector (Nakagawa et al., 2007) by LR reaction (Invitrogen). The resulting vectorwas transformed into Agrobacterium tumefaciens C58C51 and used for floral diptransformation of mtl1-1/MTL1 and mtl1-2/MTL1 heterozygous plants. Trans-formed plants were selected on hygromycin, and transgenic homozygousmutants were identified by PCR analysis using the GWMTL1-2 and MTL1-8primers to find homozygous mutant plants, the LBSALK2 and GWMTL1-2primers to check for the presence of the T-DNA mutations, and the P35SU andGWMTL1-2 primers to check for the presence of the complementing transgene(MTL1-MYC).

Primers

The primers used in this work are listed in Supplemental Table S1.

Analyses of Targeting via GFP Fusion

The DNA region encoding the first 109 amino acids of MTL1 was PCRamplified with GWMTL1-1 and GWMTL1-6 primers. The AttB1 and AttB2recombination sites were then completed in a second PCR amplificationusing the GW5 and GW3 oligonucleotides. The PCR product was cloned intopDONR207 vector using the Gateway BP clonase enzyme mix and sequencedto check PCR accuracy. The obtained entry clone was then transferredinto pMDC83 expression vector (Curtis and Grossniklaus, 2003) by GatewayLR reaction to create a translational fusion between the MTL1 targetingsequence and the GFP coding sequence. The resulting vector was trans-formed into A. tumefaciens C58C51 and used for floral dip transformation ofArabidopsis Col-0 plants. Transgenic plants were selected on hygromycin,and GFP fluorescence was visualized in root cells by confocal microscopy.Prior to observation, root cell mitochondria were labeled by soaking thesamples for 5 min in a solution containing 0.1 mM MitoTracker Red (MolecularProbes).

Protein Extraction and Immunoblot Analysis

Mitochondria and chloroplasts were isolated from Arabidopsis flower budsand young leaves, respectively, using protocols as described (Uyttewaal et al.,2008a). Total and organelle protein samples were extracted in buffer containing




http://dbsgap.versailles.inra.fr/portail/

http://arabidopsis.info/

http://arabidopsis.info/



30 mM HEPES/KOH, pH 7.7, 10 mM magnesium acetate, 150 mM potassiumacetate, 10% (v/v) glycerol, and 0.5% (w/v) CHAPS and separated by SDS-PAGE. After electrophoresis, proteins on gels were transferred onto apolyvinylidene difluoride membrane (Perkin-Elmer) and incubated with spe-cific polyclonal antibodies recognizing Nad9 (diluted 1:100,000; Lamattinaet al., 1993), FDH (diluted 1:2,000; Colas des Francs-Small et al., 1992), ATPC(diluted 1:5,000; a gift of Jorg Meurer, Ludwig-Maximilians-Universität), Nad7(diluted 1:2,000; Pineau et al., 2008), AOX (diluted 1:50; Elthon et al., 1989), andPORIN (diluted 1:10,000; a gift of D. Day, University of Western Australia).Following the hybridization of peroxidase-conjugated secondary anti-bodies, signals were detected using Western Lightning ECL plus reagents(Perkin-Elmer) and visualized by a LAS4000 chemiluminescence analyzer (FujiFilm). The apparent molecular masses of the proteins were estimated withprestained molecular mass markers (Fermentas). Analysis of the submito-chondrial distribution of MTL1 protein was performed as described (Uyttewaalet al., 2008a).

Blue Native-PAGE and Complex I Activity Assay

The equivalent of 500 mg of total mitochondrial proteins was solubilizedin buffer containing 50 mM Bis-Tris/HCl, pH 7, 750 mM 6-aminohexanoic acid,0.5 mM EDTA, and 5 g g21 mitochondrial protein of digitonin. Samples wereelectrophoresed in a 4% to 13% (w/v) Nupage Bis/Tris native gel (Life Tech-nologies) according to the manufacturer’s instructions. Following electropho-resis, the gels were either stained with Coomassie Blue or incubated in a buffercontaining 0.1 M Tris-HCl, pH 7.4, 0.2 mM NADH, and 0.2% (w/v) nitrobluetetrazolium to reveal the NADH dehydrogenase activity of complex I. A fixingsolution containing 30% (v/v) methanol and 10% (v/v) acetic acid was thenused to stop the reaction.

RNA Extraction and Analysis

Total RNA was isolated using TRIzol reagent (Life Technologies) followingthe manufacturer’s recommendations and treated with RNase-free DNase I(RNeasy Mini Kit; Qiagen). For the RT-PCR experiment, approximately 500 ngof total RNA was reverse transcribed using random hexamers (Eurofins) and400 units of reverse transcriptase (Thermo Scientific) in a final volume of 20 mL.For PCR amplification, 2mL of a 50-fold dilution of the cDNA solutionwas usedas a template. Amplification of the MTL1 gene transcript was performed usingGWMTL1-1 and GWMTL1-6 primers. PCR on the BIO2 gene transcript wasused as a control using the BIO2-6 and BIO2-8 primers. Quantitative RT-PCRanalyses of mitochondrial transcript abundance and splicing efficiency wereperformed as in described by Haïli et al. (2013). For northern-blot analysis,15 mg of RNAwas loaded on formaldehyde agarose gels and transferred onto anylon membrane (Genescreen). RNA integrity, loading, and transfer wereverified by staining the membrane with Methylene Blue. Hybridizations with32P-radiolabeled gene-specific probes were performed under high-stringencyconditions overnight at 65°C in Church buffer (7% [w/v] SDS, 0.25 M

Na2HPO4, pH 7.4, 2 mM EDTA, and 200 mg mL21 heparin). Blots were washed,and hybridized transcripts were detected and analyzed using a phosphor-imager screen (Fla5000; Fuji Film). Primers used to generate gene-specificprobes are listed in Supplemental Table S1.

Circular RT-PCR analysis was performed as described by Haïli et al. (2013).Primer pairs used to perform the different circular RT-PCR tests on complex Igenes are listed in Supplemental Table S1.

Polysome Association Analysis

Polysomeswere prepared as described previously (Barkan, 1993) except thatthey were isolated from flower buds and not from leaf tissue. Ten fractions of900 mL starting from the top of the gradients were collected. RNAs from eachfraction were extracted using TRIzol reagent (Life Technologies) following themanufacturer’s recommendations, precipitated with ethanol, and resuspendedin 30 mL of RNase-free water. Ten microliters of each RNA fraction was frac-tionated on 1.5% (w/v) formaldehyde agarose gels, transferred onto a nylonmembrane (Genescreen), and hybridized with gene-specific probes as de-scribed above.

Gas-Exchange Measurements

Respiratory oxygen consumption was measured as described by Haïli et al.(2013).

Supplemental Data

The following supplemental materials are available.

Supplemental Figure S1. Phenotypes of mtl1 mutant plants in comparisonwith the wild type.

Supplemental Figure S2. AnMTL1-MYC translational fusion can function-ally complement mtl1 mutants.

Supplemental Figure S3. Respiratory activity measurements in wild-typeand mtl1 plants.

Supplemental Figure S4. RNA gel-blot analysis determining the steady-state levels of mitochondria-encoded complex I mRNAs in the wild typeand mtl1 mutants.

Supplemental Figure S5. Quantitative analysis of intron splicing and ma-ture mRNA accumulation in mtl1 mutants.

Supplemental Figure S6. Circular RT-PCR analysis of mitochondrial com-plex I genes in the wild type and mtl1 mutants.

Supplemental Figure S7. Comparative analysis of mtl1 and bir6 mutants.

Supplemental Figure S8. The MTL1 protein cosediments with mitochon-drial polysomes on sucrose density gradients.

Supplemental Figure S9. Quantitation of polysome sedimentation hybrid-ization results.

Supplemental Table S1. DNA primers used in this analysis.

ACKNOWLEDGMENTS

We thank Guillaume Tcherkez for technical assistance with respirationmeasurements and Françoise Budar for critical reading of the article.

Received October 8, 2015; accepted November 3, 2015; published November 4,2015.

LITERATURE CITED

Ban T, Ke J, Chen R, Gu X, Tan MHE, Zhou XE, Kang Y, Melcher K, ZhuJK, Xu HE (2013) Structure of a PLS-class pentatricopeptide repeatprotein provides insights into mechanism of RNA recognition. J BiolChem 288: 31540–31548

Barkan A (1993) Nuclear mutants of maize with defects in chloroplastpolysome assembly have altered chloroplast RNA metabolism. PlantCell 5: 389–402

Barkan A (2011) Expression of plastid genes: organelle-specific elaborationson a prokaryotic scaffold. Plant Physiol 155: 1520–1532

Barkan A, Rojas M, Fujii S, Yap A, Chong YS, Bond CS, Small I (2012) Acombinatorial amino acid code for RNA recognition by pentatricopep-tide repeat proteins. PLoS Genet 8: e1002910

Barkan A, Small I (2014) Pentatricopeptide repeat proteins in plants. AnnuRev Plant Biol 65: 415–442

Barkan A, Walker M, Nolasco M, Johnson D (1994) A nuclear mutation inmaize blocks the processing and translation of several chloroplastmRNAs and provides evidence for the differential translation of alter-native mRNA forms. EMBO J 13: 3170–3181

Beick S, Schmitz-Linneweber C, Williams-Carrier R, Jensen B, Barkan A(2008) The pentatricopeptide repeat protein PPR5 stabilizes a specifictRNA precursor in maize chloroplasts. Mol Cell Biol 28: 5337–5347

Bonen L (2004) Translational machinery in plant organelles. In H Daniell, CChase, eds, Molecular Biology and Biotechnology of Plant Organelles.Springer, Dordrecht, The Netherlands, pp 323–345

Bonen L (2008) Cis- and trans-splicing of group II introns in plant mito-chondria. Mitochondrion 8: 26–34

Bonen L, Calixte S (2006) Comparative analysis of bacterial-origin genesfor plant mitochondrial ribosomal proteins. Mol Biol Evol 23: 701–712

Brown GG, Colas des Francs-Small C, Ostersetzer-Biran O (2014) GroupII intron splicing factors in plant mitochondria. Front Plant Sci 5: 35

Cai W, Okuda K, Peng L, Shikanai T (2011) PROTON GRADIENTREGULATION 3 recognizes multiple targets with limited similarity andmediates translation and RNA stabilization in plastids. Plant J 67: 318–327


Haïli et al.















Choi B, Acero MM, Bonen L (2012) Mapping of wheat mitochondrialmRNA termini and comparison with breakpoints in DNA homologyamong plants. Plant Mol Biol 80: 539–552

Cohen S, Zmudjak M, Colas des Francs-Small C, Malik S, Shaya F, KerenI, Belausov E, Many Y, Brown GG, Small I, et al (2014) nMAT4, amaturase factor required for nad1 pre-mRNA processing and matura-tion, is essential for holocomplex I biogenesis in Arabidopsis mito-chondria. Plant J 78: 253–268

Colas des Francs-Small C, Ambard-Bretteville F, Darpas A, Sallantin M,Huet JC, Pernollet JC, Rémy R (1992) Variation of the polypeptidecomposition of mitochondria isolated from different potato tissues.Plant Physiol 98: 273–278

Colas des Francs-Small C, Falcon de Longevialle A, Li Y, Lowe E, TanzSK, Smith C, Bevan MW, Small I (2014) The pentatricopeptide repeatproteins TANG2 and ORGANELLE TRANSCRIPT PROCESSING439 areinvolved in the splicing of the multipartite nad5 transcript encoding asubunit of mitochondrial complex I. Plant Physiol 165: 1409–1416

Coquille S, Filipovska A, Chia T, Rajappa L, Lingford JP, Razif MFM,Thore S, Rackham O (2014) An artificial PPR scaffold for programmableRNA recognition. Nat Commun 5: 5729

Curtis MD, Grossniklaus U (2003) A Gateway cloning vector set for high-throughput functional analysis of genes in planta. Plant Physiol 133:462–469

Elthon TE, Nickels RL, McIntosh L (1989) Monoclonal antibodies to the alter-native oxidase of higher plant mitochondria. Plant Physiol 89: 1311–1317

Fisk DG, Walker MB, Barkan A (1999) Molecular cloning of the maize genecrp1 reveals similarity between regulators of mitochondrial and chlo-roplast gene expression. EMBO J 18: 2621–2630

Gully BS, Shah KR, Lee M, Shearston K, Smith NM, Sadowska A, BlytheAJ, Bernath-Levin K, Stanley WA, Small ID, et al (2015) The designand structural characterization of a synthetic pentatricopeptide repeatprotein. Acta Crystallogr D Biol Crystallogr 71: 196–208

Gutierres S, Sabar M, Lelandais C, Chetrit P, Diolez P, Degand H, BoutryM, Vedel F, de Kouchkovsky Y, De Paepe R (1997) Lack of mito-chondrial and nuclear-encoded subunits of complex I and alteration ofthe respiratory chain in Nicotiana sylvestris mitochondrial deletionmutants. Proc Natl Acad Sci USA 94: 3436–3441

Haïli N, Arnal N, Quadrado M, Amiar S, Tcherkez G, Dahan J, Briozzo P,Colas des Francs-Small C, Vrielynck N, Mireau H (2013) The penta-tricopeptide repeat MTSF1 protein stabilizes the nad4 mRNA in Ara-bidopsis mitochondria. Nucleic Acids Res 41: 6650–6663

Hammani K, des Francs-Small CC, Takenaka M, Tanz SK, Okuda K,Shikanai T, Brennicke A, Small I (2011) The pentatricopeptide repeatprotein OTP87 is essential for RNA editing of nad7 and atp1 transcriptsin Arabidopsis mitochondria. J Biol Chem 286: 21361–21371

Hammani K, Giegé P (2014) RNA metabolism in plant mitochondria.Trends Plant Sci 19: 380–389

Hazle T, Bonen L (2007) Comparative analysis of sequences precedingprotein-coding mitochondrial genes in flowering plants. Mol Biol Evol24: 1101–1112

Howard MJ, Lim WH, Fierke CA, Koutmos M (2012) Mitochondrial ri-bonuclease P structure provides insight into the evolution of catalyticstrategies for precursor-tRNA 59 processing. Proc Natl Acad Sci USA109: 16149–16154

Hsieh WY, Liao JC, Chang CY, Harrison T, Boucher C, Hsieh MH (2015)The SLOW GROWTH3 pentatricopeptide repeat protein is required forthe splicing of mitochondrial NADH Dehydrogenase Subunit7 intron 2 inArabidopsis. Plant Physiol 168: 490–501

Hsu YW, Wang HJ, Hsieh MH, Hsieh HL, Jauh GY (2014) ArabidopsismTERF15 is required for mitochondrial nad2 intron 3 splicing andfunctional complex I activity. PLoS ONE 9: e112360

Kazama T, Nakamura T, Watanabe M, Sugita M, Toriyama K (2008)Suppression mechanism of mitochondrial ORF79 accumulation by Rf1protein in BT-type cytoplasmic male sterile rice. Plant J 55: 619–628

Ke J, Chen RZ, Ban T, Zhou XE, Gu X, Tan MHE, Chen C, Kang Y,Brunzelle JS, Zhu JK, et al (2013) Structural basis for RNA recognitionby a dimeric PPR-protein complex. Nat Struct Mol Biol 20: 1377–1382

Keren I, Bezawork-Geleta A, Kolton M, Maayan I, Belausov E, Levy M,Mett A, Gidoni D, Shaya F, Ostersetzer-Biran O (2009) AtnMat2, anuclear-encoded maturase required for splicing of group-II introns inArabidopsis mitochondria. RNA 15: 2299–2311

Keren I, Tal L, des Francs-Small CC, Araújo WL, Shevtsov S, Shaya F, FernieAR, Small I, Ostersetzer-Biran O (2012) nMAT1, a nuclear-encoded

maturase involved in the trans-splicing of nad1 intron 1, is essential formitochondrial complex I assembly and function. Plant J 71: 413–426

Khrouchtchova A, Monde RA, Barkan A (2012) A short PPR protein re-quired for the splicing of specific group II introns in angiosperm chlo-roplasts. RNA 18: 1197–1209

Koprivova A, des Francs-Small CC, Calder G, Mugford ST, Tanz S, LeeBR, Zechmann B, Small I, Kopriva S (2010) Identification of a penta-tricopeptide repeat protein implicated in splicing of intron 1 of mito-chondrial nad7 transcripts. J Biol Chem 285: 32192–32199

Kühn K, Carrie C, Giraud E, Wang Y, Meyer EH, Narsai R, des Francs-Small CC, Zhang B, Murcha MW, Whelan J (2011) The RCC1 familyprotein RUG3 is required for splicing of nad2 and complex I biogenesisin mitochondria of Arabidopsis thaliana. Plant J 67: 1067–1080

Lamattina L, Gonzalez D, Gualberto J, Grienenberger JM (1993) Higherplant mitochondria encode an homologue of the nuclear-encoded 30-kDasubunit of bovine mitochondrial complex I. Eur J Biochem 217: 831–838

Lurin C, Andrés C, Aubourg S, Bellaoui M, Bitton F, Bruyère C, CabocheM, Debast C, Gualberto J, Hoffmann B, et al (2004) Genome-wideanalysis of Arabidopsis pentatricopeptide repeat proteins reveals theiressential role in organelle biogenesis. Plant Cell 16: 2089–2103

Manavski N, Guyon V, Meurer J, Wienand U, Brettschneider R (2012) Anessential pentatricopeptide repeat protein facilitates 59 maturation andtranslation initiation of rps3 mRNA in maize mitochondria. Plant Cell24: 3087–3105

Nakagawa T, Kurose T, Hino T, Tanaka K, Kawamukai M, Niwa Y,Toyooka K, Matsuoka K, Jinbo T, Kimura T (2007) Development ofseries of Gateway binary vectors, pGWBs, for realizing efficient constructionof fusion genes for plant transformation. J Biosci Bioeng 104: 34–41

O’Toole N, Hattori M, Andrés C, Iida K, Lurin C, Schmitz-Linneweber C,Sugita M, Small I (2008) On the expansion of the pentatricopeptiderepeat gene family in plants. Mol Biol Evol 25: 1120–1128

Pfalz J, Bayraktar OA, Prikryl J, Barkan A (2009) Site-specific binding of aPPR protein defines and stabilizes 59 and 39 mRNA termini in chloro-plasts. EMBO J 28: 2042–2052

Pineau B, Layoune O, Danon A, De Paepe R (2008) L-Galactono-1,4-lactone dehydrogenase is required for the accumulation of plant respi-ratory complex I. J Biol Chem 283: 32500–32505

Prikryl J, Rojas M, Schuster G, Barkan A (2011) Mechanism of RNA sta-bilization and translational activation by a pentatricopeptide repeatprotein. Proc Natl Acad Sci USA 108: 415–420

Pyle AM, Fedorova O, Waldsich C (2007) Folding of group II introns: a modelsystem for large, multidomain RNAs? Trends Biochem Sci 32: 138–145

Ringel R, Sologub M, Morozov YI, Litonin D, Cramer P, Temiakov D (2011)Structure of human mitochondrial RNA polymerase. Nature 478: 269–273

Schmitz-Linneweber C, Williams-Carrier R, Barkan A (2005) RNA im-munoprecipitation and microarray analysis show a chloroplast penta-tricopeptide repeat protein to be associated with the 59 region of mRNAswhose translation it activates. Plant Cell 17: 2791–2804

Small ID, Peeters N (2000) The PPR motif: a TPR-related motif prevalent inplant organellar proteins. Trends Biochem Sci 25: 46–47

Sørensen MA, Fricke J, Pedersen S (1998) Ribosomal protein S1 is requiredfor translation of most, if not all, natural mRNAs in Escherichia coli invivo. J Mol Biol 280: 561–569

Takenaka M, Zehrmann A, Brennicke A, Graichen K (2013) Improvedcomputational target site prediction for pentatricopeptide repeat RNAediting factors. PLoS ONE 8: e65343

Uyttewaal M, Arnal N, Quadrado M, Martin-Canadell A, Vrielynck N,Hiard S, Gherbi H, Bendahmane A, Budar F, Mireau H (2008a)Characterization of Raphanus sativus pentatricopeptide repeat proteinsencoded by the fertility restorer locus for Ogura cytoplasmic male ste-rility. Plant Cell 20: 3331–3345

Uyttewaal M, Mireau H, Rurek M, Hammani K, Arnal N, Quadrado M,Giegé P (2008b) PPR336 is associated with polysomes in plant mito-chondria. J Mol Biol 375: 626–636

Williams-Carrier R, Kroeger T, Barkan A (2008) Sequence-specific bindingof a chloroplast pentatricopeptide repeat protein to its native group IIintron ligand. RNA 14: 1930–1941

Yagi Y, Hayashi S, Kobayashi K, Hirayama T, Nakamura T (2013) Eluci-dation of the RNA recognition code for pentatricopeptide repeat pro-teins involved in organelle RNA editing in plants. PLoS ONE 8: e57286

Yamazaki H, Tasaka M, Shikanai T (2004) PPR motifs of the nucleus-encoded factor, PGR3, function in the selective and distinct steps ofchloroplast gene expression in Arabidopsis. Plant J 38: 152–163





Yap A, Kindgren P, Colas des Francs-Small C, Kazama T, Tanz SK,Toriyama K, Small I (2015) AEF1/MPR25 is implicated in RNA editingof plastid atpF and mitochondrial nad5, and also promotes atpF splicingin Arabidopsis and rice. Plant J 81: 661–669

Yin P, Li Q, Yan C, Liu Y, Liu J, Yu F, Wang Z, Long J, He J, Wang HW,et al (2013) Structural basis for the modular recognition of single-stranded RNA by PPR proteins. Nature 504: 168–171

Zhelyazkova P, Hammani K, Rojas M, Voelker R, Vargas-Suárez M,Börner T, Barkan A (2012) Protein-mediated protection as the pre-dominant mechanism for defining processed mRNA termini in landplant chloroplasts. Nucleic Acids Res 40: 3092–3105

Zmudjak M, Colas des Francs-Small C, Keren I, Shaya F, Belausov E, Small I,Ostersetzer-Biran O (2013) mCSF1, a nucleus-encoded CRM protein requiredfor the processing of many mitochondrial introns, is involved in the biogenesisof respiratory complexes I and IV in Arabidopsis. New Phytol 199: 379–394

Zoschke R, Kroeger T, Belcher S, Schöttler MA, Barkan A, Schmitz-Linneweber C (2012) The pentatricopeptide repeat-SMR protein ATP4promotes translation of the chloroplast atpB/E mRNA. Plant J 72: 547–558

Zoschke R, Qu Y, Zubo YO, Börner T, Schmitz-Linneweber C (2013)Mutation of the pentatricopeptide repeat-SMR protein SVR7 impairsaccumulation and translation of chloroplast ATP synthase subunits inArabidopsis thaliana. J Plant Res 126: 403–414


Haïli et al.



Documents

The MTL1 Pentatricopeptide Repeat Protein Is Required for Both … · The MTL1 Pentatricopeptide Repeat Protein Is Required for Both Translation and Splicing of the Mitochondrial