BnaA02.YTG1,encoding a tetratricopeptide repeat protein,is required for early chloroplast biogenesis in Brassica napus

Haiyan Zhang,Xiaoting Li,Yebitao Yang,Kaining Hu,Xianming Zhou,Jing Wen,Bin Yi,Jinxiong Shen,Chaozhi Ma,Tingdong Fu,Jinxing Tu*

National Key Laboratory of Crop Genetic Improvement,Huazhong Agricultural University,Wuhan 430070,Hubei,China

Keywords:Brassica napus Chloroplast MORFs RNA editing TPR

ABSTRACT Chloroplasts are essential for plant growth and development,as they play a key role in photosynthesis.The chloroplast biogenesis process is complex and its regulatory mechanism remains elusive.We characterized a spontaneous Brassica napus(rapeseed)mutant,ytg,that showed a delayed greening phenotype in all green organs and retarded growth.We identified BnaA02.YTG1 encoding a chloroplastlocalized tetratricopeptide repeat protein widely expressed in rapeseed tissues.We speculated that the ytg phenotype was caused by the deletion of BnaA02.YTG1 based on sequence comparison of 4608 (with normal green leaves,isolated from the elite Chinese rapeseed cultivar ZS11)and ytg combined with transcriptome data and CRISPR/Cas9 gene editing results.The homologous gene (BnaC02.YTG1) restored the phenotype of the mutant.BnaA02.YTG1 interacted with MORF2,MORF8,and OZ1.RNA editing of the ndhD-2,ndhF-290,petL-5,and ndhG-50 plastid transcripts was affected in ytg.These findings suggested that BnaA02.YTG1 participates in RNA editing events.We predicted 29 RNA editing sites in the chloroplast of Brassica napus by comparison with the Arabidopsis chloroplast genome.We conclude that BnaA02.YTG1 affects the posttranscriptional regulation of plastid gene expression and suggest that a tetratricopeptide repeat protein is involved in the chloroplast RNA editing in rapeseed.

1.Introduction

Chloroplasts,semiautonomous organelles in higher plants,are responsible for photosynthesis and function in metabolic networks in photosynthesis cells.Chloroplasts contain a very small genome and the vast majority of their approximately 3000 proteins are nucleus-encoded [1].Thus,the enzymatic and photosynthetic machinery of the chloroplast is encoded by both the chloroplast and nuclear genomes [2].In Arabidopsis,plastidic RNA polymerase sigma(SIG)factors are nuclear-encoded proteins that contribute to plant greening and plastid development by regulating gene transcription in chloroplasts and modulating retrograde signaling from the plastid to the nucleus [3].SIG2 and SIG6 play partially redundant roles in plastid transcription,with the loss of either resulting in delayed chloroplast biogenesis [4–6].

RNA editing,a ubiquitous phenomenon [7],seems to be confined to chloroplasts and mitochondria [8,9].Transcripts encoding chloroplast and mitochondrial proteins are profoundly affected by RNA editing,which often changes the encoded amino acid predicted from the DNA sequence [10–14].RNA editing can increase the diversity of the transcriptome and the proteome,which is beneficial for life to adapt to complex environmental changes[15–17].Some RNA editing mutants show reduced or lost editing activity at several sites and displays various macroscopic phenotypes from pale or albino,in the case of chloroplasts to growth retardation or even embryonic lethality[18].Cytidine to uridine(C-to-U)editing is the main type of RNA editing in plants.Several factors are involved in plant RNA editing.The first type of RNA editing factor is nuclear-encoded pentatricopeptide repeat (PPR) proteins,composed of a 35-amino acid motifs (defined as the PPR motif) [19–21].CRR4,a member of the PLS-type PPR proteins,was the first RNA editing factor to be discovered,and crr4 is defective specifically in RNA editing,creating an ndhD initiation codon [22].The second type of RNA editing factor is the multiple organellar RNA editing factor (MORF),also known as the RNA editing factor interacting protein(RIP)[23,24].In Arabidopsis,MORF is a small protein family with 10 members [25] that interact with site-specific RNA editing proteins.MORF2,MORF9,and a potential pseudogene are chloroplast-localized;MORF8 is dual-targeted to both chloroplasts and mitochondria.MORF2 and MORF9 are the two main RNA editing factors in chloroplasts and the editing rates of almost all editing sites are reduced in morf2 and morf9 mutants [24].More recently discovered factors involved in RNA editing are an organelle RNA recognition motif-containing (ORRM) family consisting of six members [26–30];protoporphyrinogen IX oxidase 1 (PPO1),an enzyme in the biosynthetic pathway of tetrapyrrole[31];and organelle zinc finger 1 (OZ1) [10].All the above factors constitute the RNA editosome.In addition,RNA editing in chloroplasts may contribute to a type of chloroplast-to-nucleus signaling defined by genomes uncoupled (gun) mutants in Arabidopsis [32].

Besides the PPR proteins,the most notable helical repeat proteins in the chloroplast are tetratricopeptide repeat (TPR) proteins[19,33–35].Similar to PPR proteins,TPR proteins contain tandem repeats of 34 amino acids(known as the TPR motif,one amino acid shorter than the PPR motif) [36,37].In A.thaliana,there are 141 TPR proteins [38].Since their discovery in 1990,it was found that TPR proteins seem to be mainly involved in promoting protein–protein interactions [39,40] and are essential for the biogenesis of thylakoid membranes.TPR proteins participate in almost all steps of the biosynthesis of the photosynthetic apparatus,including the concerted synthesis and assembly of lipids,pigments,metal cofactors,and dozens of proteins [38].CGL71 is involved in protecting PSI from oxidative disruption during assembly [41],and the PSI assembly process is blocked in the pyg7 mutant [42].Ycf3,which interacts directly with the PSI subunits PsaA and PsaD,is essential for the accumulation of the PSI complex[43–45].MET1 is a thylakoid-associated TPR protein involved in PSII supercomplex formation and repair in Arabidopsis[46,47].Most TPR proteins act in the formation of multi-subunit protein complexes by mediating protein–protein interactions and participate in a variety of molecular functions including RNA metabolism,transcriptional regulation,and protein biosynthesis and transport [39,40].Toc64,a translocon receptor at the outer envelope of chloroplasts (forming the Toc complex) with a dual function,and an integral membrane protein called Tic40 are involved in protein import into chloroplasts [48–50].

In this study,we discovered a novel rapeseed mutant,yellow to green (ytg),and identified the causal gene using map-based cloning.YTG encodes a TPR protein (BnaA02.YTG1).BnaA02.YTG1 may participate in RNA editing by interacting with MORF2,MORF8,and OZ1.We propose that the loss of BnaA02.YTG1 accounts for lower RNA editing rates at four plastid RNA editing sites (ndhD-2,ndhD-290,petL-5,and ndhG-50) and the disturbed expression of chloroplast-related genes.These findings suggest that BnaA02.YTG1 is required for early chloroplast development and may have a function in RNA editing in rapeseed by enhancing the function of RNA editosomes.We found that the chloroplast genome has been relatively conserved during evolution and that RNA editing sites can be identified via sequence comparison.Our identification of 29 RNA editing sites in rapeseed chloroplasts may shed light on plastid RNA editing in polyploidy crops.

2.Materials and methods

2.1.Plant materials and mapping population

Two plant materials were used:ytg,a chlorophyll-deficient mutant isolated from the progeny of a restorer recurrentselection population[51],and 4608,with normal green leaves,isolated from the elite Chinese rapeseed cultivar ZS11.The ytg mutant was crossed with 4608 to produce an F1generation.To further develop the advanced segregating populations,F1plants were successively backcrossed with the ytg mutant three and four times for generating BC3F1and BC4F1populations,respectively.

2.2.Chlorophyll measurements

Leaf samples of 4608 and the ytg mutant were harvested from the plants at the five-leaf stage under normal conditions.Fresh tissue (30 mg) from the top first to fifth leaves of 4608 and the mutant were used.Spectrophotometric quantification was performed with a UV-1800 spectrophotometer (Mapada,Shanghai,China).Chlorophyll determination followed Arnon [52].Five technical replicates of each experiment were performed.

2.3.Transmission electron microscopy

The top first and fourth leaves from 4608 and ytg were harvested.All plants were at the same stage under normal conditions.Leaves were cut into 1×1 cm sections,fixed in 2.5% (w/v) glutaraldehyde in a 0.1 mol L-1phosphate buffer(pH 7.4),and further fixed in 1% OsO4in the same buffer.Subsequent steps followed Yi et al.[53].

2.4.Measurement of major agronomic and quality traits

Plants homozygous and heterozygous at the YTG locus were identified in the BC4F1population and named NIL-4608 and NILytg,respectively.Yield and seed-quality traits were evaluated at harvest maturity,including plant height (PH),measured from the base of the aboveground plant to the tip of the main inflorescence;length of main inflorescence (LMI),measured from the base of highest primary effective branch to the tip of the main inflorescence;branch number (BN),measured as the number of effective branches;silique length (SL),measured as the mean length of 10 well-developed siliques selected from the middle part of the main inflorescence immediately above the first side branch;seeds per silique(SS),identified as the mean seed number of siliques selected for SL evaluation;silique number (SN),measured as the silique number of the main inflorescence;thousand-seed weight (TSW),measured in grams and estimated from the mean of three measurements of the weight of 1000 well-filled seeds;and oil,linoleic acid (LA),and glucosinolate contents (GSL),measured in openpollinated seeds of the whole plant using near-infrared reflectance spectroscopy.

2.5.Map-based cloning of the YTG locus

Initial mapping of the YTG locus was performed using simple sequence repeat (SSR) markers and bulked-segregant analysis.BC1F1and F2plants were randomly selected to construct DNA bulks,named G pool (green leaf) and Y pool (yellowish leaf),for genotypic analysis using the Brassica 60 K Illumina Infinium SNP array (Illumina,San Diego,CA,USA).SNP analysis followed Xu et al.[54].To identify the SNPs linked to the chlorophylldeficient phenotype,scanned SNPs differing between the G and Y pools were considered to lie in candidate gene regions.The retained SNPs were used in a BLAST search against the Darmorbzh genome database (https://www.genoscope.cns.fr/brassicanapus/)to locate the chromosome positions(E value ≤1e-12).SNPs matched to unique loci in the genome were retained for further analysis.SSR primers were designed using WebSat (https://bioinfo.inf.ufg.br/websat/) based on the available genome sequence information (http://www.genoscope.cns.fr/brassicanapus).All DNA markers used in YTG mapping are listed in Table S1.

2.6.Complementation test,overexpression test,and CRISPR

Three candidate genes (BnaA02g10470D,BnaA02g10480D,and BnaA02g10490D) were cloned into the vector pCAMBIA2300.Each construct contained a native promoter (～2 kb),exons,introns,and a 3′sequence(～1 kb).All constructs were introduced into the mutant by Agrobacterium tumefaciens-mediated transformation as previously described [49].To investigate the function of BnaA02.YTG1,the binary CRIPSR/Cas9 multiplex genome targeting vector system was used.Two single guide RNAs (sgRNA) targeting the coding regions of the BnaA02.YTG1 and BnaC02.YTG1 genes were designed using CRISPR-P 2.0 (http://crispr.hzau.edu.cn/CRISPR2/)[55].sgRNAs were inserted into the pKSE401 vector,as described previously[56].The reconstructed vector was confirmed by Sanger sequencing and introduced into 4608 with normal green leaves by A.tumefaciens-mediated transformation.CRISPR transgenic plants were genotyped by PCR amplification and Sanger sequencing.Genetic Information Research Institute (GIRI) (https://www.girinst.org/) was used for transposon prediction.

2.7.Histochemical GUS staining

The BnaA02.YTG1 promoter,a 3031-bp sequence upstream of the translation start site,was amplified from 4608 and cloned into the binary vector pCAMBIA2300-GUS.This construct was transformed into Arabidopsis(Col-0)by floral dipping[57].Positive lines in the T2generation were used for GUS staining following the manufacturer’s instructions (BL622A,Biosharp,Beijing,China).

2.8.Subcellular localization of BnaA02.YTG1 and BnaC02.YTG1

The coding sequences(CDS)of BnaA02.YTG1 was cloned into the GFP-fusion expression vector pM999-35S-GFP.PEG-calciummediated transfection was used to deliver pM999-35S-BnaA02.YTG1-GFP into the A.thaliana protoplast as previously described[58].For the subcellular localization of BnaC02.YTG1,the CDS of BnaC02.YTG1 was amplified from ZS11 and then inserted downstream of the double 35S promoter at the Kpn I site in frame with GFP in the pMDC83 vector.It was then transferred to tobacco epidermal cells in a process mediated by Agrobacterium,as previously described [59].Chloroplast autofluorescence,which was red,was used as a control.GFP fluorescence signals and chlorophyll spontaneous fluorescence were visualized with a FluoView FV1200 laser scanning confocal microscope (Olympus,Tokyo,Japan).

2.9.Real-time quantitative reverse transcription PCR

Total RNA was extracted from tissue samples with an RNA prep plant kit containing the DNase I treatment reagent (DP441,TIANGEN,Beijing,China).First-strand cDNA was reverse-transcribed using a RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific,Waltham,MA,USA),using 2 μg of total RNA.To investigate the expression pattern of BnaA02.YTG1,total RNA isolated from roots,stems,leaves,buds,and siliques was subjected to qRT-PCR.qRT-PCR reactions were performed on a CFX96 Touch Real-Time PCR detection system (Bio-Rad Laboratories,Hercules,CA,USA)using a Green Real-Time PCR Master Mix(Toyobo,Osaka,Japan).The PCR reactions,cycling protocol,and melting curves were performed according to the manufacturer’s instructions.Each sample was represented by three biological and three technical repeats.Actin7 in B.napus was used as an internal control.All primers used for qRT-PCR are listed in Table S1.

2.10.Yeast two-hybrid assays

Gal4-based yeast two-hybrid analysis was performed using the Matchmaker Gold Yeast Two-Hybrid System (PT4084-1;Clontech,San Francisco,CA,USA),and a cDNA library of all ZS11 tissues was constructed according to the manufacturer’s instructions.Amino acids 165 to 543 of BnaA02.YTG1 were cloned into the pGBKT7 vector (GAL4-BD domain) as a bait plasmid.The full-length CDS regions of BnaMORFs and other editing factors were cloned into the pGADT7 vector (GAL4-AD domain) as prey plasmids.The bait plasmids and cDNA library,or pairs of constructs,were cotransformed into AH109 yeast cells.Co-transformants were placed on SD/–Leu/–Trp or SD/–Ade/–His/–Leu/–Trp agar plates.

2.11.Split-luciferase complementation (SLC)

The CDS of BnaC04.MORF2,BnaC03.MORF8,BnaA10Toc33,BnaC09.Toc34,BnaC01.Tic22-III,and BnaC09.OZ1 were PCRamplified and cloned into the JW771-cLUC vector.The CDS of BnaA02.YTG1 was amplified and cloned into the JW772-nLUC vector.Split-luciferase complementation assays were performed as previously described [60,61].

2.12.Comparison of four major chloroplast genomes and determination of chloroplast RNA editing sites in Brassica napus

Progressive Mauve [62] with default parameters was used to investigate the collinearity of the chloroplast genomes of A.thaliana(https://www.ncbi.nlm.nih.gov/nuccore/NC_000932.1),B.rapa(https://www.ncbi.nlm.nih.gov/nuccore/NC_040849.1),B.oleracea(https://www.ncbi.nlm.nih.gov/nuccore/NC_041167.1),and B.napus(https://www.ncbi.nlm.nih.gov/nuccore/NC_016734.1).Basic local alignment search tool (BLAST) was used to perform alignments among the above four chloroplast genomes.Because the chloroplast was highly conserved during evolution from A.thaliana to B.napus,RNA editing site information for A.thaliana was used to confirm the RNA editing sites in the chloroplast genome of B.napus.

2.13.RNA editing analysis

Total RNA was isolated from the top first and fourth leaves of transgenic-positive (named HB1 and HB4) and -negative plants(named hb1 and hb4) in the complemented transgenic line PT1.The total RNA was then reverse-transcribed as templates.PCR fragments containing chloroplast RNA editing sites were obtained using specific primers(Table S1)[63].The PCR products were purified and used for Sanger DNA sequencing.The C-to-T (C-to-U in RNA) editing level of each site was measured using the relative heights of the nucleotide peaks in the sequence chromatograms,and calculated by dividing the height of ‘T’ by the sum of the heights of‘T’and‘C’.The C-to-T editing level of each site was then evaluated as previously described [32].

The results were further verified by pyrosequencing at the sites where the editing rate changed.The RT-PCR products were sent to Sangon Biotech(Shanghai)Co.,Ltd.(Shanghai,China)for sequencing.The primers used are listed in Table S1.

2.14.Chlorophyll fluorescence analysis

Chlorophyll fluorescence was measured using a PAM fluorometer Dual-PAM-100 (Walz,Effeltrich,Germany).The transient increase in chlorophyll fluorescence after turning off actinic light(AL) was monitored as previously described [64].For determining minimum fluorescence (F0),leaves were kept in darkness for at least 30 min prior to use.Fluorescence decay kinetics analysis followed the instrument software operating instructions.hb1 and HB1 were used.

2.15.RNA-seq analysis

Total RNA was obtained as described above.RNA-seq libraries were constructed according to the user manual and sequenced to produce 150-base paired-end reads.The samples were sequenced using the Illumina HiSeq2000(San Diego,CA,USA)at Nanjing Personal Gene Technology Co.,Ltd.(Nanjing,China).Three biological replicates were used for each sample.Clean reads were aligned to the Darmor-bzh genome.Differentially expressed genes (DEGs)were identified by fold change ＞1 and P-value ＜0.05.Gene Ontology (GO) enrichment analysis and visualization of the results was performed with TBtools [65].

3.Results

3.1.ytg mutant showed a delayed-greening phenotype

In compared with 4608,the newly emerged true leaves and cotyledons of the mutant were yellowish and became green with development.Mature leaves partially recovered the green color(Fig.1A,C).Similarly,the buds and siliques showed a delayedgreening phenotype.Consistent with these phenotypes,the chlorophyll content of the newly emerged true leaves of the mutant was significantly lower than that of 4608(Fig.1D).The chlorophyll content of the mutant leaves increased as the leaves turned green,but remained lower than that of 4608(Fig.1E).The chlorophyll content of the top first leaf in the mutant was significantly lower than that of 4608,but that of the top fourth leaf was not significantly different.The vegetative growth of the mutant was stunted and the bolting and flowering of the mutant were also delayed (result not shown).In summary,all green organs showed delayed greening.

3.2.Chloroplast development was impaired in ytg

In 4608,the chloroplasts showed a fusiform shape and had well-structured thylakoid membranes in both the top first and fourth leaves.However,in the mutant,the thylakoid membranes in the chloroplasts of the top first leaf were much less abundant.In contrast,the thylakoid membrane organization and abundance in the chloroplasts of the top fourth leaf of the mutant were comparable to those in 4608 (Fig.1B).These results indicated that chloroplast development was impaired and chloroplast biogenesis was delayed in the ytg mutant.

3.3.Many major traits are influenced in the ytg mutant

Fig.1.Phenotypic characterization of the ytg mutant.(A)Cotyledons,seedlings,and buds of 4608 and ytg mutant.Chl a,chlorophyll a;Chl b,chlorophyll b;Car,carotenoid;Total Chl,Chl a and Chl b. (B) Transmission electron micrographs of chloroplasts from 4608 and ytg mutant.P,plastoglobule;S,starch granule;T,thylakoid membrane.(C)Leaves of 4608 and ytg mutant at five-leaf stage.(D)Pigment contents in the leaves of 4608 and ytg mutant.(E)Chlorophyll content in 4608 and ytg mutant at different leaf ages.Values in (D) and (E) represent mean±SD (n=3).Scale bars,1 cm in (A),see the bottom left corner of each image in (B),and 3 cm in (C).

The relationship between leaf color and agronomic traits was investigated by randomly sampling all mature plants from NIL-4608 and NIL-ytg in the BC4F1populations.The TSW,PH,LMI,SN,BN,and oil contents of NIL-ytg were significantly lower than those of NIL-4608 and the LA content higher(Fig.S1).Thus,chlorophyll deficiency reduced yields and affected quality traits.

3.4.Identification of the YTG locus in rapeseed

The ytg mutant was crossed with 4608.The color of the heterozygous F1plants was normal green.Selfing the F1plants generated 187 F2plants,comprising 142 green-leaf and 45 yellowishleaf plants.The normal-green and chlorophyll-deficient plants in the F2population showed a ratio of 3:1 (χ2=0.04 ＜=3.84).These results indicated that the phenotype conferred by ytg was caused by a single recessive nuclear mutation.

To investigate the molecular mechanisms responsible for the phenotypes observed in ytg,we used map-based cloning to identify the YTG locus.A Brassica 60 K Illumina Infinium SNP array was used to screen the G-and Y-pools from the BC1F1and F2populations.In the BC1F1population,804 SNPs were polymorphic between the two bulks,with 596 being localized on chromosome A02 of B.napus.In the F2population,431 SNPs were polymorphic between the two bulks,of which 300 were localized on chromosome A02 of B.napus (Fig.2A).SNPs were distributed mainly in the 2–12 Mb region,using the coordinates of the Darmor-bzh genome(Fig.2B).Based on the 10-Mb candidate region,SSR markers were designed to fine-map the YTG locus.Polymorphic SSR markers were used to screen respectively 3222 and 12,278 plants of the BC3F1and BC4F1generations.In this manner,the YTG locus was further localized to a 9.9-kb region between markers 02–20 and 02–21 (Fig.2C) that contained three putative open reading frames(ORFs).The three ORFs were sequenced and compared.The 4608 sequences were all identical to the reference genome sequence,but no fragments were amplified in the ytg mutant.Of these three genes,BnaA02g10480D was found to encode a TPR protein homologous to WTG1(AT5G53080),which is required for chloroplast biogenesis [66].We accordingly assigned BnaA02g10480D as the candidate gene,but could not exclude the other two(BnaA02g10470D and BnaA02g10490D) as candidate genes.We introduced 7181 bp of BnaA02g10470D,5683 bp of BnaA02g10480D,and 4596 bp of BnaA02g10490D,containing the transcriptional regulatory elements and full-length coding sequence of each gene,into the ytg mutant.The leaf phenotype of the ytg mutant was fully rescued by the introduction of the genomic fragment BnaA02g10480D,but not by those of BnaA02g10470D or BnaA02g10490D (Figs.3A;S2).Because the leaf phenotype of the ytg mutant changed from yellow to green instead of white to green,we named BnaA02g10480D as BnaA02.YTG1 instead of BnaA02.WTG1.Normal-green and chlorophyll-deficient plants showed a ratio of 3:1 in the T1population of BnaA02.YTG1-transformed line PT1 (χ2=0.796 ＜=3.84).This ratio indicated that a single copy of BnaA02g10480D was contained in the transformed line PT1.BnaA02.YTG1 (BnaA02T0141700ZS) and BnaA02g10490D (BnaA02T0141800ZS) were both highly expressed in transgenic-positive plants,but not detected in transgenicnegative plants,while the BnaA02g10470D (BnaA02T0141600ZS)was almost not expressed in transgenic-positive and negative plants(Table S5;Fig.S6A).We speculated that the ytg phenotype was caused by the deletion of BnaA02.YTG1.According to Darmor-bzh genome information,there are two YTG1 copies in the genome:

BnaA02.YTG1 (BnaA02g10480D) and BnaC02.YTG1 (BnaC02g14610D).These copies are 89.6%and 90.4%identical at the nucleotide and protein levels,suggesting that these genes share similar functions(Figs.S3,S4).The sequences of BnaC02.YTG1 in 4608 and ytg revealed no variation,but a BLAST search against the ZS11 genome sequence revealed a 698 bp insertion in the third exon.Transposon prediction identified the insert as a hAT transposon,which is a DNA transposon(Class II)(Fig.S5).Of all transposable element activities,their inactivation by insertion is the best characterized [67].We accordingly speculated that the insertion of the hAT transposon would affect the transcription of BnaC02.YTG1.The finding that its transcription started at the third exon confirmed this speculation (Fig.S6B).To further investigate the function of BnaC02.YTG1,we constructed an overexpression vector of BnaC02.YTG1 without the hAT transposon insertion from ZS11 to transform into ytg,which rescued the leaf color phenotype (Fig.S7).The results of subcellular localization experiments showed that BnaC02.YTG1 was a chloroplast-localized protein(Fig.S8).These results indicated that BnaC02.YTG1 was also required for early chloroplast growth and development and that the insertion of hAT transposon affected its function.

To further elucidate the functions of BnaA02.YTG1,diverse mutants were generated with the CRISPR/Cas9 system.Genetic transformations of 4608 were performed.Among T0transgenicpositive plants derived from 4608,three plants that were edited only in BnaA02.YTG1 and not BnaC02.YTG1 were identified.They all showed chlorophyll-deficient leaf color phenotypes similar to that of the ytg mutant.CR-a7 harbored biallelic mutations consisting of insertions or deletions(InDels)that caused frameshifts at the sgRNA1 and sgRNA2 loci,while CR-b404 contained indel mutations with frameshifts at the sgRNA1 site.CR-c7 harbored the same homozygous insertion mutations with frameshifts at the sgRNA1 and sgRNA2 sites (Fig.3B).These results showed that BnaA02.YTG1 was responsible for the leaf phenotype in ytg.

3.5.BnaA02.YTG1 was expressed ubiquitously in different tissues and its protein was localized in the chloroplast

We quantified BnaA02.YTG1 expression using qRT-PCR.BnaA02.YTG1 was ubiquitously transcribed in all examined tissues and was expressed mainly in young tissues such as flower buds,siliques,leaves,and cotyledons,with lower expression levels in cauline leaves,stems,and roots of B.napus(Fig.4A).The BnaA02.YTG1 transcription pattern was also monitored by GUS expression driven by the BnaA02.YTG1 promoter in transgenic Arabidopsis.Cotyledons,rosette leaves,young buds,and siliques showed higher levels of GUS expression.Few GUS signals were observed in mature siliques and seeds (Fig.4C).The GUS expression patterns were generally consistent with the levels detected by qRT-PCR.

BnaA02.YTG1 encodes a putative TPR protein of 543 amino acids with a calculated molecular mass of 60.631 kDa,containing six conserved TPR domains identified by PROSITE,https://prosite.expasy.org/ (Fig.3B).Amino acid sequence analysis revealed that it contained a chloroplast signal peptide.For its subcellular localization,the coding sequence of BnaA02.YTG1 was fused with GFP.The fluorescent signals of the BnaA02.YTG1-GFP fusion protein co-localized with chloroplast autofluorescence signals(Fig.4B),suggesting that BnaA02.YTG1 was indeed a chloroplastlocalized protein.

3.6.BnaA02.YTG1 directly interacts with MORFs (component of the plastid RNA editosome) and some chloroplast membrane proteins

Fig.2.Map-based cloning of YTG in rapeseed.(A) Distribution of retained SNPs with reference to the B.napus cv.Darmor-bzh genome.(B) Distribution of SNPs on chromosome A02.Blue arrows indicate the peaks of the SNP distributions in the two populations.(C) Mapping of YTG and candidate gene analysis with reference to the B.napus cv.Darmor-bzh reference genome.The candidate gene is highlighted in red.

TPR domains may bind preferentially to host proteins,and there are few reports concerning BnaA02.YTG1.Owing to its selfactivation,BnaA02.YTG1 was not directly used as bait.We performed yeast two-hybrid screening with BnaA02.YTG1-1 (amino acids:165–543) as bait,and a cDNA library of all ZS11 tissues was used as prey,from which 8 positive clones were obtained.Plasmids from all positive clones were extracted and transformed into E.coli for sequencing.Frameshift mutations,self-activating clones,and duplicated clones were excluded.BnaA02.YTG1-1 interacted with BnaMORF8 and BnaMORF2.BnaA02.YTG1 was chloroplast-localized,as stated above.MORF8/RIP1 and MORF2/RIP2 are required for efficient editing at most Arabidopsis mitochondrial editing sites,and also play an important role in chloroplast editing [23,24].We selected two homologous genes with the highest sequence similarity between B.napus and Arabidopsis(BnaC03.MORF8 and BnaC04.MORF2) to perform directed yeast two-hybrid point-to-point assays.Consistent with the library screening results,BnaA02.YTG1-1 interacted with BnaC03.MORF8 and BnaC04.MORF2 (Fig.5A).To avoid missing chloroplast membrane proteins,we cloned the ORFs of 17 known Arabidopsis chloroplast outer and inner-membrane proteins as prey,and performed yeast two-hybrid analysis using BnaA02.YTG1-1 as bait.BnaA02.YTG1 interacted with BnaA10.Toc33,BnaC09.Toc34,and BnaC01.Tic22-III (Fig.5A).

To verify the yeast two-hybrid results,we conducted SLC assays in tobacco (Nicotiana benthamiana) leaves.In agreement with the yeast two-hybrid results,all combinations of BnaA02.YTG1-nLUC and cLUC-BnaC04.MORF2,BnaA02.YTG1-nLUC and cLUC-BnaC03.MORF8,BnaA02.YTG1-nLUC and cLUC-BnaA10.Toc33,BnaA02.YTG1-nLUC and cLUC-Bna.C09.Toc34,and BnaA02.YTG1-nLUC and cLUC-BnaC01.Tic22-III resulted in strong luciferase signals in tobacco leaves (Fig.5B).

3.7.Editing efficiency of many RNA editing sites was reduced in ytg mutant

Given that MORF8/RIP1 and MORF2/RIP2 play an important role in chloroplast editing [23] and the interaction of BnaA02.YTG1 with MORF8/RIP1 and MORF2/RIP2 were confirmed,we compared the RNA editing rates of HB1 and hb1 by bulk sequencing.

Progressive Mauve was used for multiple alignments among the chloroplast genomes of A.thaliana to B.rapa,B.oleracea,and B.napus.All showed high collinearity (Fig.S9).BLAST alignment showed that the shared identities of the chloroplast genomes of A.thaliana to B.rapa,B.oleracea,and B.napus were 96.39%,96.49%,and 96.53%,respectively (Table S2).Thus,the chloroplast was highly conserved during the evolution from A.thaliana to B.napus.There are 43 plastid RNA editing sites in A.thaliana[68,69].Alignment of the B.napus and A.thaliana chloroplast genomes revealed 29 RNA editing sites in B.napus.Of these,25 sites in hb1 were edited in the same manner as those of HB1.However,the other four editing sites,ndhD-2,ndhF-290,petL-5,and ndhG-50,showed significant decreases in editing rates (Fig.6A,B;Table S3).To detect whether these RNA editing defects were restored as leaves turned green,we examined the level of RNA editing in hb4,finding that the editing rates of the four editing sites were mostly restored (Fig.6B).Because in comparison with bulk sequencing,pyrosequencing is a more sensitive and quantitative method of assessing RNA editing,we reassessed the extent of RNA editing using pyrosequencing.As expected,the results were consistent with bulk sequencing results (Fig.6C).Although the mutation of BnaA02.YTG1 reduced the editing extent of the four editing sites,the editing of some C targets was not affected at all.

Fig.3.Functional identification of BnaA02.YTG1.(A)Phenotype of the ytg mutant,complemented line PT1,and 4608.(B)Targeting of BnaA02.YTG1 and BnaC02.YTG1 with two sgRNAs using the CRISPR/Cas9 system.Sequences of the engineered alleles in the representative plants are shown.Deletions are indicated in black by dashes or sequence gap lengths (bp),and insertions are indicated by blue letters.WT,wild type.Scale bars,3 cm.

Fig.4.Expression pattern of BnaA02.YTG1 and subcellular localization of the BnaA02.YTG1 protein.(A) Expression of BnaA02.YTG1 in various organs,including stem leaves(SA),stems (SB),opened flowers (SC),roots (SE),buds (SF),young siliques (SG),young leaves (S),and cotyledon (SJ).Values represent mean±SD (n=3).(B) Subcellular localization of the BnaA02.YTG1 protein.Chlorophyll autofluorescence is shown in red.The image appears yellow where the signals from GFP and chloroplasts overlap.Scale bars,20 μm.(C) GUS staining in organs of Arabidopsis transformed with the native promoter of BnaA02.YTG1. Scale bars,2 mm.

Fig.5.Yeast two-hybrid(A)and SLC analyses(B)of BnaA02.YTG1 interaction with BnaC04.MORF2,BnaC03.MORF8,BnaA10.Toc33,Bna.C09.Toc34,BnaC9.OZ1,and BnaC01.Tic22-III.BD-53/AD-RecT was the positive control,and BD-Lam/AD-RecT was the negative control.AD,pGADT7/activation domain;Ade,adenine;BD,pGBKT7/binding domain;His,histidine;Lam,human lamin C;Leu,leucine;RecT,recombination of the SV40 large T antigen;SD medium,synthetic defined medium;Trp,tryptophan.

3.8.BnaA02.YTG1 may interact with BnaC9.OZ1 to affect RNA editing

To determine whether BnaA02.YTG1 interacted with editing factors at the ndhD-2,ndhF-290,petL-5,and ndhG-50 sites(Table S3),we employed a series of yeast two-hybrid point-topoint assays.BnaA02.YTG1 interacted with BnaC09.OZ1,which showed the highest identity with OZ1 in Arabidopsis,but not with the other site-specific recognition proteins CRR4,OTP84,OTP82,or ECB2 (Fig.5A).Direct interaction between BnaA02.YTG1 and BnaC09.OZ1 was verified using an SLC assay(Fig.5B).These results provided evidence that BnaA02.YTG1 affects RNA editing via physical interactions with MORFs and OZ1 in chloroplasts.

3.9.The activity of NDH complex and cytochrome b6f complex in ytg may be impaired

The chloroplast NDH complex is involved in cyclic electron flow around photosystem I [64,70].Given that ndhD-2,ndhF-290,and ndhG-50 in hb1 showed significantly decreased editing rates,we identified ytg mutants by their lack of post-illumination increase in chlorophyll fluorescence(Fig.7A)using a modified system of fluorescence imaging [64,71].This transient rise in fluorescence is due to the dark reduction of the plastoquinone pool (PQ),which is dependent on NDH activity.In contrast,the fluorescence rise was undetectable in hb1 (Fig.7A).Thus,the activity of the NDH complex was affected in ytg1.

Fig.6.RNA editing analyses.(A) Comparison of RNA editing in HB1 and hb1 plants by bulk sequencing.(B) Comparison of RNA editing rates in HB1,hb1 and hb4.Values represent mean±SD (n=3).(C) Validation of the editing extent of the four sites in hb1,HB1,and hb4 by pyrosequencing.

Fig.7.Chlorophyll a fluorescence induction analysis.(A)Monitoring NDH activity using chlorophyll fluorescence analysis after turning off AL.The bottom curve indicates a typical trace of chlorophyll fluorescence in HB1.The transient rise in fluorescence ascribed to NDH activity was monitored by chlorophyll fluorimetry.Insets are magnified traces from the boxed area.F0,minimum fluorescence yield;Fm,maximum fluorescence yield.(B)Using fluorescence decay kinetics analysis(Qa_Decay)to measure electron transfer on the donor and acceptor sides of PSII.The slope at a point on the curve represents the electron transfer rate at that point.

PSI and PSII are interconnected via the membrane-bound cytochrome b6f complex and mobile electron carriers,PQ and plastocyanin (PC).Cytochrome b6f complex catalyzes the rate-limiting step of linear electron transport,oxidizing PQ generated by PSII and reducing PC [72].We further investigated how the acceptor and donor side functions of PSII.The activity of electron transfer from QAto QBin hb1 was lower than that of HB1,and the activity of electron transfer from QBto PQ in hb1 was lower than that of HB1 (Fig.7B).These results suggested that electron transfer on both the acceptor and donor sides of PSII in hb1 was impaired.Electrons are transferred to cytochrome b6f after transferring from QBto PQ.We speculated that the decreased electron transfer activity from QBto PQ in hb1 was due to the decreased electron transfer from PQ to cytochrome b6f.This may be a feedback regulation mode in which the ability of cytochrome b6f to accept electrons is reduced.

3.10.Transcriptome analyses of transgenic Brassica napus plants

To investigate the global effect of loss of BnaA02.YTG1 on plant growth and development,we performed transcriptome analysis of HB1,HB4,hb1,and hb4 using RNA-seq.A total of 240 differentially expressed genes (DEGs),including 130 up-regulated and 110 down-regulated genes,were identified in hb1 compared with those in HB1 (Fig.8A,B).Only 25 DEGs were identified in hb4 in comparison with those in HB4 (Table S6).

GO enrichment analysis was performed to identify the functional implications of DEGs induced by the expression of BnaA02.YTG1.Down-regulated genes were significantly enriched for GO terms associated with the chloroplast membrane system(Fig.8E),while up-regulated DEGs were enriched for GO terms associated with biotic stress response (Fig.8D).These results suggested that the expression of BnaA02.YTG1 induced a large number of chloroplast membrane system-associated genes.Because BnaA02.YTG1 was required for chloroplast biogenesis during the early stage of leaf development,we hypothesized that the DEGs found only at the early stage were required for chloroplast development (Fig.8C).We accordingly applied GO enrichment analysis to the 237 DEGs identified at the early stage.Of these,26 DEGs were associated with the photosynthetic membrane system(Table S5).These results indicated that the expression of genes involved in photosynthesis was repressed in ytg.The expression of some plastid genes was also affected (Fig.S10).

4.Discussion

4.1.BnaA02.YTG1 may affect RNA editing in chloroplasts through interactions with plastid RNA editosomes

Studies of cytidine-to-uridine RNA editing in plants have focused on Zea mays,Oryza sativa,and Arabidopsis thaliana [73].With advances in second-generation sequencing,it has been vastly improved to detect and quantify RNA editing on a global scale.The chloroplast genome sequencing information of A.thaliana,B.rapa,B.oleracea,and B.napus has been released.In the present study,considering the high collinearity between A.thaliana and B.napus in the chloroplast genome,we used the homologous alignment method to predict 29 editing sites in B.napus.This study may shed light on the development of plastids,RNA editing,and crops photosynthesis.

TPR proteins have been reported [38,41,74–77] to be involved in the biogenesis of the photosynthetic apparatus,including chloroplast import,chloroplast gene expression,assembly and stabilization of photosynthetic complexes,and chlorophyll biosynthesis.For RNA editing,the sequence-specific recognition of these editing sites is performed by a large family of PPR proteins.However,little research has focused on TPR proteins in RNA editing,and only WTG1 has been reported in Arabidopsis [66].

We cloned BnaA02.YTG1,which encodes a TPR protein.We confirmed that BnaA02.YTG1 interacts with the known editing factors MORF2,MORF8,and OZ1(Fig.5).The editing of ndhD-2,ndhF-290,petL-5,and ndhG-50 transcripts was affected in ytg young leaves but rescued in mature leaves.It has been reported [10,23,24] that MORF2,MORF8,and OZ1 were involved in the editing of ndhD-2,ndhF-290,and petL-5;MORF2 and OZ1 were involved in the editing of ndhG-5.Our results are consistent with those of previous studies.However,it seems that RNA editing events have no effect on the RNA abundance of these four genes (Fig.S10).These results suggested that BnaA02.YTG1 might affect RNA editing and was not directly involved in the identification of editing sites.

Based on the above results,we speculate that BnaA02.YTG1 participated in RNA editing via interactions with MORFs and OZ1,which can stabilize the editosome.If so,BnaA02.YTG1 would enrich our knowledge of the chloroplast RNA editing system,given the few reports concerning the involvement of TPR in the RNA editing process.Chlorophyll fluorescence analysis revealed that the activity of the NDH complex in ytg was damaged,and suggested that the function of cytochrome b6f might also be affected.Mutants with complete disruption of NDH complex or knock-out of petL showed no visible phenotypes [71,72,78].Thus BnaA02.YTG1 may regulate other metabolic processes of the chloroplast via its protein–protein interaction domains (TPR),such as protein translocation by interacting with Toc33,Toc34,and Tic22 (Fig.5).If so,the impaired RNA editing may be a phenomenon of secondary effects.

4.2.BnaA02.YTG1 is a TPR protein required for early chloroplast development in B.napus

Fig.8.RNA-seq analysis of positive and negative transgenic plants of transgenic line PT1.(A)Number of differentially expressed transcripts.(B)Heatmap of DEGs.(C)Venn diagram showing the unique and shared relationships of DEGs in the top first and fourth leaves.DEGs found in only the top first leaves were assigned as DEGs at the early stage.(D) Top 30 GO terms for DEGs up-regulated in hb1.(E) Top 19 GO terms for DEGs down-regulated in hb1.

The regulatory mechanisms of chloroplast biogenesis and development are complex,and abnormal chloroplast development usually affects leaf coloration and seedlings viability.There have been many reports of mutants with delayed greening phenotypes.In the dg1 mutant,young leaves were very pale at first and then gradually greened,finally resembling those of wild-type plants[79].The ecb2-2 mutant,harboring a point mutation in the PPR motif of the AtECB2 protein,showed a delayed greening phenotype[80].In the dg238 mutant,young leaves exhibited a chlorotic phenotype but were stunted as the plant developed[81].The mutation of BnaC07.HO1 (Heme oxygenase 1) showed a yellow-green leaf phenotype at the seedling stage and became green,similar to the wild type,at the later stage [82].The gry340 mutant displayed a yellow leaf phenotype before reaching the three-leaf stage,but gradually turned green with development and recovered a normal green color after the nine-leaf stage [83].In the present study,the ytg mutant displayed a delayed greening phenotype(Fig.1),and we cloned BnaA02.YTG1 encoding a TPR-containing protein.

The TSW,PH,LMI,SN,and BN were significantly lower in NILytg than in NIL-4608.The LA content of NIL-ytg was significantly higher and the oil content lower than those of NIL-4608,possibly owing to abnormal lipid metabolism (Fig.S1).We infer that BnaA02.YTG1 increases the yield from the increase in biomass.

The greening of the cotyledons,true leaves,buds,and siliques of the ytg phenotype was reduced,supporting the notion that BnaA02.YTG1 functions mainly in young tissues.For example,the young leaves of the ytg mutant were initially yellowish and gradually greened during development,indicating that BnaA02.YTG1 functions in the early stages of chloroplast development.Unlike the above mutants that showed delayed greening,the ytg mutant showed a chlorotic phenotype with defective chloroplasts during the early stage and then became almost normal,but the chlorophyll content of the mature leaves of the ytg mutant was still lower than that of 4608.We initially speculated that the ytg mutant displayed genetic dosage effects,meaning that some homologous proteins or factors with similar functions show functional redundancy [5,84].Protein alignments revealed another protein(BnaC02.YTG1) in B.napus with homology to BnaA02.YTG1.However,the nucleotide sequence of BnaC02.YTG1 harbored a hAT transposon insertion in the third exon of the gene in both parents(Fig.S5).The RNA-seq results showed that the transcription of the gene started at the third exon(Fig.S5).Therefore,it was expressed in the fragments (Fig.S6B).The phenotype of the ytg mutant was restored by the overexpression of BnaC02.YTG1 without hAT insertion.Thus,BnaC02.YTG1 in ytg had no effect on leaf color regulation,leading us to reject the speculation.Another possible explanation is that chloroplast development is also determined by the balance between plastid gene expression and the degradation of their proteins [85–87].In the early stages of chloroplast development,proteins were accumulated by fast protein synthesis whose rate was far greater than that of degradation.Normal chloroplasts would be produced and chlorophyll content increased when the accumulation of some proteins exceeded a threshold.As a result,the leaves turned green.In recent years,there have been some reports on delayed-greening mutants.In A.thaliana,for example,the dg1 mutant showed a leaf albino phenotype,and the ys1 mutant showed a yellow leaf phenotype,both of which were caused by delayed chloroplast development [79,88].Further research [89] revealed that DG1 encodes a chloroplast-localized PPR protein that can interact with MORF2 and other proteins to participate in RNA editing at accD-1568,ndhD-2,and petL-5.Similarly,YS1,encoding a DYW protein,is required for RNA editing of rpoB[88].To date,only a few genes associated with delayed greening or chloroplast development have been cloned and the mechanism of this process remains elusive.

4.3.BnaA02.YTG1 interacts with TOC/TIC proteins and may affect protein transport

The chloroplast genome is small and chloroplast development is controlled mainly by nuclear genes,given that more than 95% of chloroplast proteins are nuclear-encoded [90].Identifying and cloning such nuclear genes may clarify the regulatory mechanism of chloroplast development.Proteins localized in the chloroplast are usually first synthesized as precursors with a transit peptide(TP),and then transferred to the surface of the chloroplast with cytosolic chaperones [91].This process requires the TOC/TIC complex to import thousands of cytoplasmically synthesized preproteins [91].Toc33 and Toc34 residing in the outer envelope membrane of chloroplasts exhibit some preferences for binding to transit peptides [92–94],and may be involved in the import of distinct preproteins.Toc33 is preferentially involved in the import of photosynthetic proteins,and Toc34 is involved in the import of non-photosynthetic chloroplast proteins [94].The Tic22 protein is the only known component in the intermembrane space that has been proposed as being involved in protein transport into chloroplasts [95].In this study,we showed using yeast two-hybrid and SLC assays that BnaA02.YTG1 could interact with Toc33,Toc34,and Tic22.As mentioned above,the chlorophyll content of the mature leaves was still lower than that of 4608,although the editing rates of ndhD-2,ndhF-290,petL-5,and ndhG-50 was mostly restored.Based on these results,we propose that the loss of BnaA02.YTG1 impaired normal protein transport from the nucleus to the chloroplast.If this is true,BnaA02.YTG1 would add another level of complexity to proteins transport.

5.Data availability

All data generated and analyzed in this study are available upon request.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

CRediT authorship contribution statement

Haiyan Zhang:Writing -original draft,Conceptualization,Data curation,Validation,and Formal analysis.Xiaoting Li:Investigation.Yebitao Yang:Investigation.Kaining Hu:Formal analysis.Xianming Zhou:Investigation.Jing Wen:Supervision.Bin Yi:Supervision.Jinxiong Shen:Supervision.Chaozhi Ma:Supervision.Tingdong Fu:Supervision.Jinxing Tu:Writing-Review&Editing,Supervision,Project administration,and Funding acquisition.

Acknowledgments

We thank Dr.Wei Hua (Oil Crops Research Institute of the Chinese Academy of Agricultural Sciences)for providing the all-tissues cDNA library of ZS11.This research was supported by the National Key Research and Development Program of China(2016YFD0100305).

Appendix A.Supplementary data

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2021.06.010.