Long Zhng,Ynzhou Qi, Mingming Wu,Lei Zho,Zhicho Zho,Cilin Lei,Yunyun Ho, Xiowen Yu, Yinglun Sun, Xin Zhng, Xiuping Guo, Yulong Ren,,Jinmin Wn,,
a National Key Facility for Crop Gene Resources and Genetic Improvement, Institute of Crop Sciences, Chinese Academy of Agricultural Sciences, Beijing 100081, China
b State Key Laboratory for Crop Genetics and Germplasm Enhancement, Jiangsu Plant Gene Engineering Research Center,Nanjing Agricultural University, Nanjing 210095, Jiangsu, China
Keywords:
ABSTRACT
Rice accumulates large amounts of starch in the endosperm that serves as a carbon source for the plant, human and livestock.Normal endosperm development contributes to grain quality and yield.Rice endosperm has been used as an ideal model system for dissecting the molecular machinery of starch synthesis because of its rich genetic resources and small genome.Many floury endosperm mutants and the causative genes have been identified.Some of them encode enzymes involved in starch synthesis itself, including ADP-Glucose Pyrophosphorylase Large Subunit 2 (OsAPL2), starch synthesis enzyme-encoding gene OsSSIIIa/FLO5, and starch branching enzyme-encoding gene BEIIb/Amylose-extender(AE) [1-3].Several transcription factors regulate the expression of starch synthesis-related genes, including bHLH144,OsbZIP58, and RSR1 [4-6].Interestingly, mutations in genes involved in rice storage protein trafficking cause floury endosperm phenotypes,including GLUTELIN PRECURSOR ACCUMULATION1 (GPA1)/Rab5a, GPA2/VPS9a, GPA3, GPA4/GOT1B, and GPA5 [7-11].Together, these studies suggest that endosperm development involved many biological processes.Novel factors required for endosperm development remain to be explored.
Pentatricopeptide repeat (PPR) proteins are nuclearencoded proteins characterized by tandem arrays of a degenerate 31-36 amino acid repeat (PPR motif) [12].PPR proteins generally contain 2-26 PPR motifs and are classified into two subgroups:the classical P-type PPR members composing only of 35-amino acid motifs,and PLS-type PPR members possessing short(S),canonical,and long(L)motifs[13].Most PPR proteins are predicted to be localized to mitochondria and/or chloroplasts, and are required for the post-transcriptional processing of RNA.Among them,P-type members participate in RNA cis- and trans-splicing, RNA stabilization, and translational activation, whereas PLS-type members mostly participate in RNA editing [12-14].
Group II introns commonly exist in mitochondrial genome of eukaryotes and encode reverse transcriptase/maturase with catalytic function, which undergo self-splicing [15].In higher plants,most of group II introns belong to genes encoding subunits of electron transfer chain complex I and require PPR proteins to perform either cis-splicing or trans-splicing[16].So far, several P-type PPR proteins have been demonstrated to be involved in intron splicing of mitochondrial transcripts and seed development, such as ABO5, MTL1, TANG2,OTP439, OTP43, and SLO3 in Arabidopsis thaliana [17-21], as well as DEK2, DEK35, DEK37, EMP8, EMP10, and EMP16 in maize [22-27].As PPR proteins are a large family in plants,the underlying splicing mechanisms of numerous unidentified P-type PPR proteins remain poorly understood.
The rice genome contains 491 PPR genes,of which 246 PPR genes accounting for about 50% belong to P-type, and 245 genes belong to PLS-type [28].Currently, only four PPR proteins have been identified to be involved in rice mitochondrial function and endosperm development.The mutation in PLStype PPR protein OGR1 causes a defect in RNA editing at seven specific sites on five mitochondrial genes, inlcuding nad2,nad4, cox2, cox3, and ccmC [29].P-type PPR proteins, such as OsNPPR1, OsNPPR3, and FLO10, play an important role in the splicing of mitochondrial genes [30-32].Compared with the marked progress in Arabidopsis and maize, more P-type PPR proteins involved in intron splicing and endosperm development remain to be identified in rice.
In this study, we identified a novel floury endosperm mutant ppr5.Map-based cloning and complementation test demonstrated that PPR5 encodes a new P-type PPR protein that is localized to mitochondria.Loss of PPR5 function significantly affected the splicing efficiency of nad4 intron 3, compromising the assembly of mitochondrial complex I and activity of NADH dehydrogenase.Our results suggest that PPR5 is involved in mitochondrial function and endosperm development in rice by affecting the splicing of mitochondrial nad4 intron 3.
The ppr5 mutant was obtained from an ethyl methanesulfonate (EMS)-induced mutant pool of japonica rice cultivar Kitaake.As ppr5(-/-) homozygous mutant plants were seedling lethal, the vitreous brown rice kernels (+/+and+/-) was planted for reproduction.A plant which produced seeds with both vitreous and floury endosperm was the ppr5(+/-)heterozygous plant.A ppr5(+/-)heterozygous plant was hybridized with an indica variety Dular to produce an F2population for map-based cloning.To identify developing endosperm with the mutant phenotype, half-grain assays were performed during grain filling.Briefly,the distal(embryoless)half from ppr5(+/-) heterozygous plant was ground into flour for DNA extraction.Primers 5′-CCCAATAGAGGTGTACGGCT-3′and 5′-TCCTCACACCATCCTTGCTT-3′were used for amplification and sequencing of mutation site.The remaining half(with embryo) defective in the PPR5 gene was used for subsequent experiments.All plants were grown in a paddy field of Institute of Crop Sciences, CAAS in Beijing.
For the cross-section observations, brown rice was transversely broken by two forceps, coated with gold, and examined with a Hitachi S-4800 scanning electron microscope.Semi-thin sections of endosperms at 9 and 15 days after fertilization (DAF) were prepared according to a previously described process [33].The sections were counter stained with safranin O and iodine solution, and viewed with an Olympus BX53 light microscope.Given the phenotype difference and technical difficulty to prepare resin section, the grains at 9 DAF were further subjected to ultrastructure observation, RNA extraction, and ATP determination.For ultrastructural observations of mitochondria, endosperms(approximately 2 mm in thickness) at 9 DAF were fixed in 2.5% glutaraldehyde for 24 h, dehydrated and embedded in LR White resin.The samples (approximately 80 nm in thickness) were sectioned using a Leica UC7 ultramicrotome.The sections were examined with a Hitachi H-7650 transmission electron microscope.
Dry brown rice was ground and passed through a 100 μm mesh sieve to obtain flour.Total starch content was measured with a Megazyme Total Starch Assay kit (K-TSTA).Protein content (N×5.95) was measured using an Elementar Analysensysteme Gmbh Vario EL cube CHN-Nitrogen analyzer(Elementar, Germany).The amylose content was assessed by the iodine colorimetric method [33].The fine structure of amylopectin was measured following a fluorophore-assisted capillary electrophoresis (FACE) method [34].The pasting properties of brown rice flour were measured using a rapid visco analyzer (RVA, 3D, Newport Scientific, Narrabeen,NSW, Australia).Briefly, a 3 g sample was mixed with 30 mL of H2O, held at 50 °C for 1 min, progressively heated to 95 °C at 12 °C min-1, held at 95 °C for 2.5 min, cooled to 50 °C at 12 °C min-1, and held at 50 °C for 1.4 min.The ATP contents in developing endosperm of WT and ppr5 mutant at 9 DAF were measured with an ATP assay kit (Beyotime, Jiangsu).
F1plants from cross ppr5(+/-)/Dular(+/+) were of two types,either produced only vitreous brown rice kernels (+/+) or both vitreous (+/+and+/-) and floury (-/-) kernels.A population of 852 floury kernels from the latter were germinated and DNA was extracted from the ultimately lethal F2seedlings for gene mapping and map-based cloning using 170 genome-wide polymorphic insertion/deletion(InDel)markers.The sequences of new markers for fine mapping are listed in Table S1.
For genetic complementation, a 4.6-kb genomic fragment containing the entire PPR5 coding region,a 2120 bp upstream regulatory region, and a 1083 bp downstream regulatory sequence was cloned by PCR and inserted into the binary vector pCAMBIA1305 to generate the transformation cassette pCAMBIA1305-PPR5.The PPR5 CDS sequence of WT without the stop codon was cloned in-frame in front of the green fluorescent protein (GFP) in the pCAMBIA1305-GFP vector under the control of the UBIQUITIN1 promoter to create UBI:PPR5-GFP.As the ppr5 floury kernels can induce embryogenic callus,recombinant plasmids pCAMBIA1305-PPR5 and UBI:PPR5-GFP were introduced into calli by Agrobacterium strain EHA105.Twenty-five independent transgenic lines carrying pCAMBIA1305-PPR5 and 36 transgenic lines carrying UBI:PPR5-GFP were obtained.
Wild-type PPR51-150CDS sequences were amplified and inserted into the pCAMBIA1305-GFP vector under the control of the CaMV 35S promoter to produce a 35S:PPR51-150-GFP plasmid.For transient expression, the plasmid was transfected into N.benthamiana leaf epidermal cells as described previously [33].After 2 days, the N.benthamiana leaves were injected with the mitochondrial dye Mitotracker Red(Invitrogen) for 30 mins before imaging.Fluorescent signals were observed using a Zeiss LSM710 confocal laser microscope.UBI:PPR5-GFP was transformed into ppr5 mutant and the roots of T1transgenic lines were analyzed using a Zeiss LSM710 confocal laser microscope.
Total RNA for PPR5 expression pattern analysis was extracted from WT roots, stems, leaf sheaths, leaves, panicles, and the developing seeds at 3-21 DAF using an RNA Prep Pure Plant kit (Tiangen Co., Beijing, China).First-strand cDNA was reverse transcribed by priming with oligo (dT18) based on a HiScript 1st Strand cDNA Synthesis kit (Vazyme, Nanjing, China).For mitochondrial and AOX gene expression analyses, total RNA was extracted from 9 DAF endosperms of WT and ppr5 mutant, and reverse transcribed using random hexamer primers.Specific primers for RNA splicing analysis were designed for intron-exon (unspliced forms) and exon-exon (spliced forms) links for each gene.Quantitative RT-PCR was performed on a CFX Connect real-time PCR system (Bio-Rad, California, USA) using an AceQ qPCR SYBR Green Master Mix Kit (Vazyme).RT-PCR cycles were adjusted between 24 and 35 depending on the transcript level of the corresponding gene.The rice Actin gene was used as an endogenous control.The RT-PCR primers used to analyze 33 mitochondrial genes expression have been documented elsewhere [31], and all the other primers are listed in Table S1.
Mitochondria were isolated from endosperm at 9 DAF using the method of Wu [31].Briefly, approximately 1 g of endosperm was ground in 5 mL of extraction buffer consisting of 75 mmol L-1MOPS-KOH (pH 7.6), 0.6 mol L-1sucrose, 4 mmol L-1EDTA, 0.2% polyvinylpyrrolidone 40, 8 mmol L-1Cys, and 0.2% bovine serum albumin.The homogenate was filtered with a two-layer miracloth and centrifuged in a gra-dient (600, 1300, and 2100×g at 4 ° C, 4 min each) to remove fiber and debris.The supernatant was centrifuged at 22,000×g for 20 min, and resuspended in 400 μL of buffer consisting of 10 mmol L-1MOPS-KOH (pH 7.2) and 0.3 mol L-1sucrose.After adding 800 μL of distilled water, the suspension was centrifuged again at 22,000×g for 20 min to obtain the mitochondria enriched pellet.The pellet was further resuspended in membrane protein extraction buffer consisting of 50 mmol L-1imidazde-HCl (pH 7.0), 500 mmol L-16-aminohexanoic acid, 1 mmol L-1EDTA, and 1% Triton X-100, incubated on ice for 30 min and centrifuged at 22,000×g, 4 °C for 10 min.The supernatant was collected and added to Coomassie Blue G-250 buffer consisting of 5% Coomassie blue G-250, 20 mmol L-1imidazde-HCl (pH 7.0), and 500 mmol L-16-aminohexanoic acid.20 μL of mitochondrial protein were loaded onto a 3% to 12% gradient gel to analyze NADH complex I activity in detection buffer (0.02 mol L-1NADH, 1 mmol L-1NBT, and 0.05 mol L-1MOPS-KOH (pH 7.6)) and stained by Coomassie Blue R-250.
Total RNA from endosperm at 9 DAF was extracted and circularized with T4 RNA Ligase (New England Biolabs, Ipswich, MA) following the user’s manual.The products were extracted by phenol/chloroform to synthesize the first-strand cDNA by HiScript 1st Strand cDNA Synthesis kit (Vazyme) with specific primer nad4 exon1-3-cRT (Table S1).The fragments from circular RT-PCR were amplified with specific primers nad4 exon1-3-cF and nad4 exon1-3-cR (Table S1), and cloned into the pEASY-Blunt Simple Cloning Vector (TransGen, Beijing) for sequencing.
Proteins from developing endosperms at 9 DAF were extracted in extraction buffer consisting of 50 mmol L-1Tris-HCl (pH 8.0), 0.25 mol L-1sucrose, 2 mmol L-1DTT, 2 mmol L-1EDTA, 1 mmol L-1phenylmethylsulphonyl fluoride, and protease inhibitor cocktail (Roche).Proteins were transferred to PVDF membranes (0.45 μm, Millipore) after SDS-PAGE.The membranes were incubated with antibodies against AOX and with monoclonal antibody against EF-1α and visual-ized using a chemiluminescence analyzer (ChemiDocTMX-RS, BioRad).
The ppr5 mutant was identified in an ethyl methane sulfonate (EMS)-treated mutant pool of japonica rice cultivar Kitaake.Mature grains of ppr5 with hulls were phenotypically similar to those of WT (Fig.1A), but the dehulled brown rice of ppr5 displayed floury endosperm (Fig.1B).Cross sections of the brown rice kernels showed that the endosperm of ppr5 was floury-white and shrunken (Fig.1C, D).Scanning electron microscopy (SEM) indicated that the compound starch grains in endosperm cells of ppr5 were loosely packed, spherical, and small, in contrast to the densely packed, irregular polyhedral starch grains in WT (Fig.1E, F).Grain length, width, and thickness of ppr5 brown rice were significantly reduced (Fig.1G-I).Correspondingly, the 1000-brown rice grain weight of ppr5 (11.06 g) was only 51% of WT (21.68 g) (Fig.1J).Starch, protein, and amylose contents were significantly lower than those in the WT (Fig.2A–C).Amylopectin fine structure analysis showed proportionate degrees of polymerization (DP) in the range 6-8 and 16-55 were increased in ppr5, whereas the proportion in the range 9-15 was significantly decreased relative to the WT (Fig.2D).Pasting properties of WT and ppr5 flours determined by a RVA are presented in Fig.2E, and the pasting parameters are given in Table S2.ppr5 flour had significantly lower peak viscosity, hot viscosity, breakdown viscosity, and final viscosity than the WT.Moreover, homozygous ppr5 seedlings showed retarded growth and failed to develop into adult plants (Fig.S1).Thus, the homozygous ppr5 mutant was defective in endosperm development and vegetative growth, but heterozygous plants with normal growth characteristics produced both normal vitreous and floury brown rice kernels (Fig.S2).
Semi-thin sections of developing endosperms of WT and ppr5 at 9 and 15 DAF were prepared to observe the starch grain morphologies.Iodine staining showed that the WT produced compound starch grains in endosperm cells (Fig.3A, a1, C, c1), whereas cells in ppr5 contained numerous small and single starch grains (Fig.3B, b1, D, d1).These observations suggested that the ppr5 mutation affected the formation of normal compound starch grains in the endosperm.
The aleurone cell and ovular vascular trace are considered to function in nutrient transport into the endosperm [31,33].Saffron O staining showed that the WT had two layers darkly stained aleurone cells, whereas the aleurone cells in ppr5 were lightly stained and formed more layers than the WT at 9 DAF (Fig.3A, B), although this was not so evident as that at 15 DAF.However, the cell contents appeared to remain sparse (Fig.3C, D).As shown in Fig.S3, the ovular vascular traces were largely comparable between developing endosperms in WT and ppr5 at 9 DAF, but the aleurone cells near the nucellar projection were lightly stained.These results indicated that the ppr5 mutation obstructed the transport of nutrients into the grain, most likely due to the disrupted development of the aleurone layer.
A map-based cloning strategy was used to isolate the candidate gene responsible for the ppr5 mutant phenotype.An F2mapping population was created by crossing ppr5(+/-)individuals with the indica rice cultivar Dular.A total of 852 floury kernels were selected and germinated, and DNA was extracted from the seedlings for gene mapping.The PPR5 locus was first mapped between InDel markers ZL5-3 and ZL5-7 on the short arm of chromosome 5.Subsequently, the locus was delimited to an 85-kb region between the InDel markers Z5-14 and Z5-18 on BAC clone OJ1430B02 that contained 10 open reading frames (ORFs) (http://rice.plantbiology.msu.edu/; Fig.4A).Sequencing analysis of all ORFs revealed a single base (G) deletion in the ORF6(Os05g0207200) at position 421 bp starting from the ATG start codon(Fig.4B,C;Fig.S4).This mutation formed a new premature stop codon which resulted in a truncated ppr5 protein composed of the N-terminal 160 amino acid residues of WT PPR5 protein (Fig.S5).
To confirm that the single base deletion in Os05g0207200 was responsible for the mutant phenotype, a 4.6-kb genomic fragment containing the transcriptional regulation elements and the full-length coding sequence of PPR5 was introduced into the ppr5 mutant.Positive transgenic lines were selected with recombined vector-specific primers(Fig.4D).Both brown rice morphology and starch grain phenotype were recovered to WT levels (Fig.4E, F).These results confirmed that Os05g0207200 indeed represents PPR5.
Sequence analysis showed that PPR5 comprised a single exon and encoded a protein with 467 amino acid residues(Fig.5A).Database searches using Pfam [35]revealed that the PPR5 protein contained 7 PPR motifs with a variable degree of conservation and belonged to the P subfamily(Fig.5A, B).The single base deletion in the ppr5 allele resulted in loss of all PPR motifs (Fig.5A).BLAST searches in NCBI showed that PPR5 is a single-copy gene in the rice genome and its homologous genes widely exist in green algae, bryophytes, pteridophytes, and higher plants(Fig.S6).In summary, PPR5 represents a new member of the P-type PPR protein family in rice.
Quantitative RT-PCR assay was performed using tissues from several organs to determine the expression pattern of PPR5.Transcripts of PPR5 were detected in roots,stems,leaves,leaf sheaths, young panicles, and developing endosperm.The highest and the lowest levels of expression were in the leaves and leaf sheaths, respectively (Fig.5C).In addition, PPR5 maintained a high expression at the early and middle developmental stages in endosperm,and then gradually decreased(Fig.5C).Consistent with the general expression pattern,PPR5 plays an important role in seedling growth and endosperm development.
Fig.2–Properties of brown rice and amylopectin chain length distribution of flour in ppr5 compared with WT.(A)Starch, (B) Protein, and (C) Amylose contents; (D)Amylopectin chain length distributions; (E) Rapid viscosity profiles of flours.Data are shown as means ± SD (Student’s t-tests, **, P<0.01).
PPR proteins are mainly targeted to chloroplasts or mitochondria, and occasionally to the nucleus [12,30].To determine the subcellular localization of PPR5, a fusion vector of PPR5 fused with GFP was constructed and transiently expressed in Nicotiana benthamiana leaves.Free GFP fluorescence was distributed evenly in the cytoplasm and nuclei,whereas PPR51-150-GFP fluorescence was detected in small dots and co-localized with the red signals from the mitochondrial dye Mitotracker Red (Fig.5D).To verify the subcellular localization of PPR5 in vivo, a fusion vector of full-length PPR5 fused with GFP under the control of the maize UBIQUITIN1 promoter was constructed and introduced into the ppr5 mutant.The brown rice grain morphology in the transgenic rice line was well recovered (Fig.S7A), indicating that PPR5-GFP was biologically functional in vivo.PPR5-GFP in transgenic root tip cells showed a similar localization pattern(small dots) to that observed in the N.benthamiana leaves(Fig.S7B).These data indicated that PPR5 is a mitochondrialocalized protein.
The rice mitochondrial genome contains 56 genes, including 34 genes encoding 33 known electron transport chain proteins.Previous studies showed that P-type PPR proteins are involved in mitochondrial RNA metabolism, including cleavage, splicing, stabilization, and translational activation[12,14].The above amino acid sequence analysis and subcellular localization experiments showed that PPR5 is a mitochondrion-targeted P-type PPR protein.Therefore, transcripts of 33 genes (9 mitochondrial intron-containing genes and 24 other protein-coding genes)in developing endosperms from the WT and ppr5 at 9 DAF were amplified by RT-PCR using specific primers (Fig.6A, B).Among 33 genes, only the mature nad4 transcript was almost undetectable in ppr5(Fig.6A, B).These results indicated that PPR5 was involved in regulation of nad4 expression in developing endosperm.
Rice mitochondria have 23 group II introns, including 3 introns in the nad4 sequence.We designed 23 specific primers across adjacent exons to examine the splicing efficiencies of all 23 mitochondrial group II introns in ppr5 using qRT-PCR.The results showed a significant decrease in splicing efficiency of nad4 intron 3 in ppr5(Fig.6C).There were also slight reductions in splicing efficiency of nad4 introns 1 and 2(Fig.6C).These results suggested that splicing efficiency of nad4 intron 3 was reduced in the ppr5 mutant.
Transcription of mature nad4 mRNA in rice involves three cis-splicing intron events [36].To verify that the marked reduction in nad4 in ppr5 was caused by improper cissplicing of nad4 intron 3, three specific primers across adjacent exons were designed to examine the nad4 precursor transcript.RT-PCR analysis showed that introns 1 and 2 were correctly spliced,whereas the cis-spliced transcript for exons 3-4 (exon 3 + exon 4) were only weakly detected in ppr5,although the precursor of intron 3 was detected(Fig.7A).Similarly, circular RT-PCR showed that ppr5 accumulated increased levels of unassembled precursor and decreased levels of mature nad4 (Fig.7B).Further sequencing analyses of the amplified products revealed that the end of WT nad4 mature RNA was similar to that of intron 3 precursor in ppr5(Fig.7C, Fig.S8), indicating that the mutation of PPR5 does not affect the stability of this precursor.These results suggested that PPR5 is required for the cis-splicing of mitochondrial nad4 intron 3 and formation of the mature nad4 transcript.
nad4 encodes a central subunit of NADH dehydrogenase in mitochondrial complex I.To test whether the nad4 defect affects the assembly of complex I,we analyzed the mitochondrial complex by blue native polyacrylamide gel electrophoresis(BN-PAGE).As shown in Fig.8A,complex I in ppr5 was significantly reduced compared with the WT.In addition, in-gel NADH-dehydrogenase activity staining analysis showed that the activity of complex I exhibited a dramatic decrease in ppr5 (Fig.8B).Since ATP production in mitochondria requires electron transfer through complex I, we measured the ATP content in developing grains at 9 DAF.As shown in Fig.8C,the ATP content in ppr5 was about one-third of that in WT.These results suggested that the splicing defect in nad4 intron 3 caused a functional reduction in the electron transport chain in complex I.
Fig.3–Abnormal compound starch grains and aleurone layer in ppr5 endosperm.(A, B) Safranin O and iodine staining showing the developmental endosperm of WT(A)and ppr5(B)at 9 DAF.(a1,b1)represent the magnified regions indicated by red outline in WT and ppr5 endosperm.(C,D)Safranin O and iodine staining showing the developmental endosperm of WT(C)and ppr5(D)at 15 DAF.(c1,d1)represent the magnified regions indicated by red outline in WT(C)and ppr5(D)endosperm,respectively.Scale bars, 50 μm.Al, aleurone layer; EN, endosperm.
Endosperms from WT and ppr5 at 9 DAF were observed by transmission electron microscopy (TEM) to detect effects of the ppr5 mutation on mitochondrial structure.Mitochondria in WT endosperm displayed distinct inner envelope cristae surrounded by a dense matrix, whereas those in ppr5 were vague in outline and lacked typical cristae (Fig.8D).These results suggested that the PPR5 mutation led to the severely impaired mitochondrial function.
Alternative oxidases (AOXs) can be activated in plants when the electron transport chain is disrupted in the cytochrome c pathway [37].To verify whether the disruption in complex I activated an alternative respiratory pathway in ppr5, we examined the expression of AOX genes (AOX1A,AOX1B, and AOX1C) at the transcriptional level.qRT-PCR showed that both AOX1A and AOX1C were markedly upregulated in ppr5 compared to the WT; AOX1B was slightly up-regulated in ppr5 (Fig.8E).Immunoblotting analysis showed that AOX proteins were also dramatically increased in ppr5 (Fig.8F).These data strongly suggested that the PPR5 mutation altered the assembly and activity of complex I,leading to up-regulated expression of genes in the alternative respiratory pathway.
Fig.4– Map-based cloning of the PPR5 locus and genetic complementation of the ppr5 mutant.(A)Fine mapping of the PPR5 locus (red arrow).The locus was mapped to an 85-kb region between markers Z5-14 and Z5-18 on the short arm of chromosome 5 that contains 10 open reading frames (ORFs).(B) A single base (G) was deleted in ORF6 (Os05g0207200) at position 421 bp from the ATG start codon.(C) Molecular identification of the single base deletion by a CAPS marker.The left and right panels showed the CAPS analysis associated with WT and ppr5.M, marker.(D) Molecular identification of transgenic plants by PCR amplification.M,marker.(E,F)Complementation of the ppr5 mutation restored to normal brown rice appearance (E) and starch grain morphology (F).Scale bars, 10 μm.
Fig.5–Expression pattern of PPR5 and subcellular localization of PPR5.(A)PPR5 protein contained 7 PPR motifs,whereas the mutated ppr5 protein putatively had no PPR motifs.(B) Amino acid sequences of 7 PPR motifs of PPR5.Amino acids fully or semiconserved are shaded black and gray, respectively.(C) Expression pattern of PPR5 in various rice tissues and in the endosperm at different developmental stages.R,root;S,stem;L,leaf;LS,leaf sheath;P,panicle.(D)Subcellular localization of PPR5 in N.benthamiana leaves.Panels display transient expression of free GFP and PPR51-150-GFP (green), Mitotracker localization (red), chlorophyll autofluorescence (blue), and merged images.Scale bars, 10 μm.
Fig.6–Loss of mature nad4 transcript and splicing deficiency of nad4 intron 3 in ppr5 endosperm.(A, B) RT-PCR analyses of expression of 9 intron-containing (A) and other protein-encoding (B) genes in the mitochondrial genome in WT and ppr5.Actin was used as a sample loading control.U,unspliced;S,spliced.(C)Quantitative RT-PCR analysis of 23 group II introns of mitochondrial genes in WT and ppr5.Primers spanning adjacent exons and introns were used to measure differences in splicing efficiency.Data are shown as means ± SD from three biological replicates.
Rice endosperm has been a preferred system to explore regulatory mechanisms of starch and/or protein synthesis [38,39].Defects in endosperm development generally affect starch and/or protein accumulation leading to loss in grain weight and/or changes in amylose: amylopectin ratio.Previous studies have indicated that the accumulation of starch and protein maintains a balance during endosperm development, when one increases, the other decreases.For example, starch synthesis in flo6, flo8, gif2, and pdil1-1 mutants is disordered, and starch content decreases while the protein content increases [34,40-42].In this study, we isolated and identified a rice floury endosperm ppr5 mutant that exhibited several distinct features.As the starch and protein contents were significantly reduced, the grain weight of ppr5 was severely reduced compared to WT (Figs.1J and 2A, B).The amylose content and amylopectin fine structure also differed between WT and ppr5 (Fig.2C, D).The endosperm phenotype caused by the ppr5 mutation is more severe than those of mutants reported previously.This is likely due to the lethality of homozygous mutant plants compared to those mutants such as flo6, flo7, and flo8 [33,40,41].Therefore, a leaky mutation in PPR5 may lead to a weak phenotypic change, which can be used for grain quality improvement in the future.Despite the fact that most of the defective endosperm mutants exhibit a flourywhite appearance, they differ in starch grain morphology.For example, loss-of-function mutation of SSG5 and ISA1 seriously inhibits initiation of starch grains and results in accumulation of phytoglycogen [43,44]; and mutations in SSG4 and SSG6 cause enlarged starch grains in the endosperm[45,46].By contrast, endosperm development in ppr5 kernels was delayed, leading to the formation of a large number of small and single starch grains (Fig.3).
Fig.7–ppr5 accumulates unassembled precursors of nad4 intron 3.(A) Gene structure of mitochondrial nad4 and RT-PCR analysis of three nad4 introns in WT and ppr5.Numbers represent primers for different amplification products.M,marker.(B)Agarose gel electrophoretic detection of circular RT-PCR amplification products of nad4 exons 1–3.p,exons 1–3 precursor;m,mature nad4.(C) Mapping of the circular RT-PCR products as determined by sequencing of nad4 exons 1–3.Each bar represents a single sequenced clone and shows the position of the 5′ and 3′ extremities.
Tissues that are responsible for the nutrition transportation in maize endosperm include the basal endosperm transfer layer (BETL) and aleurone.BETL occurs at the endosperm surface near the vascular bundle at the base of the kernel and develops from a layer of cells to a multilayer of cells with grain filling [47].Previous studies identified several loss-offunction mutations in mitochondrial-localized PPR proteins in maize, including Dek2, Dek35, Dek37, EMP8, and EMP10[22-26]that affected mitochondrial function and development of BETL, leading to stagnation of endosperm development and lethal seedlings.The phenotype of rice ppr5 was very similar to those of maize mutants with defective mitochondria-localized PPR proteins, abnormal endosperm development,and seedling lethality(Fig.1 and Fig.S1).In rice,the aleurone layer and ovular vascular trace, rather than the BETL, are considered to function in transporting nutrients into the endosperm, an energy-dependent process[31].Compared with starchy endosperm cells which are near oxygenfree and have few mitochondria, aleurone cells have 8% oxygen concentrations and contain abundant mitochondria [48].Our microscopic observations revealed that the endosperm of ppr5 showed no significant difference in the ovular vascular trace compared with WT (Fig.S3), but aleurone development was abnormal, most likely related to the dysfunctional mitochondria(Figs.3 and 8D).Since the ATP content in ppr5 endosperm was significantly lower than in the WT(Fig.8C),it was likely that the defective aleurone layer in ppr5 with abnormal mitochondrial function could not import sufficient nutrients into the developing grain and cause stress, leading to the floury and shrunken endosperm phenotype.
Fig.8–Loss of PPR5 function disrupted mitochondrial function.(A) Blue native PAGE analysis of mitochondrial complex I.(B) In-gel NADH-dehydrogenase activity analysis of mitochondrial complex I.(C) ATP concentration in grains at 9 DAF.(D) Transmission electron microscopy of mitochondria in WT and ppr5 endosperm at 9 DAF.M, mitochondria.(E) RT-PCR analysis of AOX1A, AOX1B, and AOX1C involved in the alternative respiratory pathway.Actin was used as a sample loading control.(F) Immunoblotting analysis of AOX proteins in the developing endosperm of WT and ppr5 kernels.Anti-EF-1α antibodies were used as loading control.Data are shown as means ± SD from three biological replicates, and compared with WT by Student’s t-test (**, P<0.01).
More than 20 group II introns that exist in the mitochondrial genome in flowering plants require either cis-or trans-splicing[36].P-type PPR proteins involved in group II intron splicing have been extensively studied in Arabidopsis and maize.Here, we identified PPR5 as a regulator required for endosperm development in rice through control of mitochondrial nad4 intron 3 splicing.Most group II introns belong to genes encoding subunits in complex I (NADH dehydrogenase),including cis-spliced introns in nad1, nad2, nad4, nad5, and nad7 as well as trans-spliced introns in nad1, nad2, and nad5[36].MTSF2/PPR19, nMAT1, OTP43, DEK2, and FLO10 were identified as trans-splicing factors in nad1 in Arabidopsis,maize, and rice [18,25,31,49,50].Dek37 and EMP16 are specifically required for cis-splicing of nad2 intron 1 and intron 4 in maize [24,27].Dek35 was shown to participate in cis-splicing of nad4 intron 1 [23].MTL1 is required for both translation and splicing of mitochondrial nad7 mRNA in Arabidopsis[20].PPR5 identified in this study is a P-type PPR protein containing 7 predicted PPR motifs.Splicing analyses combined with circular RT-PCR demonstrated that cis-splicing of nad4 intron 3 was significantly disrupted in ppr5 (Fig.7), suggesting that PPR5 is crucial for nad4 intron 3 splicing in rice mitochondria.Previous studies have shown that P-type PPR proteins can bind to the 5′or 3′ends of their substrates to provide a barrier to exoribonucleases or to remodel target RNA [31,51].Loss-of-function mutations of mitochondrion-targeted PPR protein usually cause reduced stability of targeted transcripts.For example, PPR78 is required for the stability of nad5T1 precursor and nad5 mature mRNA in maize.In the ppr78 mutant, the abundance of intron 1 and/or 2 spliced transcripts is significantly decreased [52].No significant difference in the 5′and 3′extremities of nad4 precursor in ppr5 were detected (Fig.7C, Fig.S8), but the amounts of unassembled precursors were increased (Fig.7).Therefore, PPR5 is required for intron splicing, but does not affect the stability of the precursor, which may be inconsistent with the function of PPR78 protein.A deletion of the single base G leading to premature termination of the PPR5 protein caused a loss of all PPR motifs (Fig.4B and 5A).Hence, the ppr5 protein corresponded to a loss-of-function mutant of PPR5, but the mature nad4 transcript was not completely abolished in ppr5.This indicated that splicing of nad4 intron 3 might also involve other proteins.
NADH dehydrogenase is the major entry point for electron transport that involves nine mitochondrially encoded subunits, including NAD1-4, NAD4L, NAD5-7, and NAD9 [23].Several PPR proteins have been reported to function in the diverse RNA processing of the nad4 transcript in Arabidopsis, maize, and rice (Table S3).EMP8, Dek35, DEK41/DEK43, and RL1 affect the cis-splicing of nad4 intron 1 and/or 3 [23,26,53-55], AHG11 is involved in nad4 RNA editing [56], RPF1 is required for 5′-end processing [57], MTSF1 stabilizes the nad4 mRNA [51].All these studies suggested that NAD4 plays a vital role in mitochondrial function.Supporting this notion, we identified the rice homolog of maize DEK41/DEK43 (Fig.S6) and found that its loss-offunction mutation results in a deficiency in mitochondrial complex I assembly and activity, and leading to seedling lethality and abnormal endosperm development (Figs.3, 7 and 8).Our study demonstrated that correct splicing of the nad4 transcript is a crucial posttranscriptional step in maintaining normal function of mitochondrial complex I in rice.
CRediT authorship contribution statement
Jianmin Wan supervised the project.Jianmin Wan and Yulong Ren designed the research.Yulong Ren screened the mutant material.Long Zhang, Yanzhou Qi, and Mingming Wu performed experiments.Lei Zhao,Zhichao Zhao,Cailin Lei,Yuanyuan Hao,Yinglun Sun,Xiaowen Yu,Xin Zhang,and Xiuping Guo provided technical assistance.Long Zhang and Yulong Ren analyzed the data and drafted the manuscript.Jianmin Wan improved the manuscript.All authors discussed the results and commented on the manuscript.
Declaration of competing interest
Authors declare that there are no conflicts of interest.
Acknowledgments
This work was supported by grants from the National Natural Science Foundation of China (31901427), the National Transgenic Science and Technology Program (2019ZX08010-003),the Agricultural Science and Technology Innovation Program of CAAS (CAAS-ZDXT2018001) and the Young Talent of CAAS to Yulong Ren.
Appendix A.Supplementary data
Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2020.09.002.