XIA Xiaodong, ZHANG Xiaobo, WANG Zhonghao, CHENG Benyi, Sun Huifeng, XU Xia,GONG Junyi, YANG Shihua, WU Jianli, SHI Yongfeng, XU Rugen

(1College of Agronomy, Yangzhou University, Yangzhou 225009, China; 2China National Rice Research Institute, Hangzhou 310006,China; 3Agro-Tech Extension Center of Haiyan County, Jiaxing 314300, China; #These authors contributed equally to this work)

Abstract: An EMS (ethy methanesulfonate)-induced lethal etiolated (le) mutant obtained from the rice variety Zhongjian 100 was characterized by lethal etiolated phenotypes, with significantly reduced levels of chlorophyll a, chlorophyll b, total chlorophyll, and carotenoids. Additionally, the mutant displayed a significantly decreased number of chloroplast grana, along with irregular and less-stacked grana lamellae.The le mutant showed markedly diminished root length, root surface area, and root volume compared with the wild type. It also exhibited significantly lower catalase activity and total protein content, while peroxidase activity was significantly higher. Using the map-based cloning method, we successfully mapped the LE gene to a 48-kb interval between markers RM16107 and RM16110 on rice chromosome 3. A mutation (from T to C) was identified at nucleotide position 692 bp of LOC_Os03g59640 (ChlD), resulting in a change from leucine to proline. By crossing HM133 (a pale green mutant with a single-base substitution of A for G in exon 10 of ChlD subunit) with a heterozygous line of le (LEle), we obtained two plant lines heterozygous at both the LE and HM133 loci. Among 15 transgenic plants, 3 complementation lines displayed normal leaf color with significantly higher total chlorophyll, chlorophyll a, and chlorophyll b contents. The mutation in le led to a lethal etiolated phenotype, which has not been observed in other ChlD mutants. The mutation in the AAA+ domain of ChlD disrupted the interaction of ChlDle with ChlI as demonstrated by a yeast two-hybrid assay, leading to the loss of ChlD function and hindering chlorophyll synthesis and chloroplast development. Consequently, this disruption is responsible for the lethal etiolated phenotype in the mutant.

Key words: Oryza sativa; lethal etiolated mutant; gene cloning; functional analysis; reactive oxygen species

Chlorophyll is one of the most important photosynthetic pigments in chloroplasts. Defects in chlorophyll biosynthesis generate drastic consequences such as decreased chlorophyll content, reduction or complete loss of key enzyme activities, inhibition of chloroplast development, ultimately leading to etiolation and albinism,decreased photosynthetic efficiency, and retardation of plant growth or even plant death.

Mutations of genes involved in the chlorophyll biosynthesis pathway in rice usually lead to yellowgreen leaves, white-green leaves, and even death at the seedling stage in severe cases. The yellow-green leaf mutant824ysdisplays decreased chlorophyll content,together with the block of chloroplast development and plant growth. This gene conferring the yellow-green leaf trait encodes divinyl reductase, which catalyzes the conversion of divinyl chlorophyll a to monovinyl chlorophyll a (Wang et al, 2010). The leaves of the mutantwgl1are white-green, and change to light-green at the later growth stage, while the old leaves turn brown necrotic when extended. The geneWGL1encodes a protochlorophyllide oxidereductase B, which is involved in chlorophyll biosynthesis and chloroplast development in rice (Kang et al, 2015). The yellowgreen leaf mutantygl1has yellow leaves at the early growth stage and gradually turns green with high photosynthetic efficiency. The geneYGL1encodes a chlorophyll synthetase and catalyzes esterification of chlorophyllide to complete the last step of chlorophyll biosynthesis (Wu et al, 2007). Two chlorophyll a oxygenase genes,OsCAO1andOsCAO2, catalyze the synthesis of chlorophyll a into chlorophyll b (Lee et al,2005).PORAencodes a protochlorophyllide oxidereductase A in rice, which catalyzes the photoreduction of protochlorophyllide to chlorophyllide in chlorophyll synthesis (Sakuraba et al, 2013).

Magnesium (Mg) chelatase is the first enzyme in the chlorophyll synthesis pathway, which catalyzes the chelation of Mg2+into protoporphyrin IX to form Mg-protoporphyrin IX (Walker and Willows, 1997;Masuda, 2008). Mg chelatase is an ATP-dependent heterologous polymerase and consists of three subunits:ChlI (40 kDa), ChlD (70 to 80 kDa) and ChlH (140 kDa). ChlI and ChlD subunits of Mg chelatase belong to the AAA+(ATPase-associated with various cellular activities) protein family. However, the ChlD subunit lacks ATPase activity. During the activation step, ChlI and ChlD subunits form homohexamers, respectively,followed by the formation of cyclic composite structures.Furthermore, Mg2+and ATP bind to the I-D ring complexes to form ATP-I-D-Mg2+complexes (Willows et al, 1996; Jensen et al, 1999; Lundqvist et al, 2010),which serves as a platform to facilitate the binding of the third subunit ChlH and substrate (protoporphyrin IX) (Lundqvist et al, 2010). The ChlH subunit first binds to the substrate (Reid and Hunter, 2002) and subsequently binds to the ATP-I-D-Mg2+complex,followed by the allostery of the ChlD subunit and the activation of ATPase activity of the ChlI subunit,resulting in ATP hydrolysis and energy generation for Mg2+chelation (Jensen et al, 1999), and then forms Mg-protoporphyrin IX.

GUN4encodes the ChlH subunit binding protein of Mg chelatase, and the C-terminal residues of GUN4 are required for the activation of ChlH in chlorophyll synthesis (Zhou et al, 2012). The expression levels ofChlI,ChlD, andChlHare regulated by light, circadian rhythm, and temperature (Papenbrock et al, 1999;Ishijima et al, 2003; Masuda, 2008; Ruan et al, 2017).Furthermore, the redox status mainly regulates ChlI and ChlH subunits, but not the ChlD subunit. The redox status in chloroplasts can influence disulfide bond formation in the ChlI subunit, thereby regulating Mg chelatase activity (Luo et al, 2012). Chloroplast thioredoxins (TRX) are key regulators of the redox state of chloroplasts, which regulate the activity and stability of target proteins by reducing their disulfide bonds (Luo et al, 2012). Notably, the activity of Mg chelatase is regulated by the interactions among its different subunits (Gibson et al, 1999; Lake et al, 2004;Lundqvist et al, 2010; Zhou et al, 2012; Luo et al,2013). The N-terminal + linker domain, which is essential for Mg chelatase activity, can interact with both the ChlI and ChlH subunits (Luo et al, 2018).Brindley et al (2015) demonstrated that five glutamate residues at the C-terminal domain of ChlD play a critical role in the first step of Mg2+insertion into protoporphyrin IX. Furthermore, the N-terminal AAA+domain of ChlD is essential for the formation of the ChlHID complex (Adams et al, 2016). InSynechocystis,the C-terminal integrin domain of ChlD controls its interaction with ChlH, and mediates the synergistic reaction of Mg chelatase in the presence of Mg2+. The interaction site between the AAA+and chelating domains of Mg chelatase is critical for the utilization of energy from ATP hydrolysis, and the AAA+domain of the ChlI subunit transmits the energy to the active site on the ChlH subunit through the ChlD subunit(Farmer et al, 2019). InChlamydomonas, the phosphorylation of ChlD is associated with Mg chelatase activity (Sawicki et al, 2017). A yeast twohybrid assay demonstrated that single amino acid substitutions block the intrinsic interactions between OsChlI and OsChlD subunits, whereas OsChlI can interact with a few TRXs. These results indicated that the regulatory mechanism of Mg chelatase activity is similar in monocot and dicot plants (Zhang et al,2015). Furthermore, the activity of Mg chelatase is involved in chloroplast-to-nucleus retrograde signaling,abscisic acid signaling, and glucose metabolism regulation (Luo et al, 2015).

Although it has been exploited that Mg chelatase not only catalyzes the synthesis of chlorophyll but also participates in several other processes in the cell,an insight into the role of this enzyme in different cellular processes is lacking. In the present study, a lethal etiolated mutantlewas used to dissect the cause of etiolated lethality at the molecular level. The study of the lethal etiolated mutant gene is significant to understand the synthesis of chlorophyll in plants, plant evolution, nuclear signal transduction in the chloroplast,and stress resistance.

Fig. 1. Phenotype, chlorophyll content and ultrastructure of chloroplasts in wild type (WT) and le mutant.

RESULTS

Phenotype, chlorophyll content, and ultrastructure of chloroplasts

Thelethal etiolated mutantlewas obtained from a rice variety Zhongjian 100 using EMS (ethy methanesulfonate)mutagenesis. The mutant was characterized by a yellow phenotype at the seedling stage (Fig. 1-A and -B). The mutant plants died approximately two weeks after germination (Fig. 1-C). One week after germination, the contents of chlorophyll a and b, carotenoids, and total chlorophyll were significantly lower in the mutant compared with the wild type (WT) (Fig. 1-D to -G).TEM (transmission electron microscope) analysis showed that the chloroplast stroma of WT plants was dense, and the grana lamellae were stacked thicker and arranged tightly, whereas the granum number ofthe chloroplastswas reducedinle, and the grana lamellae were irregular in shape with fewer stacks and loose arrangement (Fig. 1-H to -K).These results revealed that the mutation affects chlorophyll synthesis and chloroplast development in rice plants.

Root morphology

Fig. 2 showed the root morphological indices of WT and itslemutant plants. The total root length, root surface area, and root volume of WT andlewere 23.0 and 17.5 cm,2.0 and 1.3 cm2, and 20 and 12 mm3, respectively. Significant differences were observed in these root parameters between the WT andlemutant plants, indicating that thelemutation affects the growth of plant roots.

Biochemical and physiological analysis

In comparison with WT,leleaves showed significantly lower catalase activity and total intracellular protein content, and significantly higher peroxidase activity(Fig. 3). Hence, the accumulation of excess reactive oxygen species intheleplants caused damage to normal cells.

Fig. 2. Total root length (A), root surface area (B), and root volume(C) of seedlings of mutant le and its wild type (WT).

Genetic analysis

Genetic analysis of the segregation progeny ofLElerevealed that among the 646 individual lines, 481 were normal individuals, and 165 were yellowish individuals, in accordance with a 3:1 segregation ratio.We crossed theLEleheterozygous line with Reyan 1(ajaponicarice variety) and examined the resulting two F2populations (LEle/Reyan 1). The numbers of normal green plants and yellow plants were consistent with a 3:1 segregation ratio, indicating that thelelethal yellowing phenotype is controlled by a single recessive nuclear gene (Table S1).

Fine mapping of LE gene

Ten normal and 10 yellowish plants from the F2population of LEle/Reyan 1 were used to prepare WT and mutant DNA pools. We found that marker RM7076 on rice chromosome 3 may be linked to the lethal etiolated geneLE. Subsequently, 36 yellowish lethal individuals were used to verify the role of RM7076 and its nearby markers. Markers RM7076 and RM16131 had only two and three single exchanges, respectively,among the 36 individuals, preliminarily indicating that the location of theLEgene isbetween RM7076 and RM16131 on chromosome 3. Subsequently, theLEgene was narrowed down to a location between markers RM16107 and RM16110 at a physical distance of about 48 kb (Fig. 4-A). The 48 kb region between RM16107 and RM16110 contains eight possible candidate genes. Comparison of the genomes of WT andlemutant identifiedLOC_Os03g59640as the only gene that differed in DNA sequences (single-substitution of C for T) (Fig. 4-B). Hence, this gene, which encodes the Mg chelatase ChlD subunit (OsChlD), was tentatively identified as a candidate gene for the mutant.

Fig. 3. Physiological and biochemical parameters of mutant le and its wild type (WT).

Complementary validation

Allelism validation

Our previous study reported another mutantHM133ofOsChlDthat contains a single substitution (A for G)mutation in exon 10 ofOsChlD, causing a change in the amino acid from arginine to glutamic acid and resulting in pale green leaves throughout the growth period (Shi et al, 2016). To further examine the association ofHM133with thelemutant, the heterozygous lineLElewas crossed withHM133. The F1population displayed two phenotypes: normal leaf color and pale green leaf, and the F2population of normal leaf color exhibited a 3:1 (normal leaf : yellow leaf) segregation ratio, whereas the F2population of the pale green leaf individuals showed two phenotypes:pale green leaf and yellow leaf with a ratio of 3:1.PCR sequencing of two F1pale green leaf plants demonstrated that thelelocus and theHM133locus were heterozygous, and both of the heterozygous plants survived (Fig. S1), revealing thatLEis allelic toOsChlDand confers the pale green leaf phenotype of the mutantHM133. The mutation inOsChlDin exon 4 from T to C inleis responsible for the yellow leaf and lethality.

Fig. 4. Fine mapping of LE gene conferring lethal etiolated mutation (A) and differences in gene sequence of mutant le and its wild type (WT)at LOC_Os03g59640 (B).

Transformation validation

To verify thatLOC_Os03g59640controls the yellow leaf phenotype ofle, the coding sequence ofLOC_Os03g59640from the WT plant was cloned into the vector pCAMBIA1300-NLuc to form pCAMBIA-1300-NLuc-ChlD. The construct was transformed into embryogenic calli derived fromLElemature seeds using theAgrobacterium tumefaciens-mediated method.The forward and reverse primers S-2 were designed to amplify the segment from an upstream intron to a downstream exon of thelemutation site. A 739-bp product could be amplified from all tested plants (Fig.S2). To further verify the SNP (single nucleotide polymorphism) information at the mutation site in the 739-bp PCR product, the product was diluted 1000-fold and used as a template with the dCAPS (derived cleaved amplified polymorphic sequences) primer pair D-1. A product size of 125 bp indicated successful amplification (Fig. 5-B). The remaining PCR products were digested withKpnI, resulting in a single 103-bp product in complementary lines C-3, C-6, and C-15(Fig. 5-C), demonstrating thatlelewas their endogenous genotype as only thelegenotype in the mutation site can be digested.When DNA extracted from the three transformants was amplified with primer D-1 and digested byKpnI, all three generated 103-bp and 125-bp bands (Fig. 5-D). These plants carried T-DNA from the WT target gene and were complementary plants, exhibiting leaf color similar to the WT plant(Fig. 5-A).

The contents of chlorophyll a and b, carotenoids,and total chlorophyll in transformed plants were significantly higher than those in mutantleplants (Fig.5-E).TEM analysis showed the presence of dense stromata with thick and tightly-arranged grana lamellae in the transformed plants, which was similar to WT(Fig. 5-F). The results demonstrated that the candidate gene was indeed the target gene responsible for the lethal etiolated phenotype.

Bioinformatics analysis

Fig. 5. Verification, phenotypes, chlorophyll contents, and ultrastructure of chloroplasts of complementary plants.A, Phenotypes of wild type (WT), le mutant and complementary plants (C-3,C-6, and C-15).

According to the National Rice Data Center (https://www.ricedata.cn/), theLE(ChlD)gene (2 571 bp) is located on rice chromosome 3. It consists of 15 exons and codes a protein of 754 amino acids, which includes a chloroplast signal peptide, an AAA+domain, and a Mg chelatase domain. A mutation (T to C) at nucleotide position 692 bp changes the protein sequence of the mutant (leucine to proline). The region from amino acids 46 to 372 is the AAA+domain without ATPase activity, whereas the region from amino acids 373 to 485 is the acidic proline-rich region (linker domain)that connects the C- and N-termini of the ChlD subunit. The region from amino acids 486 to 754 constitutes the C-terminal integrin I domain of the ChlD subunit and contains a metal ion coordination motif that is thought to bind to both the ChlI and ChlH subunits.

BLASTp analysis of the ChlD protein sequence demonstrated significant homology to the ChlD of other plants, such asSetaria viridisandPanicum hallii(94%),Sorghum bicolorandZea mays(93%),Lolium perenne(92%),Nicotiana tabacum(91%),Gossypium hirsutum(88%),Hordeum vulgare(85%),Cyanothecesp. PCC 7425 (75%),Thermostichus vulcanus(73%), andCyanidioschyzon merolae(71%) (Fig. S3). Phylogenetic tree analysis also showed that rice CHLD protein is genetically closely related toS. viridisandP. hallii(Fig. S3).

Yeast two-hybrid verification

Fig. 6. Interaction of ChlDle and ChlI detected by yeast two-hybrid assay.

Yeast two-hybrid vectors were constructed to investigate whether the point mutation in theChlDgene would affect the interaction between the ChlD and ChlI subunits. The ChlD subunit of WT interacted with the ChlI subunit, as shown in Fig. 6, but neither the ChlD subunit ofHM133nor the ChlD subunit oflecould interact with the WT ChlI subunit. This result suggested that ChlD subunit mutations inleandHM133affect the interaction with ChlI, which may be the cause of the altered chlorophyll synthesis in mutants.

DISCUSSION

Severe impairment in chlorophyll synthesis and chloroplast development in le mutant

Tetrapyrrole synthesis is regulated by glutamine-tRNA reductase. Similar to mutations inArabidopsis, barley,and maize, thelemutant in the present study had very low levels of chlorophyll a, chlorophyll b, carotenoids,and total chlorophyll. These findings implied that Mg chelatase is crucial for regulating the pace of chlorophyll biosynthesis. The mutant had a severely reduced number of grana, with irregularly shaped and loosely arranged lamellae. This demonstrated that the chloroplast development of the mutant was also negatively affected. Due to mutations in ChlI, ChlD,and ChlH, barley produced different leaf color mutants (Jensen et al, 1996). Chlorophyll a and b contents inlewere very low, and carotenoid content was about 10% of WT, which can explain the lethal etiolated phenotype in the mutant. Numerous albino mutants in rice have extremely low levels of chlorophyll a, b, and carotenoids, and their mutant genes are involved in the pathways for chlorophyll synthesis and chloroplast development. However, a ChlD mutation that causes yellowing and lethality of plants has not been reported in rice so far. Therefore,our research contributes to the body of knowledge regarding chlorophyll synthesis and the molecular mechanism of chloroplast development.

Single-base mutations in exon 4 of ChlD subunit resulting in lethal etiolated phenotype

We used the map-based cloning method to map the gene and elucidate the cause of the deadly yellowing ofle. There are eight potential candidate genes within the region between markers RM16107 and RM16110 on rice chromosome 3, but onlyLOC_Os03g59640exhibited a single base change (from T to C) when comparing the sequences ofleand WT among the eight candidate genes (Fig. 4).LOC_Os03g59640encodes the ChlD subunit of Mg chelatase, a crucial enzyme in chlorophyll synthesis. We crossed the light green lineHM133(which has a single-base substitution of A for G in exon 10 of the ChlD subunit, resulting in an amino acid change from arginine to glutamate)with theleheterozygous mutantLEle. The ratio of normal green leaves to light green leaves was 3:1 in the F1generation ofHM133andLEle. Sequencing of two light green plants revealed that both individuals were heterozygous at theHM133andLEloci.Additionally, we demonstrated that the introduction of theChlDgene through genetic transformation can restore a typical phenotype. The results of subcellular localization indicated that ChlD was localized to the chloroplast, consistent with Ruan et al (2017). The reduced Mg chelatase activity, altered expression of genes involved in chlorophyll production and chloroplast development, and ultimately the yellowing phenotype are all attributed to mutations inChlD(Zhang et al, 2006). Allelic mutations inChlDincludeYgl7,Chl1,Ygl98,HM133, andYgl3. Among these,Chl1, with a mutation in the linker domain, only manifests a pale yellow-green leaf phenotype at the seedling stage (Zhang et al, 2006). Mutations inYgl98,HM133,andygl7occurred in the C-terminal integrin I domain present a light green leaf phenotype (Sun et al,2011; Deng et al, 2014; Shi et al, 2016). However,ygl7forms a protein with new functions and can partially fulfill theChlDfunction, promoting photosynthesis and enhancing light energy use efficiency.Ygl3has mutations in the AAA+domain, whileleandYGLhave mutations as well.Ygl3exhibit a yellow-green leaf phenotype throughout its growth period,lehas deadly yellowing, andYGLdisplays a light green phenotype that transitioned to the normal phenotype at high temperatures (Ruan et al, 2017). Notably, there are significant functional differences among mutations at various positions in theChlD.

Single-base mutations in ChlD subunit in le affect interaction of ChlI subunit with ChlD subunit

A leucine-to-proline substitution occurred at position 231 bp due to the base change at 692 bp in theChlDopen reading frame from T to C inle. The AAA+domain in the ChlD subunit (46-372 aa) lacks ATPase function, while the linker domains (acidic proline-rich sections) connecting the C- and N-termini of the ChlD subunit spans from 373-485 aa. The metal ion coordination motif found in the C-terminal integrin I domain of the ChlD subunit, covering 486-754 aa, is thought to bind to both the ChlI and ChlH subunits.

We examined the protein’s secondary structure on the SWISS-MODEL website (http://swissmodel. expasy.org/), revealing that this amino acid substitution could potentially impact the secondary structure of ChlD protein and, consequently, its ability to interact with the ChlI protein. The yeast two-hybrid assay provided further support for this finding. The ChlD protein inHM133has a mutation in the C-terminal integrin I domain, which prevents it from interacting with ChlI.However, the mutation does not affect leaf color or the ability to produce normal seeds, indicating that only a portion of ChlD protein function has been lost.

ChlDlewith a mutation in the AAA+domain, was unable to interact with ChlI, resulting in yellowing and seedling death. Our data demonstrated that this mutation led to the loss of ChlD function, Mg chelatase activity disruption, chlorophyll synthesis and chloroplast development impairment, ultimately causing seedling death.

On molecular mechanism of breeding at a high light efficiency

The primary organ for photosynthesis in plants is the leaf, and the growth and development of rice are physiologically dependent on photosynthesis. As a complex subcellular system, the chloroplast still harbors many unidentified developmental mechanisms and photosynthesis regulatory networks. To develop novel rice varieties with high light efficiency, it is necessary to expand the excavation of rice leaf color gene germplasm resources, clone and identify more leaf color genes, and thoroughly investigate their mechanisms.Rice growth and development are regulated by a variety of processes, including photosynthesis, chlorophyll synthesis and degradation, and biomass improvement.Leaf color mutants serve as excellent research tools for studying plant photosynthesis, chloroplast growth,and the production of secondary metabolites. They can also be utilized in breeding programs. Due to different growth periods or temperature fluctuations, certain white stripes and albino green mutants may exhibit varying leaf hues, although their basic agronomic features are not noticeably different from WT. In rice,over-expression of the chloroplast geneD1can enhance biomass, yield, and the net CO2assimilation rate,demonstrating excellent breeding value (Chen et al,2020). This gene encodes phosphoribosylamine-glycine ligase, which controls chlorophyll metabolism, cell division and chloroplast development. Because the over-expressedVAL1plants exhibit significantly higher photosynthetic capacity, it can be employed in rice breeding to increase yield (Zhang et al, 2018). In doing so,Ygl7exerts certain ChlD functions, promoting photosynthesis and enhancing the efficiency of photosynthetic activity. Clarifying its molecular biological mechanism contributes to a better understanding of the photosynthetic process and may prove valuable for breeding purposes (Deng et al, 2014).

METHODS

Materials and rice planting

We used theindicavariety Zhongjian 100, the lethal etiolated mutantle, and the pale-green leaf mutantHM133(Shi et al,2016) as materials. Addentionally, the other organisms used wereEscherichia coliDH5α,E. coliBL21,A. tumefaciensEH105, and the yeast strain Y2H Gold. This study also utilized the following plasmids: pGADT7, pGBKT7, pCAMBIA1300-CLUC and pCAMBIA1300-NLUC. Rice plants were grown at the Fuyang Experimental Base of the China National Rice Research Institute during the summer, while the transgenic plants were planted in the transgenic field under normal irrigation and fertilization. Chlorophyll contents, root parameters,and physiological and biochemical parameters were measured at 14 d after sowing in 1/2 MS culture medium with three replications. DPSD9.01 software was used for significance analysis.

Determination of root morphological indices

Root scanning was performed with three replicates for each treatment. Briefly, root samples were placed in a transparent tank containing deionized water, and the position of the roots was adjusted with a forcep to avoid overlap. Subsequently, the complete root images were scanned using a digital scanner(Epson Perfection V800, SEIKO EPSON CORP,Jakarta,Indonesia), and the data were collected and stored. The total root length, total root surface area, and total root volume were subsequently analyzed using the WinRhizo PRO 2016 Root Analysis system (Regent Instruments, Quebec, Canada) (He et al,2009).

Chlorophyll content assay

After cutting the leaves of fresh rice seedlings into small fragments of approximately 2 mm, 0.01 g of leaves were soaked in 1 mL of 95% alcohol and placed in the dark for 48 h,with 95% ethanol as the blank control. This experiment was performed in triplicate for each group of samples. After incubation in the dark, the absorbance of leaf pigment extracts and the control solution was measured at 470, 645, 663, and 652 nm using a Spectra Max i3x Multi-Mode Microplate Reader(Molecular Devices, Sunnyvale, USA). The contents of chlorophyll a, chlorophyll b, carotenoids, and total chlorophyll were calculated following the method of Shang (2022).

Observation of chloroplast ultrastructure

The second leaves from the top of WT andlemutantseedlings were cut into 1 mm fragments and vacuum-fixed in a 2.5%glutaraldehyde solution. Further processing of the sample for TEM observation was performed using the method described by Lv et al (2015). The ultrastructure of chloroplasts was determined at the College of Agriculture and Biotechnology,Zhejiang University, China.

Determination of physiological and biochemical indicators

WT andleseedling leaves were used to determine the contents of total protein, MDA, and H2O2, as well as the activities of CAT, SOD, and POD using kits from Nanjing Jiancheng Biological Co. Ltd, Nanjing, China, according to the manufacturer’s instructions.

Gene localization

A total of 623 evenly distributed SSR (simple sequence repeat)markers on the rice genome were randomly selected to perform polymorphism screening of parentalLEleand Reyan 1.Selected polymorphic markers were compared between the DNA pools of yellow and normal leaves to identify markers related to the gene responsible for the etiolated phenotype.Subsequently, 36 etiolated F2individuals were randomly selected for linkage marker analysis. The genetic mapping analysis was performed using the MapMaker 3.0 software.Furthermore, 614 individual plants with yellow leaves were analyzed to complete fine mapping and generate physical maps.

Candidate gene prediction

Based on the results of the fine mapping, candidate genes in the target region were amplified using primer S-1 (forward 5′-TC AACTGCCAGTGCCAGAC-3′; reverse, 5′-CGGTTACCTCG GAGAAGC-3′) designed with Primer Premier 5, followed by sequencing. The target fragments were recovered and purified using the Gel Extraction & Clean-up Kit (EASY-DO, Hangzhou,China).

Complementary validation

The intermediate vector pSAT1-nVenus-N_(pE3308)-ChlD was double-digested withBamHI andSacI, and then the CDS of ChlD was connected to pCAMBIA1300-NLuc using T4 DNA ligase (TaKaRa, Beijing, China). Subsequently, it was transformed intoE. coliDH5α competent cells. The positive recombinants were screened by colony PCR and verified by sequencing. The recombinant plasmid along with heterozygous seeds (LEle),was sent to Wuhan Boyuan Company, China, for transformation.To detect mutation sites, the primer pair S-2 (forward, 5′-GT TCTCCTCTTATGGCAGATG-3′; reverse, 5′-CCTCTTTGCT GGACTCTTG-3′) was used in the first step. The forward and reverse S-2 primers were designed to amplify regions upstream of an intron and downstream of an exon, respectively,containing thelemutation site. The resulting 739-bp product contained the original genetic background information of the transformants at the mutation site. Subsequently, SNP analysis at the mutation site was conducted. The 739-bp PCR product served as a template in a PCR with dCAPS primer pair D-1(forward, 5′-ATGAGGGCGTAAGCAAGGTAC-3′; reverse, 5′-AGATCCTTCCTCTGGGTTGT-3′). A product size of 125 bp indicated successful amplification. The remaining amplified products from the second PCR were digested withKpnI (only thelegenotype at the mutation site can be digested), and a single 103 bp band demonstrated that the endogenous genotype of the transgenic plants waslele.To verify whether these plants carried T-DNA from the WT target gene, the extracted DNA was used as a template in a PCR performed directly using the primer pair D-1. The amplified products were digested withKpnI and the transgenic lines that showed bands of 103 and 125 bp were considered complementation lines.

Yeast two-hybrid verification

TheCHLDandCHLIgenes were amplified using the primer pairs CHLD-Y2H (forward, 5′-CGACGTACCAGATTACGCT CATATGATGGCGATGGCCACCACCGC-3′; reverse, 5′-AT CGATGCCCACCCGGGTGGAATTCTCACGAACTCTTCAG GTCCG-3′) and CHLI-Y2H (forward, 5′-GATCTCAGAGGA GGACCTGCATATGATGGCTTCCGCCTTCTCCCC-3′; reverse,5′-CAGGTCGACGGATCCCCGGGAATTCTTAGGTGAAGA CTTCATAAA-3′). These were cloned into PGBKT7 and PGADT7 vectors, respectively, and then transformed intoE. coliDH5α competent cells. The plasmids were subsequently extracted and sequenced (Shangya Biological Co. Ltd, Zhejiang Province,China). This experiment was conducted following specific steps as described by Shang (2022).

Bioinformatics analysis

Basic information onChlDwas obtained from the National Rice Data Center (https://www.ricedata.cn/). BLASTp analysis(http://blast.ncbi.nlm.nih.gov/Blast.cgi) was performed using the protein sequence of ChlD to identify homologous proteins in other species. Homologous protein sequences were aligned,analyzed, and subjected to phylogenetic analysis using the MEGA4.0 and Genedoc software.

ACKNOWLEDGEMENTS

This study was supported by the National Natural Science Foundation of China (Grant No. 32072049), and Open Foundation of State Key Laboratory of Rice Biology, China(Grant No. 20210208).

SUPPLEMENTAL DATA

The following materials are available in the online version of this article at http://www.sciencedirect.com/journal/rice-science;http://www.ricescience.org.

Fig. S1. Verification oflemutation site in F1generation ofLEleandHM133.

Fig. S2. Structure ofChlDand its specific amplification location.

Fig. S3. Alignment of ChlD homologous proteins in different species with phylogenetic analysis.

Table S1. Genetic analysis ofle, a lethal etiolated rice mutant.

Table S2. Predicted genes and their encoded proteins between markers RM16107 and RM16110.