Chloroplast C-to-U editing, regulated by a PPR protein BoYgl-2, is important for chlorophyll biosynthesis in cabbage

Abstract

Leaf color is an important agronomic trait in cabbage (Brassica oleracea L. var. capitata), but the detailed mechanism underlying leaf color formation remains unclear. In this study, we characterized a Brassica oleracea yellow-green leaf 2 (BoYgl-2) mutant 4036Y, which has significantly reduced chlorophyll content and abnormal chloroplasts during early leaf development. Genetic analysis revealed that the yellow-green leaf trait is controlled by a single recessive gene. Map-based cloning revealed that BoYgl-2 encodes a novel nuclear-targeted P-type PPR protein, which is absent in the 4036Y mutant. Functional complementation showed that BoYgl-2 from the normal-green leaf 4036G can rescue the yellow-green leaf phenotype of 4036Y. The C-to-U editing efficiency and expression levels of atpF, rps14, petL and ndhD were significantly reduced in 4036Y than that in 4036G, and significantly increased in BoYgl-2 overexpression lines than that in 4036Y. The expression levels of many plastid- and nuclear-encoded genes associated with chloroplast development in BoYgl-2 mutant were also significantly altered. These results suggest that BoYgl-2 participates in chloroplast C-to-U editing and development, which provides rare insight into the molecular mechanism underlying leaf color formation in cabbage.

Introduction

The chloroplast is a crucial organelle for photosynthesis and provides necessary energy for plant growth and development. Chlorophyll is the critical pigment responsible for absorbing and transferring light energy to the photosynthetic system. Leaf color mutants are usually caused by impaired chloroplasts and blocked chlorophyll synthesis. Leaf color variations affect the photosynthetic rate, resulting in stunted plant growth and reduced yields. Leaf color mutants can be used to rapidly identify variety purity in crop hybrid breeding and understand plant photosynthetic mechanisms, chlorophyll anabolic pathways, and gene regulatory networks [1–5].

Pentatricopeptide repeat (PPR) is a large protein family that participates in post-transcriptional RNA modification, such as RNA editing, RNA processing, RNA translation, RNA stability and RNA splicing [6–10]. PPR proteins can be classified into P and PLS subfamilies and play essential roles in chloroplast development [11]. In Arabidopsis, Wang et al. identified the P-type PPR protein ECD2; ECD2 RNAi lines had an albino cotyledon phenotype. Further analysis revealed that ECD2 is related to chloroplast gene expression and group II intron splicing [12]. In soybean, Feng et al. identified a PLS-type PPR protein GmPGL2, and plants with mutant GmPGL2 exhibited pale-green leaves. GmPGL2 is essential for C-to-U editing of ndhB, ndhD, rps16, ndhF, and ndhE, rps18 genes in chloroplasts [13]. In rice, Lan et al. identified a young leaf white stripe (ylws) mutant. YLWS encodes a P-type PPR protein, which mutation causes defects in chloroplast RNA editing of ndhA, rps14 and ndhB, and RNA splicing of ndhA, rps12, atpF and rpl2 genes [14]. Huang et al. identified a PLS-DYW subfamily PPR protein OsPPR16, which is required for RNA editing of rpoB-545 in rice plastids [15]. In maize, a qKW9 mutant with smaller ears and fewer kernels was identified. qKW9 encodes a PLS-type PPR protein that affects chloroplast C-to-U editing of ndhB and photosynthesis [16].

Chromosomal deletion is a classic type of structural variation, and several chromosome fragment deletions are associated with crop-specific phenotypes. Li et al. identified a 13.96-kb chromosomal fragment deletion in watermelon. Through expression and co-segregation marker analysis, two genes Cla97C02G045390 and Cla97C02G045400 in the chromosomal deletion region were found to be responsible for the seed size in watermelon [17]. In Brassica napus, Zhang et al. localized the delay-green leaf gene BnaA02.YTG1 to a 9.9-kb region and found that the fragment was deleted in the ytg mutant by sequence analysis. By functional analysis, the BnaA02g10480D gene in the chromosomal deletion region was found to control the delay-green leaf phenotype and participate in the chloroplast RNA editing in rapeseed [18]. In Cucurbita pepo, Zhu et al. mapped the dark-green stem color gene CpDsc-1 to a 65.2-kb interval, and a 14-kb chromosomal fragment deletion between Cp4.1LG15g03360 and Cp4.1LG15g03420 genes was identified in the candidate region. Expression and co-dominant marker analysis indicated that Cp4.1LG15g03420 may be the main cause of the dark-green stem phenotype in zucchini [19]. In Arabidopsis, a 14-kb deletion on chromosome 3 was identified in the T-DNA mutant SALK_008491, and the loss of both NHD1 and PGDH3 causes SALK_008491 highly sensitive to dynamic light stress [20].

In this study, we identified a new spontaneous yellow-green leaf mutant BoYgl-2 in cabbage. Using BC₁ and F₂ populations, the BoYgl-2 gene was fine mapped to a tiny region by BSA-seq and linkage analysis, and a large chromosomal deletion was detected in the BoYgl-2 locus. BoYgl-2 encodes a nuclear-targeted P-type PPR protein that affects chloroplast RNA editing. This finding lays a foundation for revealing the molecular mechanism underlying leaf color formation in cabbage.

Results

Characterization of the BoYgl-2 mutant

Compared with 4036G (normal-green leaf), the 4036Y mutant exhibited a yellow-green leaf phenotype at seedling stage and returned to normal-green leaf at mature stage (Fig. 1A). The Chl a and Chl b contents were significantly reduced in 4036Y than that of 4036G at the seedling stage, but had no significant difference at the mature stage (Fig. 1B, Supplemental Fig. S1). To investigate the effect of the BoYgl-2 mutation on chloroplast biogenesis, the chloroplast ultrastructures of the leaves of 4036G and 4036Y seedlings were observed by transmission electron microscopy (TEM). The chloroplasts in 4036G leaves exhibited well-structured and well-organized thylakoid membranes (Fig. 1C). However, in the 4036Y mutant, the chloroplasts lacked organized thylakoid membranes (Fig. 1D). These results suggest that BoYgl-2 is involved in Chl biosynthesis and chloroplast development.

Figure 1

Phenotypic characterization of the BoYgl-2 mutant. (A) 4036G with normal-green leaves (left) and 4036Y with yellow-green leaves (right). Bar = 2.5 cm. (B) Pigment contents in the leaves of 4036G and 4036Y at the seedling stage. Error bars represent the standard errors of three biological replicates (Student’s t-test: **P < 0.01). (C) Chloroplast ultrastructure in the seedling leaves of 4036G. (D) Chloroplast ultrastructure in the seedling leaves of 4036Y. T, thylakoid membrane; S, starch grain.

Fine mapping of the BoYgl-2 gene

To identify the BoYgl-2 gene, the 4036Y mutant was crossed with 4036G. All the leaves of the F₁ plants were normal green. The F₂ population comprised 2508 individuals, with 1885 normal-green leaf and 623 yellow-green leaf individuals, the segregation ratio is 3:1. All BC₁P₁ individuals showed normal-green leaf. The BC₁P₂ population contained 653 yellow-green leaf and 659 normal-green leaf individuals, with a segregation ratio of 1:1 (Table 1). These results indicate that the yellow-green leaf phenotype is controlled by a single recessive nuclear gene.

Table 1

Genetic analysis of the yellow-green leaf trait in BC₁ and F₂ populations

Populations	Total plant number	Number of normal–green leaf plants*	Number of yellow–green leaf plants*	Expected ratio	χ2a
F1	16	16	0	-	-
F2	2508	1885	623	3:1	0.03
BC1P1	120	120	0	-	-
BC1P2	1312	659	653	1:1	0.03

^*Normal-green and yellow-green leaf plants were identified at the seedling stage by visual inspection

^aχ² >χ²_0.05 = 3.84 was considered significant

BSA-seq analysis was performed to preliminarily map the BoYgl-2 gene. The highest peak region (P<0.01), which contains 5.36 Mb (0–5.36 Mb) on chromosome 3 according to the cabbage reference genome (TO1000), was taken as the candidate interval associated with BoYgl-2 (Fig. 2A). To fine map the BoYgl-2, 13 InDel markers with polymorphism between the parents were developed within the 5.36-Mb candidate region and then used to analyze a total of 1276 recessive individuals (yellow-green leaf) from the BC₁P₂ and F₂ populations. A linkage map consisting of 13 InDel markers (Supplementary Table S1) was constructed. The InDel marker B36-11 was found to be closely linked to BoYgl-2, with a genetic distance of 0.1 cM. Based on Based on the location of marker in the reference genome, BoYgl-2 was ultimately mapped to a 232-kb region (C03: 0 bp-232,068 bp) (Fig. 2B).

Figure 2

Fine mapping of the BoYgl-2 gene. (A) Plot of the Δ(SNP-index) value obtained from the two bulks. The top red line indicates the threshold line (99% confidence level). The x-axis is the position of nine chromosomes and the y-axis is the Δ(SNP-index) value. (B) Linkage map of the BoYgl-2. InDel marker B36-11 was closely linked to the BoYgl-2 gene.

A 162-kb chromosomal deletion was identified at the BoYgl-2 locus

To identify the candidate gene for BoYgl-2, visual analysis of reads in the candidate interval was performed by IGV (Integrative Genomics Viewer) using the resequencing data of 4036G and 4036Y. A 162-kb chromosome deletion (0 bp-161,975 bp) was detected at the BoYgl-2 locus in 4036Y mutant, reads with high sequencing depth in this region were genome highly repetitive sequences (Fig. 3A). Subsequently, five markers, DEL10, DEL35, DEL85, DEL131, and DEL163 (at positions 10 kb, 35 kb, 85 kb, 131 kb, and 163 kb, respectively), were developed based on the TO1000 genome. These markers were then used to verify the parents and another cabbage normal-green leaf inbred line, 0120. The results showed that the target bands of all the markers could be amplified in 0120 and 4036G, but only the target band of DEL163 (outside the chromosomal deletion fragment) could be amplified in 4036Y (Fig. 3B), verifying the large chromosomal terminal deletion at the BoYgl-2 locus in the 4036Y mutant.

Figure 3

Identification of the chromosomal deletion at the BoYgl-2 locus. (A) Visual analysis of the sequence reads of 4036G and 4036Y at the BoYgl-2 locus by IGV. (B) Amplicons of the markers DEL10, DEL35, DEL85, DEL131 and DEL163 in 0120, 4036G and 4036Y. M represents the DNA ladder.

Functional verification of BoYgl-2

Based on the cabbage reference genome (TO1000), 53 genes were identified within the 232-kb candidate interval (Supplementary Table S2). According to the comparative genomic annotation in A. thaliana (TAIR), only one gene, Bo3g001140, was strongly related to the formation of leaf color. Bo3g001140 is a homolog of the AT1G02420 gene in Arabidopsis, which encodes a pentatricopeptide repeat (PPR) protein involved in chloroplast RNA processing and is located in the chromosomal deletion region. Many PPR mutants exhibit a yellow-green leaf or albino phenotype [12–14, 21]. Thus, we designated Bo3g001140 as the candidate gene for BoYgl-2.

To test whether Bo3g001140 is responsible for the BoYgl-2 phenotype, we transformed the full-length CDS of Bo3g001140, driven by the CaMV 35S promoter, into 4036Y mutants, and obtained three independent overexpressing transgenic T₁ lines. All the overexpression lines showed a normal-green leaf phenotype similar to 4036G (Fig. 4A), and qRT-PCR analysis showed that the expression level of Bo3g001140 was significantly higher in overexpression lines than that in 4036Y mutant (Supplemental Fig. S2). These results indicate that Bo3g001140 is the BoYgl-2 gene controlling leaf color formation in cabbage.

Figure 4

Functional analysis of the BoYgl-2 gene, RNA editing and expression analysis of atpF, rps14, petL and ndhD. (A) Phenotypes of BoYgl-2 overexpression lines at the seedling stage. Bar = 3 cm. (B) C-to-U editing efficiency of atpF, rps14, petL and ndhD in 4036G, 4036Y and BoYgl-2 overexpression lines at the seedling stage. (C) Expression levels of atpF, rps14, petL and ndhD in 4036G, 4036Y and BoYgl-2 overexpression lines at the seedling stage. Error bars represent the standard errors of three biological replicates (Student’s t-test: **P < 0.01).

BoYgl-2 encodes a nuclear-localized P-type PPR protein

Sequence analysis revealed that BoYgl-2, which encodes a putative protein of 484 amino acids, contained nine PPR motifs and belonged to the P-type subfamily according to the TPRpred database (https://toolkit.tuebingen.mpg.de/tools/tprpred) (Fig. 5A). Next, a phylogenetic tree of the BoYgl-2 protein and its nine homologues from other cruciferous species was constructed to analyze their evolutionary relationship. The results showed that BoYgl-2 was conserved in these homologous and shared a closer relationship with Brassica cretica-F2Q69_00000017 (98.55%) and Brassica carinata-Bca52824_006135 (97.73%), indicating that they may have originated from the same ancestor gene (Fig. 5B). To determine the subcellular localization of BoYgl-2, we performed a transient expression assay in tobacco leaves. The GFP fluorescence of BoYgl-2:GFP fusion protein was colocalized with the nuclear marker fluorescence (Fig. 5C). The results demonstrated that BoYgl-2 is a nuclear-localized PPR protein.

Figure 5

Phylogenetic analysis and subcellular localization of BoYgl-2. (A) Schematic diagram of the BoYgl-2 protein with nine PPR motifs. Blue boxes represent P-type repeats. (B) Phylogenetic tree of the BoYgl-2 protein and its nine homologues from other cruciferous species. Numbers represent bootstrap values. (C) Subcellular localization of the BoYgl-2:GFP fusion protein. Bar = 20 μm.

Chloroplast RNA editing is impaired in the BoYgl-2 mutant

Many PPR proteins participate in chloroplast RNA editing [13, 15, 22, 23]. We determined whether this was the case for BoYgl-2. By RNA-seq analysis, 29 C-to-U editing sites were identified in the chloroplast of the parental lines. Editing-specific markers were then used to examine the C-to-U editing efficiency of the 29 editing sites in 4036G and 4036Y at the seedling stage. Five editing sites (rpoC1-21382, accD-57118, psbF-62793, rps12-clpP-68694, and ndhD-115201) showed a less than 10% increase, and five editing sites (rpoB-23461, rpoB-25327, lhbA-34880, rpoA-77634, and rpl2-exon2-152763) exhibited no difference in C-to-U editing efficiency in 4036Y. The C-to-U editing efficiency of the remaining 19 editing sites was reduced to varying degrees in 4036Y, among which the atpF-12328, rps14-36077, petL-64504, and ndhD-116086 were reduced by more than 36% (Table 2). In BoYgl-2 overexpression lines, the C-to-U editing efficiency of atpF-12328, rps14-36077, petL-64504 and ndhD-116086 was significantly increased and recovered to be similar to 4036G (Fig. 4B). In addition, qRT-PCR analysis showed that the expression levels of atpF, rps14, petL and ndhD were significantly decreased in 4036Y than that in 4036G, and significantly increased in BoYgl-2 overexpression lines than that in 4036Y mutant (Fig. 4C). These results indicate that BoYgl-2 is important for improving the C-to-U editing efficiency of atpF, rps14, petL and ndhD in cabbage chloroplast.

Table 2

Chloroplast RNA editing efficiency analysis in 4036G and 4036Y.

Gene	Region	Position	4036G	4036Y	Difference	Nucleotide change	Amino acid change
rps16	intronic	5515	62.50%	58.33%	-4.17%	/	/
atpF	exonic	12328	83.33%	45.83%	-37.50%	C->T	Pro-Leu
rpoC1	exonic	21382	68.75%	78.72%	9.97%	C->T	Ser-Leu
rpoB	exonic	23461	100.00%	100.00%	0.00%	C->T	Ser-Leu
rpoB	exonic	25327	89.58%	89.58%	0.00%	C->T	Ser-Leu
rpoB	exonic	25342	89.58%	83.33%	-6.25%	C->T	Ser-Leu
lhbA	exonic	34880	100.00%	100.00%	0.00%	C->T	Ser-Leu
rps14	exonic	36077	64.58%	27.08%	-37.50%	C->T	Pro-Leu
accD	exonic	56509	97.92%	93.75%	-4.17%	C->T	Ser-Leu
accD	exonic	57118	91.67%	95.83%	4.17%	C->T	Pro-Leu
psbF	exonic	62793	89.58%	91.67%	2.08%	C->T	Ser-Phe
petL	exonic	64504	82.98%	46.81%	-36.17%	C->T	Pro-Leu
rps12-clpP	intergenic	68694	43.75%	62.50%	18.75%	/	/
clpP	exonic	68755	60.42%	45.83%	-14.58%	C->T	His-Tyr
rpoA	exonic	77634	100.00%	100.00%	0.00%	C->T	Ser-Phe
ndhB	exonic	94173	81.25%	52.08%	-29.17%	C->T	His-Tyr
ndhB	exonic	94556	68.75%	54.17%	-14.58%	C->T	Ser-Leu
ndhB	exonic	94598	62.50%	41.67%	-20.83%	C->T	Ser-Leu
ndhB	exonic	95361	77.08%	70.83%	-6.25%	C->T	Ser-Phe
ndhB	exonic	95958	91.67%	89.58%	-2.08%	C->T	Ser-Leu
ndhF	exonic	111249	89.58%	79.17%	-10.42%	C->T	Ser-Leu
ndhD	exonic	114778	93.75%	87.50%	-6.25%	C->T	Ser-Leu
ndhD	exonic	115201	77.08%	79.17%	2.08%	C->T	Pro-Leu
ndhD	exonic	115210	95.83%	93.75%	-2.08%	C->T	Ser-Leu
ndhD	exonic	115414	95.83%	93.75%	-2.08%	C->T	Ser-Leu
ndhD	exonic	115705	100.00%	95.83%	-4.17%	C->T	Ser-Leu
ndhD	exonic	116086	93.75%	54.17%	-39.58%	C->T	Thr-Met
ndhG	exonic	117780	97.92%	87.50%	-10.42%	C->T	Ser-Phe
rpl2-exon2	splicing	152763	100.00%	100.00%	0.00%	/	/

Expression analysis of chloroplast-related genes

Many studies have shown that PPR proteins are involved in chloroplast development [4, 13, 21, 23]. We analyzed the expression patterns of chloroplast-related genes in 4036G and 4036Y. The expression levels of the PEP-dependent genes psaA and rbcL and ribosome gene rps2 were significantly reduced, and the expression levels of the NEP-dependent gene rpoA and ribosome genes rps15, rps18, rpl20, rpl23, rpl33, and 23S rRNA were significantly increased. However, the expression levels of other chloroplast-encoded genes psbA (PEP-dependent), rpoB, rpoC1, rpoC2 (NEP-dependent), rps7, and 16S rRNA (ribosomal) did not differ significantly between 4036Y and 4036G. The expression levels of the photosynthesis-associated genes, RBCS1A and CAB1, and the Chl biosynthesis gene, CAO, were significantly down-regulated, while the expression levels of other nuclear-encoded genes PsaD2 (photosynthetic), DVR, and CHLG (Chl biosynthetic) were not significantly altered in 4036Y compared with 4036G (Fig. 6). These results suggest that the expression patterns of genes related to chloroplast development, Chl biosynthesis and photosynthesis were significantly affected in the BoYgl-2 mutant.

Figure 6

Expression patterns of genes associated with chloroplast development in 4036G and 4036Y. BoActin served as the equal loading control. Error bars represent the standard errors of three biological replicates (Student’s t-test: **P < 0.01).

Discussion

Most PPR proteins are localized and function in chloroplasts or mitochondria. In Arabidopsis, rice and maize, some chloroplast-targeted PPR proteins have been identified, which are related to pale-green, albino, pale-yellow, and white-striped leaf formation [4, 14, 15, 21, 22, 24]. And several mitochondrion-localized PPR proteins have been identified to be associated with leaf albino, pale-green and dark curling and seed development [25–28]. In addition, very few PPR proteins have been reported to be localized in the nucleus. Hao et al. identified a rice nuclear-targeted PPR protein OsNPPR1 that is essential for mitochondrial function and endosperm development [29]. In Arabidopsis, a nuclear-localized PPR protein GRP23, and a mitochondrion and nuclear dual-localized PPR protein were identified, which are essential for early embryo development [30, 31]. In this study, we identified the first nuclear-targeted P-type PPR protein BoYgl-2 in cabbage, which is important for chloroplast development and chlorophyll biosynthesis, providing novel insights into the molecular mechanism underlying leaf color formation in cabbage.

PPR proteins are crucial for the normal activities of chloroplasts and mitochondria. Many PPR gene mutations result in variable leaf color phenotypes. In rice, Chen et al. identified a PLS-type PPR protein, PGL12. A single-base mutation (C to T) in the pgl12 gene generates a premature stop codon, which severely affects the 16S rRNA processing and plastid ndhA RNA splicing, resulting in the pale-green leaf phenotype of pgl12 mutant [21]. Wang et al. identified a P-type PPR protein WSL4. A 2-bp deletion in the WSL4 gene, generating a premature stop codon, causes defects in chloroplast RNA editing of rpoB, and RNA splicing of atpF, ndhA, rpl2, and rps12 genes, which results in the white-stripe leaf phenotype of WSL4 mutant [23]. In soybean, Feng et al. identified a pale-green leaf mutant Gmpgl2. A single-base deletion in the Gmpgl2 gene produces a truncated protein that lacks portion of the E2 and E+ motifs, which affects the C-to-U editing of NDH complex subunits and ribosome genes [13]. In maize, a P-type PPR protein PPR8522 was identified, and a 3.3-kb MuDR insertion in PPR8522 is responsible for the emb8522 mutant seedling albino phenotype [24]. In the present study, a nuclear-localized P-type PPR protein, BoYgl-2, was identified, which is completely absent in the 4036Y mutant. The C-to-U editing efficiency and expression levels of atpF, rps14, petL and ndhD were significantly decreased in 4036Y than that in 4036G, and significantly increased in BoYgl-2 overexpression lines than that in 4036Y, indicating that the deletion of BoYgl-2 impaired the C-to-U editing of atpF, rps14, petL, and ndhD and thus affected the chlorophyll biosynthesis in BoYgl-2 mutant. BoYgl-2 is localized in the nucleus, but how it participates in C-to-U editing of chloroplast genes and affects chloroplast function in cabbage remains unclear. We speculated that BoYgl-2 may indirectly regulate chloroplast RNA editing and chlorophyll biosynthesis through a new pathway, which requires further experimental verification in the future.

Marker-assisted selection is an effective method for genetic breeding in many B. oleracea crops. Zhang et al. identified a lobed leaf gene BoLMI1a in ornamental kale. dCAPS marker DMLMI1 and co-dominant marker CMLMI1 were developed based on the BoLMI1a promoter variations, which can be used for marker-assisted selection of leaf shape in ornamental kale breeding [32]. Zhang et al. identified a petal color gene BoCCD4.2 in Chinese kale. Based on the CACTA-like transposon insertion in BoCCD4.2, a co-dominant marker TE3 was developed that can be used for marker-assisted selection of white/yellow petal color in Chinese kale [33]. In cabbage, Han et al. identified a male-sterile gene BoTPD1, a BoTPD1-specifc marker based on the 182-bp InDel co-segregated with male sterility and can be used for marker-assisted selection [34]. Ji et al. identified a wax-deficient gene BoCER2. According to a G-to-A substitution in the BoCER2 coding region, a BoCER2-specifc KASP marker was designed, which can be used for marker-assisted selection for glossiness [35]. In this study, we identified a yellow-green leaf gene BoYgl-2, and a 162-kb chromosomal deletion was identified at the BoYgl-2 locus. Four markers, DEL10, DEL35, DEL85, and DEL131 were developed in the chromosomal deletion region that can only be amplified in normal-green leaf materials. These markers can be used for marker-assisted selection of leaf color in cabbage breeding.

Materials and methods

Plant materials

The 4036G (P₁) is a cabbage inbred line with normal-green leaves; the 4036Y (P₂) is a yellow-green leaf mutant isolated from 4036G. 4036G was crossed with 4036Y to produce F₁ lines. F₁ lines were self-pollinated to generate F₂ population; the BC₁P₁ and BC₁P₂ populations were generated by backcrosses of F₁ × 4036G and F₁ × 4036Y, respectively. All plant materials used in this study were grown in a greenhouse (16 h light/8 h dark photoperiod; 25 °C ± 3 °C) at the Institute of Vegetables and Flowers, Chinese Academy of Agriculture Sciences (IVFCAAS, Beijing, China).

Pigment content and transmission electron microscopy analysis

Chlorophyll and carotenoid contents were determined as previously described [18]. Fresh leaves (~0.2 g) of 4036G and 4036Y were collected at the seedling stage, and then placed in 5 ml of 80% acetone for 24 h in the dark. Chl a, Chl b and carotenoid contents were measured using a DU800 spectrophotometer (Beckman Coulter, USA) at wavelengths of 663, 645 and 470 nm, respectively. Three biological replicates were performed per sample.

Transmission electron microscopy (TEM) was performed on 4-week-old leaves of 4036G and 4036Y as previously described [5].

BSA-seq analysis and fine mapping of the BoYgl-2 gene

Young leaves from thirty normal-green leaf and thirty yellow-green leaf BC₁ individuals (4-week-old) were collected to construct two bulks. High-quality genomic DNAs from the two bulks and two parental lines were extracted using the FastPure Plant DNA Isolation Mini Kit (Vazyme, Nanjing, China) to construct Illumina libraries [32], which were subsequently sequenced on an Illumina Hi-Seq 2500 sequencer by Biomarker Technologies Co., Ltd. (Beijing, China). Δ(SNP-index) analysis was performed as previously described [36].

InDel markers were designed by comparing the resequencing data from the candidate region of the 4036G and 4036Y parents. The yellow-green leaf individuals from the BC₁P₂ and F₂ populations were analyzed by the markers with polymorphism between the parents. Primer design, PCR analysis, and genetic and physical map construction were performed according to the method described before [32].

Vector construction and cabbage transformation

For the complementation experiment, the full-length CDS of BoYgl-2 was amplified from the 4036G and cloned into the binary expression vector pBWA(V)BS-CaMV 35S. The recombinant plasmid was then transformed into 4036Y mutant by Agrobacterium-mediated cabbage transformation. Positive transgenic plants were identified by PCR amplification. All primers used for vector construction and identification are listed in Supplementary Table S1.

Phylogenetic analysis and subcellular localization

Homologues of the BoYgl-2 protein were obtained from the National Center for Biotechnology Information (NCBI) database by a BLASTP search. Phylogenetic analysis was performed as described previously [32].

The full-length CDS of BoYgl-2 without the stop codon was cloned into the green fluorescent protein (GFP) expression vector pBWA(V)HS-35S-GFP. Far red fluorescent protein (mKate) expression vector was used as the nuclear marker. The fusion construct and the mKate construct were transformed into Agrobacterium GV3101 and then injected into tobacco leaves as described previously [5]. GFP fluorescence signals were detected under a confocal laser-scanning microscope (Carl Zeiss, Germany).

RNA editing analysis

Total RNA was isolated from the young leaves of 4-week-old seedlings of 4036G and 4036Y using the FastPure Universal Plant Total RNA Isolation Kit (Vazyme, Nanjing, China). cDNA was synthesized using the HiScript III 1st Strand cDNA Synthesis Kit (Vazyme, Nanjing, China). The cDNAs were subsequently used for Illumina library construction and sequencing with an Illumina Hi-Seq 4000 sequencer by Biomarker Technologies Co., Ltd. The transcripts of chloroplast genes were identified by referencing the cabbage chloroplast genome [37]. RNA editing site analysis was performed as previously reported [13].

Specific markers for RNA editing sites were designed based on the cabbage chloroplast reference genome (Supplementary Table S1). RNA editing-specific markers were then used to amplify the cDNAs from the young leaves of 4-week-old seedlings of 4036G and 4036Y. PCR products were then cloned into a T-vector using the TA/Blunt-Zero Cloning Kit (Vazyme, Nanjing, China). Fifty positive clones were selected from each sample for sequencing and RNA editing efficiency analysis.

Expression analysis

The expression patterns of genes related to chloroplast development were analyzed by quantitative real-time PCR (qRT–PCR). A FastPure Universal Plant Total RNA Isolation Kit (Vazyme, Nanjing, China) was used to extract total RNA from 4-week-old seedling fresh leaves of 4036G and 4036Y according to the manufacturer’s instructions. cDNA was synthesized using the TIANGEN FastKing RT Kit (Tiangen, Beijing, China). qRT–PCR was carried out using the ChamQ Universal SYBR qPCR Master Mix (Vazyme, Nanjing, China) on a CFX96 Real-Time System (Bio-Rad, USA). Relative expression levels of the genes were calculated using the 2^−ΔΔCt method [38]. Three biological and three technical replicates were performed for all experiments. BoActin was used as the internal control gene. The qRT–PCR primers are listed in Supplementary Table S1.

Acknowledgments

This work was supported by a grant from the National Natural Science Foundation of China (31872948).

Author Contributions

X.H. and Y.Z. conceived and designed the experiments. B.Z., Y.W., and S.L. performed the experiments and analyzed the data. B.Z. wrote the manuscript. X.H. and Y.Z. revised the manuscript. W.R. conducted the expression analysis. L.Y., M.Z., H.L., Y.W., and J.J. provided the valuable suggestions on the manuscript. All authors read and approved the manuscript.

Data Availability

All the data generated or analyzed in this study are included in this published article and its supplementary information files. All the sequence data of the present study have been deposited in the NCBI Sequence Read Archive (SRA) database under BioProject PRJNA1015626.

Conflict of Interest statement

The authors declare that they have no competing interests.

Author notes

These authors contributed equally to this work.