Fine Mapping and Candidate Gene Analysis of the Leaf-Color Gene ygl-1 in Maize

A novel yellow-green leaf mutant yellow-green leaf-1 (ygl-1) was isolated in self-pollinated progenies from the cross of maize inbred lines Ye478 and Yuanwu02. The mutant spontaneously showed yellow-green character throughout the lifespan. Meanwhile, the mutant reduced contents of chlorophyll and Car, arrested chloroplast development and lowered the capacity of photosynthesis compared with the wild-type Lx7226. Genetic analysis revealed that the mutant phenotype was controlled by a recessive nuclear gene. The ygl-1 locus was initially mapped to an interval of about 0.86 Mb in bin 1.01 on the short arm of chromosome 1 using 231 yellow-green leaf individuals of an F2 segregating population from ygl-1/Lx7226. Utilizing four new polymorphic SSR markers, the ygl-1 locus was narrowed down to a region of about 48 kb using 2930 and 2247 individuals of F2 and F3 mapping populations, respectively. Among the three predicted genes annotated within this 48 kb region, GRMZM2G007441, which was predicted to encode a cpSRP43 protein, had a 1-bp nucleotide deletion in the coding region of ygl-1 resulting in a frame shift mutation. Semi-quantitative RT-PCR analysis revealed that YGL-1 was constitutively expressed in all tested tissues and its expression level was not significantly affected in the ygl-1 mutant from early to mature stages, while light intensity regulated its expression both in the ygl-1 mutant and wild type seedlings. Furthermore, the mRNA levels of some genes involved in chloroplast development were affected in the six-week old ygl-1 plants. These findings suggested that YGL-1 plays an important role in chloroplast development of maize.


Introduction
The color of plant leaves can significantly vary due to differences in the content of chlorophyll in the chloroplast. This type of mutation is known as chlorophyll deficient and is also called leaf color mutation. Such mutations have been reported in many higher plants since the 1930s, including rice [1], soybean [2], maize [3], barley [4], wheat [5], tomato [6], oilseed rape [7], and Arabidopsis [8].
There are various leaf color mutants known in maize, such as albino, etiolated, light green, stripe or spot, virescent (v), and zebra according to their phenotypic characteristics [30]. According to the maizegdb database (http://www.maizegdb.org/), more than 210 leaf color mutant loci have been identified, distributed over all 10 maize chromosomes. Most genes associated with leaf color mutation are just approximately located to different chromosomes or chromosomal regions, with only a subset cloned and characterized. Among the cloned genes, three encode chlorophyll biosynthesis enzymes or their related proteins. elm1 encodes a phytochromobilin synthase, and its seedling mutant plants have a pale green leaf, pronounced elongation of the mesocotyl, and disrupted de-etiolation responses to red and far-red light [31,13]. Mature mutant plants maintain the pale green leaf phenotype and flower earlier than wild-type plants under long days [13]. Oy1 encodes the I subunit of magnesium chelatase (ZmCHLI), and its seedling mutant plants exhibit yellow-green leaves that can be enhanced by growth at low temperature (24°C) [32]. Mature, field-grown mutant plants show a pale-green phenotype consistent with chlorophyll deficiency. elm2 encodes a heme oxygenase (HO) that is required for chlorophyll biosynthesis, and its mutants spontaneously develop a yellowish phenotype during their entire life [14]. This gene was isolated by map-based cloning, and the 31 bp deletion in the coding region which reduced endogenous HO activity and disrupted the de-etiolation responses to red and far-red signal at seedling stage. Two genes involved in chloroplast development were also characterized. hcf60 encodes the chloroplast ribosomal small subunit protein 17, and its mutants show temperature and light-dependent chlorophyll deficiencies [33]. vyl encodes the ClpP5 subunit of the chloroplast Clp protease, and its mutants exhibit a yellow leaf phenotype after emergence but can recover and become indistinguishable from wild-type plants after about 2 weeks [15].
The above-mentioned leaf-color mutant loci provided useful information in unraveling the chlorophyll synthesis and chloroplast development in maize. However, its mechanisms of leaf color mutation are still far from being fully understood and new genes related to leaf color mutation are of value. Among them, chlorophyll deficient mutants related to the chloroplast signal recognition particle (cpSRP) pathway could be useful tools for studying the progress of chloroplast development in maize [34][35]. The cpSRP contains four proteins that work together: cpSRP54, cpSRP43, cpFtsY, and ALB3 [36].
In this study, we isolated a new recessive yellow-green leaf mutant ygl-1 in maize. The mutant spontaneously exhibited yellow-green leaves phenotype throughout all growth stages. By map-based cloning we isolated the ygl-1 gene. It was predicted to encode a cpSRP43 protein and acts as a chloroplast precursor. Physiological and microscopic analysis revealed that chloroplast development was affected. Transcript levels of some genes associated with chloroplast development were significantly changed in ygl-1 six-week old plants. Our findings provided direct evidence that the YGL-1 gene plays an essential role in chloroplast development of maize.

Plant materials and mapping population
The ygl-1 mutant was identified in self-pollinated progenies from the cross of two maize inbred lines Ye478 and Yuanwu02. An F 2 segregating population was constructed from the cross of a ygl-1 mutant and normal green leaf variety Lx7226. A total of 231 recessive yellow-green leaf individuals from three ears were selected and used for initial mapping. Normal green plants from the same three ears were self-pollinated to generate an F 3 segregating population. A total of 2247 individuals from the F 3 segregating population were used for fine mapping. Later, a large F 2 population was obtained from a cross of the ygl-1 mutant and normal green leaf variety Lx7226. To fine map ygl-1, 2930 individuals from the F 2 population were screened. Plants with crossovers were self-pollinated to generate offspring for phenotypic verification. All the plants used for mapping ygl-1, chlorophyll content determination, transmission electron microscopy analysis and semi-quantitative RT-PCR were grown in the greenhouse of the Shandong Academy of Agricultural Sciences.

Measurement of chlorophyll and carotenoid contents and photosynthetic characteristics
Leaf samples of wild type Lx7226 and the ygl-1 mutant (of approximately 100mg fresh weight, respectively) were cut and submerged in 10ml of a 2:1 solution of 95% (V/V) acetone and ethanol for 48h at 26°C under dark conditions. The concentrations of total chlorophyll (total Chls), chlorophyll a (Chl a), chlorophyll b (Chl b), and carotenoids (Car) were measured with UH5300 UV/Vis spectrophotometer (Hitachi) at 663, 645, and 470nm, respectively. Pigment contents were calculated according to the following equations with slight modification obtained from the methods of Arnon [37].

Transmission electron microscopy analysis
Leaf samples of wild type Lx7226 and the ygl-1 mutant were harvested at different stages of development from the greenhouse. Only the top fully expanded leaves excluding the flag leaf were taken as samples. The fresh leaves were quickly cut into strips of 2mm width. The strips were fixed in a solution of 2.5% (v/v) glutaraldehyde and further fixed with 1% osmium tetroxide. The tissues were dehydrated through a series of acetone solutions and embedded in EMBed-812 prior to thin sectioning on Ultracut UCT (Leica). They were viewed using a transmission electron microscope H-7500 (Hitachi).

DNA extraction
Leaf tissue of each plant was harvested at the two-week old stage, and genomic DNA was extracted by the CTAB method [38]. At the same time, 15 green and yellow-green plants each were randomly selected from the F 2 population to construct DNA pools used to screen for polymorphic markers between ygl-1 and Lx7226.

Marker development and gene prediction
SSR markers used for linkage analysis were obtained from the Maize Genetics and Genomics Database (www.maizegdb.org). To fine map the region of the target locus, additional SSR markers were obtained from previous studies [39]. Gene prediction and annotation within the located region was conducted by Softberry and according to the Maize Genetics and Genomics Database.

Gene cloning and sequence analysis
The DNA sequences of candidate gene GRMZM2G007441 were amplified from both the ygl-1 and its wild-type parents using primer pairs 21-3F (5 0 -CCCAAACGAACATGACCTAAAGC) and 21-6R (5 0 -TTCATCAAGTAATCTCTATCACCTGC). The target DNA fragments were cut from the gel and purified by the DNA Gel Extraction kit (TransGen Biotech). The fragments were connected to the pEASY-T1 cloning vector and sequenced at Invitrogen Biotech. The mutant site of the candidate gene was confirmed by the comparison and analysis of the sequencing results using DNAStar software. Multiple DNA sequence analysis was performed using DNAMAN version 5.0 software.

Multiple amino acid sequence alignment and phylogenetic analysis
The full length protein sequence of YGL-1 (GRMZM2G007441) was obtained from GenBank (http://www.ncbi.nlm.nih.gov). Its homologs in the ygl-1 mutant and 6 wild type lines were obtained according to the translation to amino acid sequences from DNA sequences using BioXM 2.6 software. BLASTP searches against the nr database were carried out with the amino acid sequence of the maize gene YGL-1 used as a query and its homologs from other species were retrieved from GenBank (http://www.ncbi.nlm.nih.gov). Multiple amino acid sequence alignment was conducted using DNAMAN version 5.0 software. Phylogenetic analysis was conducted by using MEGA version 5.1 with the rooted neighbor-joining tree using percentage identities based on a multiple sequence alignment generated with the DNAMAN 5.0 software.

Semi-quantitative RT-PCR
Total RNA was isolated using an RNA Isolation Kit (Tiangen Biotech). High-quality firststrand cDNA was generated using oligo (dT) and PrimeScript™ II (Takara). Amplification of ygl-1 and YGL-1 cDNA (GenBank accession: DAA43190.1) was performed using primer pairs (21JF: 5 0 -CCCAGCATCCCGAACCCAA, 21JR: 5 0 -AGATGGCGTTGTGGGAGGG). Specific primers of 19 genes used for semi-quantitative RT-PCR were designed using primer 5.0 or selected from previous study [31]. The maize actin gene (GRMZM2G126010) was used as a control in the experiment. The primer pairs were designed using primer 5.0 and shown in S1 Table. Results Characterization of the ygl-1 mutant populations all showed a segregation ratio of 3:1 (green: yellow-green plants, χ2 < χ2 0.05 = 3.84, P>0.05; Table 1). These results indicated that the yellow-green leaf phenotype in the ygl-1 mutant was controlled by a single recessive nuclear gene.
To characterize the yellow-green leaf phenotype of ygl-1, the pigment contents of total Chls, Chl a, Chl b, and Car at seedling and tasseling stages were examined (Table 2). At the seedling stage, the contents of total Chls, Chl a, Chl b, and Car were 1.65, 1.39, 0.26, and 0.34 mg g -1 in the ygl-1 mutant, significant reductions of 24.32%, 20.4%, 40.2%, and 1.96%, respectively, compared to the wild type Lx7226. At the tasseling stage, total Chls, Chl a, Chl b, and Car were  further reduced to 48.43%, 35.71%, 72.79%, and 17.19% of wild type levels, respectively. At both seedling and tasseling stages, ygl-1 mutants exhibited higher Chl a/b ratios than wild type, indicating that the reduction of Chl b was greater than Chl a in the ygl-1 mutant. These results indicated that the ygl-1 mutant phenotype was mainly because of reduced levels of total Chls, Car, and an increased ratio of Chl a/b. In addition, the significantly decreased content of Chl b in ygl-1 might play an important role in its yellow-green leaf phenotype.
To determine whether the yellow-green leaf phenotype of the ygl-1 mutant affected chloroplast development, wild-type line Lx7226 and the ygl-1 mutant leaf sections at the seedling and tasseling stages were compared using transmission electron microscopy. At the seedling stage, the ygl-1 mutant had less non-appressed granal stacks (Fig 2A), while the wild-type inbred line Lx7226 had more and larger granal stacks ( Fig 2B). Rare and less dense granal stacks appeared in the ygl-1 mutant at tasseling (Fig 2C). The wild-type inbred line Lx7226 had more, wellordered and dense granal stacks in the chloroplast ( Fig 2D). However, there was no apparent difference in the structure of the chloroplast in the ygl-1 mutant and wild-type. These results indicated that chloroplast development was suppressed in the ygl-1 mutant.
We compared the photosynthetic characteristics between wild-type line Lx7226 and the ygl-1 mutant at tasseling stage. As shown in Table 3, the net photosynthesis rate (P n ), the stomatal conductance (Cs), the intercellular CO 2 concentration (Ci) and the transpiration rate (Tr) were significantly deceased in the ygl-1 mutant compared with wild-type Lx7226. The results indicated that the capacity of photosynthesis was impaired in the ygl-1 mutant, which might be consistent with its chlorophyll deficiency and impaired development of chloroplasts.

Initial mapping of ygl-1
Preliminary mapping of ygl-1 was conducted using 231 recessive individuals showing yellowgreen leaf phenotype from an F 2 segregating population obtained from the cross of ygl-1/ Lx7226. A total of 224 pairs of SSR markers distributed over the 10 maize chromosomes, available from the public database maizegdb, were screened by polyacrylamide gel electrophoresis. We found only P3 in bin 1.01 on the short arm of this chromosome to exhibit polymorphism between the two parents and the DNA pools from the randomly selected 15 green and 15 yellow-green plants, respectively (see methods) ( Table 4). A total of 23 SSR primers located in bin 1.01 available from maizegdb were detected by polyacrylamide gel electrophoresis between the two parents and the two DNA pools. Five of the 23 primers had polymorphisms both between the two parents and the two DNA pools. All six polymorphic markers were used to screen the 231 recessive individuals of F 2 . Finally, two of them, P1 and P2, were positioned on one side of ygl-1 toward the telomeric end of chromosome 1, with 50 and 24 recombinants, respectively; the other three markers, P4, P5 and P6 were located on the other side of ygl-1 toward the centromeric region, with 10, 17 and 24 recombinants, respectively (Table 4, Fig 3A). No recombinant was screened out by marker P3, suggesting that this marker was closely linked to ygl-1.
These results indicated that the locus of ygl-1 was between markers P2 and P4, which delineated an interval of 2.6 Mb according to the reported B73 whole genome sequence (Fig 3A).
Four new SSR markers showing polymorphisms between two parents and two DNA pools   Fig 3B).

Fine mapping of ygl-1
To fine map ygl-1, four new SSR markers polymorphic between the two parents and two DNA pools were identified (  P13 (Fig 3C). Candidate gene annotation and sequence analysis Gene prediction in the 48 kb region of the maize genome inferred to contain ygl-1 revealed three genes (Fig 3D, S2 Table). One of the three genes, GRMZM2G007441 comprises only one exon and was predicted to encode a signal recognition particle 43 kDa (SRP43) protein and act as a chloroplast precursor. This protein has been reported to confer a pale green leaf phenotype in Arabidopsis [22]. Motivated by these results, we sequenced the GRMZM2G007441 gene from amplified genomic DNA of mutant line ygl-1 and six wild-type inbred lines, Lx7226, Ye478, Yuanwu02, B73, Chang7-2 and Zheng58, using primer pairs 21-3F and 21-6R, respectively. A single-nucleotide was deleted at the position 713 bp (-/G) from the ATG start codon in ygl-1, which resulted in a frame shift mutation (Fig 4). The resulting 960 bp truncated translational region encodes a predicted protein of 319 amino acids in ygl-1 compared to 1281 bp ORFs (open reading frame) encoding predicted proteins of 426 amino acids in all 6 wild-type lines (Fig 4 and S1 Fig). The conserved region of ANK_2 and ANK changed and CHROMO almost disappeared in ygl-1 compared to the 6 wild-type lines (S1 Fig). To our knowledge, the other two candidate genes GRMZM2G071349 and GRMZM2G096020 were not reported to participate in Chl metabolism or chloroplast development (S2 Table). BLAST search of the maize genome database revealed that YGL-1 (GRMZM2G007441) is a single-copy gene. We aligned the amino acid sequences of YGL-1 and its related proteins from Sorghum bicolor, Setaria italica, Oryza sativa (LOC_Os03g03990), Brachypodium distachyon, Arabidopsis thaliana (AT2G47450), Physcomitrella patens, Selaginella moellendorffii, Auxenochlorella protothecoides and Chlamydomonas reinhardtii. YGL-1 exhibits most sequence similarity to the related protein in Sorghum bicolor, with 87% identity. YGL-1 also exhibits 81%, 76%, 71%, 54%, 47%, 46%, 43% and 36% sequence similarity to proteins in Setaria italica, Oryza sativa, Brachypodium distachyon, Arabidopsis thaliana, Physcomitrella patens, Selaginella moellendorffii, Auxenochlorella protothecoides and Chlamydomonas reinhardtii, respectively (S2 Fig). We also investigated the possible phylogenetic relationships between YGL-1 and these proteins mentioned above. The results indicated that the proteins from higher plants including monocotyledonous plants (Z. mays, S. italica, O. sativa and B. distachyon), dicotyledonous plant (A. thaliana), bryophyte (P. patens) and pteridophyte (S. moellendorffii) are divided clearly into one subgroup, while the proteins from lower plants (A. protothecoides and C. reinhardtii) form another subgroup (Fig 5). These data indicate that the YGL-1 (GRMZM2G007441) is conserved in plants.

Expression analysis of YGL-1
To compare YGL-1 expression profiles, total RNA was extracted from roots, stems, leaves, ears and tassels of the ygl-1 mutant and wild-type plants, respectively. Semi-quantitative RT-PCR analysis was performed. YGL-1 mRNA was expressed in all tested tissues with relatively higher levels in ears than in the other four organs (Fig 6A). Expression levels of the YGL-1 transcript were similar in the ygl-1 mutant and wild type from early to mature stages (Fig 6B). These results indicated that the premature termination mutation in the coding region of ygl-1 did not result in variation of its own mRNA expression. Furthermore, the effect of light and dark growth conditions on the expression of YGL-1 was conducted. Transcript levels were higher when ygl-1 or wild-type plants were grown under dark than light conditions, indicating that light intensity regulates the YGL-1 or ygl-1 mRNA expression (Fig 6C).

The effect of the ygl-1 mutation on the expression of other genes
To investigate whether the ygl-1 mutation affected the transcript levels of genes associated with Chl biosynthesis, photosynthesis and chloroplast development, semi-quantitative RT-PCR primers for 19 genes related to these pathways were designed (S1 Table). Semi-quantitative RT-PCR analysis was conducted in six-week old leaves of ygl-1 and wild type plants. The expression of genes involved in Chl biosynthesis, such as elongated mesocotyl1 (elm1) and elongated mesocotyl2 (elm2), showed no difference between the ygl-1 mutant and wild type [13][14] (Fig 7). rbcS is a nuclear gene encoding the small subunit of Rubisco [40]. psbA and rbcL1 are chloroplast genes encoding reaction center polypeptides and the large subunit of Rubisco [41][42], respectively. Their mRNA levels were also not significantly different between the ygl-1 mutant and wild type (Fig 7). 7 nuclear genes (lhca1, lhcb1, lhcb2, lhcb3, lhcb6, lhcb7 and lhcb9), encoding the light-harvesting chlorophyll proteins (LHC-PI or LHC-PII) which are synthesized in the cytoplasm and after import into chloroplast, then they are targeted and inserted into thylakoid membrane, their transcript levels were also compared. 6 genes (lhca1, lhcb1, lhcb2, lhcb6, lhcb7 and lhcb9) were at a similar transcript level in the ygl-1 mutant and wild type (Fig 7). lhcb3, encoding the light harvesting chlorophyll a/b binding protein of PSII, was altered in the ygl-1 mutant, with no expression or too low to be visible (Fig 7). Interestingly, 6 other genes participating in chloroplast development or function were also significantly different in the ygl-1 mutant and the wild type. vyl-Chr.1 and vyl-Chr.9, encoding the ClpP5 subunit of Clp protease which was important for chloroplast development or function [15,43], were slightly decreased in the ygl-1 mutant (Fig 7). csr1 (chloroplast SRP receptor1) encoding cpFtsY important for efficient thylakoid targeting of LHCPs [35], showed severely reduced mRNA level in the ygl-1 mutant (Fig 7). X1, which was predicted to encode Zea mays signal recognition particle 54 kDa protein and act as a chloroplast receptor (GenBank), showed Fine Mapping and Candidate Gene Analysis significantly increased expression in the ygl-1 mutant (Fig 7). hcf60, a nuclear gene, encodes the chloroplast ribosomal small subunit protein 17 [33]. hcf106, another nuclear gene, is required for post-translational steps in the biogenesis of the chloroplast cytochrome b6f complex in maize [44]. Their transcript levels were also significantly increased in the ygl-1 mutant (Fig 7). hsp70, encoding a stromal chaperone protein which facilitates protein folding and plays a role in protein translocation across membranes, did not appear to be significantly affected in ygl-1 [45] (Fig 7). Overall the ygl-1 mutation might mainly affect the mRNA levels of some genes involved in chloroplast development but not genes associated with Chl biosynthesis or photosynthesis. These results suggested that the expression of these genes associated with chloroplast development might be regulated at the translational or post-translational levels, since the mutation of YGL-1 did not alter its own mRNA expression in ygl-1 (Fig 6A and 6B).

Discussion
In recent years, researchers have paid increasing attention to leaf-color mutations, with important insights into the mechanisms of these mutations gained by study of several organisms. In maize, the molecular mechanisms of leaf-color mutation and the related responsible loci are still far from being fully understood. Mutant analysis is an effective approach to explore the function of genes in chloroplast development.

SRP43 confers the yellow-green mutation in ygl-1
Taking a positional cloning approach, the ygl-1 gene was mapped to an interval of 48 kb in bin 1.01 on the short arm of chromosome 1. To our knowledge, only vyl-Chr.1 related to leaf-color Fine Mapping and Candidate Gene Analysis mutation has been previously mapped and cloned on the same chromosome [15]. Most homozygous vyl-Chr.1 plants died within 3 weeks of emergence and the surviving plants were paleyellow and very small. In contrast, homozygous ygl-1 plants were normal and can grow to Fine Mapping and Candidate Gene Analysis maturity. In addition, the vyl-Chr.1 gene is located in bin 1.03, not in the 48 kb region in bin 1.01 of ygl-1, so we assert that ygl-1 is a novel gene conferring leaf-color mutation in maize. The chloroplast signal recognition particle (cpSRP), composed of one cpSRP43 dimer and one cpSRP54 monomer, is required for binding the light-harvesting chlorophyll protein (LHCP) to form a cpSRP/LHCP transit complex which targets LHCP to the thylakoid membrane [46][47]. Mutants in SRP54 have been isolated in Arabidopsis and rice. The mutant ffc, which was generated by ethyl methane sulfonate (EMS) in Arabidopsis, had yellow first true leaves that subsequently turned pale green [48]. A mutant in SRP54 of rice, designated as YGL138(t), derived from Nipponbare treated by EMS, showed yellow-green leaf phenotype throughout development [21]. Could a mutant in SRP43 also lead to chlorophyll deficient mutation? The chaos mutant of SRP43 in Arabidopsis, resulting from insertion of an enhancer trap Ds element, showed pale leaves throughout the vegetative cycle [49,22]. The double mutant ffc/chaos had pale yellow leaves from early stage to maturity [50]. Mutations in CrcpSRP43 would lead to a low Chls content, high Chl a/b ratio and leaf-color phenotype in C. reinhardtii [51]. The two cpSRP43 mutants in Arabidopsis and C. reinhardtii showed a similar and specific decline of LHCPs in the developed chloroplast, indicating that the cpSRP43 protein plays a conserved function in green microalgae and higher plants, which is also supported by our phylogenic analysis. To our knowledge, no mutant in SRP43 has yet been reported in model monocotyledonous plants including rice and maize. We postulated that mutation in cpSRP43 of maize will result in low levels of Chls and Car, higher Chl a/b ratio and a leaf-color phenotype. In our study, a leaf-color mutant ygl-1 was isolated and it developed yellow-green leaves at all stages. The ygl-1 gene was fine mapped to a 48 kb region in the maize genome. Three candidate genes were identified in the 48 kb chromosomal region of ygl-1. One of these, GRMZM2G007441, was predicted to encode a cpSRP43 protein and act as a chloroplast precursor. We suspect that some mutational site in the GRMZM2G007441 genomic sequence which leads to mutation in the encoded product exists in ygl-1. Sequencing results revealed that 1-bp was deleted in the coding region of this gene in ygl-1, and the single nucleotide deletion led to a frame shift mutation. ygl-1 mutants showed significantly reduced contents of Chls, Car, and higher Chl a/b ratios than wild-type plants at both seedling and tasseling stages. We presumably assume that the defect in ygl-1 lead to the chlorophyll deficiency in the ygl-1 mutant, which makes it to exhibit yellow-green leaf phenotype throughout the growth period. In addition, higher Chl a/b ratios in ygl-1 might make the significantly decreased content of Chl b to play an important role in its yellow-green leaf phenotype. These results provided further evidence that GRMZM2G007441 was the gene responsible for ygl-1 in maize.

SRP43 is an important component of the SRP protein and plays a central role during chloroplast development
The cpSRP is a collection of four proteins that work together: cpSRP54, cpSRP43, cpFtsY and ALB3 [36]. These SRP proteins are believed to be involved in the proper folding of the LHCP S and their targeting to the thylakoid membrane [52]. In the chloroplast, cpSRP uses two mechanisms to target both chloroplast-and nuclear-encoded LHCP S to the thylakoid membrane [53]. The first pathway was called co-translational targeting. This system, which targets chloroplast-encoded proteins to the thylakoid membrane, requires cpSRP54 and the conserved SRP RNA of the bacterial and cytosolic ribonucleoprotein SRPs. The second pathway called posttranslational targeting, targets nuclear-encoded proteins to the thylakoid membrane using cpSRP54 together with cpSRP43, but lacks the conserved SRP RNA of the bacterial and cytosolic ribonucleoprotein SRPs.
In the latter targeting pathway, the cpSRP43 works as a novel chaperone specific for lightharvesting chlorophyll a, b-binding proteins and its role cannot be taken over by other stromal chaperones [54][55]. This protein alone is sufficient to form a soluble complex with LHCPs and to prevent aggregation [54]. This supports the idea that cpSRP43 alone is sufficient to targeting of LHCPs to the thylakoid membrane [56]. The cpSRP43 protein is also able to dissolve aggregated LHCPs [54]. This ability of cpSRP43 might be advantageous if LHCP import exceeds the capacity of cpSRP43 to form the transit complex.
The cpSRP43 consists of ANK repeat domains and CHROMO domains [54]. The cpSRP43 ANK repeat domains provide the binding site for the conserved DPLG motif in LHCPs and are sufficient for the chaperone function, whereas the chromo domains are dispensable. In our study, the cpSRP43 protein was 319 amino acids long in the ygl-1 mutant, in which the ANK repeat domains changed a lot and the CHROMO domain almost disappeared due to the 1-bp deletion in the coding region which led to a premature termination codon. We hypothesize that the deficiency chlorophyll, the impaired development of chloroplasts, the decreased capacity of photosynthesis and the yellow-green leaf phenotype of ygl-1 might be due to the changed ANK repeat domains of cpSRP43, which might be incapable of binding the conserved DPLG motif in LHCPs to function as chaperone correctly. In our semi-quantitative RT-PCR results, lhcb3, encoding light harvesting chlorophyll a/b binding protein of PSII, was specifically suppressed in ygl-1 plants compared with the wild-type plants, whereas all other components of PSI (lhca1) and PSII (lhcb1, lhcb2, lhcb6, lhcb7 and lhcb9) remained unchanged. These findings indicate that at least lhcb3 is dependent on the cpSRP43 for targetting, other 6 proteins are either less dependent or independent of cpSRP43. This observation supports the idea that mutations in cpSRP have specific effects on a subset of LHCPs [48]. hcf106, encoding an integral membrane protein which is required in 4pH pathway for targeting proteins to the thyakoid membranes [44,[57][58][59], showed significantly increased expression in the ygl-1 mutant. The increase in hcf106 might occur in response to the defect of functional cpSRP43. hsp70, a nuclear gene encoding a stromal chaperone protein [45], remained a similar transcript level in ygl-1 and wild type. This result is in consistent with previous reports in Arabidopsis ffc and chaos mutants [48]. In addition, the stable level of some gene transcripts (participating in Chl biosynthesis such as elm1 and elm2, or photosynthesis such as rbcS, psbA and rbcL1) and the differently affected level of other gene transcripts (involving in chloroplast development such as vyl-Chr.1, vyl-Chr.9, csr1, X1 and hcf60), probably caused by a feedback mechanism, suggest that ygl-1 mainly altered the transcript levels of genes involved in chloroplast development. The ygl-1 mutant had less non-appressed granal stacks and rare and less dense granal stacks at the seedling and tasseling stages, respectively. The suppressed development of chloroplasts and yellow-green leaf phenotype in the ygl-1 mutant is likely due to the apparent up/down-regulations of genes (hcf106, vyl-Chr.1, vyl-Chr.9, csr1, X1 and hcf60) that participated in chloroplast development, which response to the defect of ygl-1 function that are responsible for the normal chloroplast development. However, transcript levels of genes that involved in Chl biosynthesis and photosynthesis were stable, which makes why the contents of Chls and Car were reduced and the capacity of photosynthesis was decreased is still not very clear.
In the higher plant post-translational pathway, the cpSRP43 binds to the imported LHCPs, then cpSRP54 binds to the cpSRP43-LHC protein complex. This complex subsequently binds to a membrane-bound cpFtsY, and finally this LHC-cpSRP43-cpSRP54-cpFtsY complex is guided to the ALB3 protein [34]. In the co-translational pathway, cpSRP54 binds to a chloroplast-encoded protein that is newly synthesized by a ribosome, then cpFtsY guides this complex to the ALB3 protein that works in tandem with cpFesY. Our semi-quantitative RT-PCR results revealed that csr1, which encodes cpFtsY [35], showed severely suppressed transcript level in the ygl-1 mutant. X1, which was predicted to encode cpSRP54 and act as a chloroplast receptor (GenBank) showed significantly increased expression in the ygl-1 mutant. These results suggest that there might be alternative explanations for the respective interactions of these genes with cpSRP43. In addition, the ygl-1 mutant does not show a developmental effect on chlorophyll content, chloroplast ultrastructure, or yellow-green leaf phenotype. The transcript level of YGL-1 remained at a similar level from 1-week to 10-week stages both in ygl-1 and wild type, indicating that the requirement for SRP43 is not dependent on plant age. Taken together, we suggest that the cpSRP43 protein plays a central role in the post-translational targeting pathway during chloroplast development.

Conclusion
In this study, a novel yellow-green leaf mutant ygl-1 was isolated from maize. The mutant spontaneously showed a distinguishable yellow-green leaf phenotype throughout the growing period and was controlled by a single recessive nuclear gene. The YGL-1 locus was delimited within a 48 kb region in the maize genome harboring three candidate genes. A single base pair deletion discovered in cpSRP43 might lead to reduced contents of Chls and Car, an increased ratio of Chl a/b, defective chloroplast development and declined capacity of photosynthesis, which conferred the yellow-green character in ygl-1. The mutation in ygl-1 might regulate the expression levels of some genes associated with chloroplast development at translational or post-translational levels, while its own mRNA level was not affected. Our results are a valuable foundation for further study and functional characterization of ygl-1, which will facilitate dissection of the mechanism of cpSRP43 protein involvement in chloroplast development of maize.