Identification and Fine Mapping of a White Husk Gene in Barley (Hordeum vulgare L.)

Barley is the only crop in the Poaceae family with adhering husks at maturity. The color of husk at barely development stage could influence the agronomic traits and malting qualities of grains. A barley mutant with a white husk was discovered from the malting barley cultivar Supi 3 and designated wh (white husk). Morphological changes and the genetics of white husk barley were investigated. Husks of the mutant were white at the heading and flowering stages but yellowed at maturity. The diastatic power and α-amino nitrogen contents also significantly increased in wh mutant. Transmission electron microscopy examination showed abnormal chloroplast development in the mutant. Genetic analysis of F2 and BC1F1 populations developed from a cross of wh and Yangnongpi 5 (green husk) showed that the white husk was controlled by a single recessive gene (wh). The wh gene was initially mapped between 49.64 and 51.77 cM on chromosome 3H, which is syntenic with rice chromosome 1 where a white husk gene wlp1 has been isolated. The barley orthologous gene of wlp1 was sequenced from both parents and a 688 bp deletion identified in the wh mutant. We further fine-mapped the wh gene between SSR markers Bmac0067 and Bmag0508a with distances of 0.36 cM and 0.27 cM in an F2 population with 1115 individuals of white husk. However, the wlp1 orthologous gene was mapped outside the interval. New candidate genes were identified based on the barley genome sequence.


Introduction
Barley (Hordeum vulgare L., 2n = 2H = 14) is the only Poaceae family crops with its caryopsis adhering to husk, which provides economically important source of human and animal nutrition and underpins the malting and brewing industries. Barley also can be used as a genetic model species for Triticeae genomics [1]. In general, the husk is green, which not only protects the flower and grain from pathogens, insects, and external environments, but also is the essential photosynthetic organ to accumulate dry matter of grain [2,3]. Studies have shown that photosynthesis in husk occurs via C3 and C4 pathways, and they can provide carbohydrates and amino acids for grain growth [4,5]. Therefore, the changes in husk color could affect on the compositions of the grain and the yield of barley.
The white husk is a special type of husk color variation. In barley, a white husk mutant was first described in 1958 as a natural mutant from an unknown variety [6]. Genetic analysis of this mutant identified that the white husk characteristic is controlled by a recessive nuclear gene named alm, located on the short arm of chromosome 3H using a full set of morphological traits as markers [7][8][9]. Linkage maps, including the alm gene, have been constructed on the basis of morphological and molecular markers [10]. However, these maps exhibit low density near the alm gene. In rice, five white husk genes have been identified and they are wp1 [11], wp2 [12], wp(t) [13], wslwp [14], and wlp1 [15]. To date, only wlp1 has been cloned and one base substitution of C to T in the coding region of chloroplast ribosome L13 protein could result in white husk and leaf under low temperature [15].
The recent release of the barley draft genome assembly [16] provides a resource for pursuing the target gene, however, the precise genomic sequence is still not available. Early molecular genetic studies, based on sets of cloned DNA probes hybridizing to genomic DNAs of different species, have further developed our understanding of the conservation of gene synteny within the grass family, and lead to the concept of alignment of different grass genomes at the macro level [17][18][19][20]. Since completing the draft rice genome sequence [21], rice has been successfully used as a model crop to isolate functional genes in barley based on syntenic cloning [22][23][24]. In this study we report the discovery of a white husk barley mutant, its husk morphological features and genetics. The husk gene was initially mapped to chromosome 3H and a candidate gene identified using the syntenic cloning approach. Further fine-mapping identified that the rice orthologous gene is not the functional gene for the mutant in barley. These results will pave the way for map-based cloning of the target gene and unraveling the molecular mechanisms of the mutant. The white husk gene/trait will be utilized as a visualized marker in barley molecular breeding in future.

Plant materials
A mutant with a white husk at the heading and flowering stages was identified from a commercial malting barley variety Supi 3 in the field of Jiangsu Province of China in 2006 and designated wh (white husk). The mutant wh was crossed with a green husk variety Yangnongpi 5 to obtain F 1 hybrid seeds in 2009. The F 1 was then crossed with wh and Yangnongpi 5, respectively to produce BC 1 F 1 progenies. The F 2 populations, BC 1 F 1 progenies, their F 1 and parents were planted in the field at Zhejiang Academy of Agricultural Sciences, Hangzhou, China to observe the husk color of individuals at the heading stage, and then to analyze inheritance of the white husk.
Two hundred and twenty-eight individuals with white husks from the wh/Yangnongpi 5 F 2 population were used for preliminarily mapping of the target gene. One thousand one hundred and fifteen individual plants with white husks of an enlarged F 2 population (4657 individuals) were used to fine-map the white husk gene.

Agronomic traits and malting quality analysis
Agronomic traits (plant height, 1000-grain weight and maturity) of wh and its wild-type Supi 3 were investigated as described by Zhang and Liu [25]. Micro-malting and malting quality analysis was conducted in the laboratory of the National Barley Improvement Centre (Hangzhou, China) using 120 g of grain from the sieving fraction >2.2 mm. The malting process and determination of malting quality such as malt extract, viscosity, diastatic power, α-amino nitrogen and kolbach index were performed as described by Wang et al. [26].

Morphological analyses of white husks by SEM and TEM
The lemmas of wh and its wild-type Supi 3 at the booting and heading stages were obtained, cut into small pieces (2 mm) and fixed in 2.5% glutaraldehyde (v/v) in 100 mM phosphate buffer (PBS, pH 7.0) for 6 to 8 h. The samples were washed three times with the same PBS, post-fixed in 1% osmium tetroxide for 1 h, and dehydrated with an ethanol series. Some of the samples were then treated for scanning electron microscopy (SEM), with the remaining samples used for transmission electron microscopy (TEM). SEM samples were initially transferred to an alcohol and iso-amyl acetate mixture (v:v = 1:1) for approximately 30 min and then to pure iso-amyl acetate for approximately 1 h. The specimens were dehydrated using a Hitachi Model HCP-2 critical point dryer with liquid CO 2 . The dehydrated specimens were coated with gold-palladium and observed using a Philips Model TM-1000 SEM. TEM samples were treated with gradient ethanol concentrations and transferred to absolute acetone for 20 min. The specimens were then placed in a 1:1 absolute acetone and Spurr resin mixture for 1 h at room temperature, before being transferred to a 1:3 mixture of absolute acetone and resin for 3 h. The resulting sample was transferred to the final Spurr resin mixture and stored overnight. The specimen was stained with uranyl acetate and alkaline lead citrate, and ultrathin sections (80 nm) were prepared. These sections were examined using a TEM (Hitachi H-7650, Japan).

Genomic DNA extraction
Seedling leaves from F 1 , F 2 , and parent plants were collected and frozen in liquid nitrogen for DNA extraction. The cetyltrimethylammonium bromide (CTAB) method was used to extract genomic DNA from 0.6 to 1.0 g of young leaf tissue [27]. The quality and concentration of DNA was estimated using a NanoDrop ND-2000 spectrophotometer (Thermo Scientific, Walthman, MA, USA). DNA samples were diluted with double-distilled H 2 O to obtain a final concentration of 100 ng/μL and stored at -20°C. Bulk segregant analysis was initially used to map the wh gene. Two DNA pools were assembled using equal amounts of DNA from 10 F 2 plants with white or green husks.

SSR marker selection
A total of 1046 SSR markers [28][29][30][31][32] distributed in the seven chromosomes of barley were used to screen for polymorphism between the two parental lines (wh and Yangnongpi 5). Polymorphic SSR markers were first used to analyze the two DNA pools. The polymorphic markers between the two bulks were used to determine the genotype of individuals from F 2 plants with white husks.
Polymerase chain reaction (PCR) was performed in a final reaction volume of 10 μL containing 1 μL of 100 ng/μL genomic DNA, 1 μL of 10 pmol/L solution of forward and reverse primers, 3 μL of sterile ddH 2 O, and 5 μL of 2× Taq Master Mix (GeneSolution, Shanghai, China). PCR products were separated on 8% non-denaturing polyacrylamide gels and visualized by silver staining.

Data analysis
Chi-squared (χ 2 ) tests for goodness-of-fit were performed to evaluate the deviations of observed data from theoretically-expected segregation ratios. Polymorphic molecular markers and the white husk gene were subjected to linkage analysis using JoinMap 3.0 software with LOD threshold of 4 [33]. Recombination frequencies were converted to map distances in centi-Morgans (cM) by the Kosambi function [34].

Comparative genomic analysis
SSR markers flanking the target gene were first integrated into the barley consensus map [35]. Sequences of the molecular markers in the regions were used to search the barley genomic database (http://webblast.ipk-gatersleben.de/barley/viroblast.php) to identify corresponding barley contigs. The markers flanking the white panicle genes in rice were mapped to the corresponding physical map (http://archive.gramene.org/markers/), and the genes around these markers identified on the Rice Genome Browser website (http://rice.plantbiology.msu.edu/cgibin/gbrowse/rice/). For collinearity analysis of barley and rice, the sequences of these rice genes were used to blast barley genomic databases and to find their orthologous loci. Where several significant hits were found, only the best hit was adopted.
A PCR reaction was conducted using a KOD-FX kit (TOYOBO, Japan) comprising 6 μL of 2 mM dNTP, 30 μL of 2× buffer, 1.2 μL of 1 U/μL KOD-FX polymerase, 3.6 μL of 10 pmol/L solution of forward and reverse primers, 6 μL of 100 ng/μL genomic DNA, in a final volume of 60 μL with sterile distilled water. PCR cycling conditions consisted of an initial denaturation setup of 94°C for 4 min, followed by 35 cycles of 94°C for 30 s, 55°C for 30 s, 72°C for 2 min 30 s, and a final extension cycle at 72°C for 10 min. At least two independent PCR products were sequenced by BGI Tech Solutions Co., Ltd (Shenzhen, China).

Discovery of a white husk barley
In 2006, a mutant was discovered in a field of malting barley cultivar Supi 3, a leading cultivar in the Jiangsu Province, China. The husks in the mutant were colorless at the heading and flowering stages ( Fig 1A) and became yellow but slightly brighter than Supi 3 in the mature grains. Husk color of mature grains of Supi 3 and wh were measured with a Konica Minolta CM-700d over two consecutive years; the mutant values for Min L (brightness) were 2.4 units higher than Supi 3 (Table 1), indicating brighter grains. Multi-year and -location experiments showed that the white husk trait is stably inherited.
In contrast to the wild-type barley, wh had yellow spots on its leaves during vegetative growth (Fig 1B). Plant height in wh was 7-14% shorter than its wild type in two consecutive years (Fig 1C). The 1000-grain weight of wh was 82-87% and its maturation period was also 5-7 d earlier than its wild-type barley. Furthermore, the diastatic power and α-amino nitrogen contents were much higher in the wh than in Supi 3 grains (Table 1).
Scanning electron microscopy showed that mutant grains had more dense trichomes than wild-type barley grains. Yet, they were not taller or plumper; projections could not be formed on top in wh lemma surface (S1 Fig). Differences were evident in morphological characteristics and internal structure of the chloroplast between wild-type and mutant lemma at the booting and heading stages (Fig 2). The wild type had oblong chloroplasts with a dense internal structure, and normal growth at these two stages. The laminated structure of the thylakoids was regular (Fig 2A and 2B). Mutant chloroplasts had numerous empty vesicle-like structures at the booting stage, and abnormal growth. The laminated structure of the thylakoids was dispersed, and the chloroplasts produced cysts. At the heading stage, no laminated structure was observed and the membrane structure of the chloroplast was destroyed. Thus, serious growth defects were observed in the mutant barley chloroplasts (Fig 2C and 2D).

Genetic analysis of white husk barley
At the heading stage, wh husks were white, while husks of Yangnongpi 5 are normally green. When wh and Yangnongpi 5 were reciprocally crossed, F 1 generation plants had green husks. Thus, the white husk trait for wh is controlled by a recessive nuclear gene. Among the 970 F 2 plants of wh/Yangnongpi 5, 742 single plants had normal green husks and 228 had white husks. The segregation ratio fits the 3:1 model (χ 2 3:1 = 1.08, P > 0.3). Of the 93 BC 1 F 1 plants of wh/Yangnongpi 5//wh backcrossing, 50 single plants had normal green husks, and 43 had white husks; the segregation ratio was 1:1 (χ 2 1:1 = 0.39, P > 0.5). All 97 BC 1 F 1 single plants from the wh/Yangnongpi 5//Yangnongpi 5 backcrossing had normal green husks. These results confirm that the white husk trait is controlled by a recessive nuclear gene. Table 1. Characterization of wh grains. Husks in mature grains of wh were much brighter, as diastatic power and α-amino nitrogen contents significantly increased, even though malt extract decreased slightly.
Year Material L values Viscosity (mPaÁs) Malt extract (%) Diastatic power (WÁK) α-amino nitrogen (mg/100 g) Kolbach index (%)  Mapping the white husk gene wh Nine hundred and forty-four pairs of the published SSR primers synthesized in our laboratory were screened for polymorphism between the parents wh and Yangnongpi 5. Among these primer pairs, 95 were polymorphic between parents. The polymorphic SSR markers were used to test for polymorphism in the two DNA pools. Five SSR markers-namely Bmac0209, Bmag0138, Bmac0871, Bmag0508a, and Bmag0122-located on chromosome 3H were polymorphic between the two DNA pools. The five polymorphic markers were used to analyze the 228 individuals with white husks from the F 2 population. Linkage analysis showed that wh was located on the short arm of chromosome 3H.

Collinearity analysis of the white husk genomic region in barley and rice
All markers (S1 Table) in the interval with sequences were used to search for corresponding barley contigs (http://webblast.ipk-gatersleben.de/barley/). The markers were mapped to an interval between 45.93 and 51.77 cM (Fig 3A and 3B) on chromosome 3H of the POPSEQ map [36]. Sequences of all barley genes in the interval were retrieved and used to BLAST search the syntenic rice genes. The genomic region in barley chromosome 3H showed good synteny with rice chromosome 1 with conserved gene order (Fig 3). Of the five white panicle genes in rice [11][12][13][14][15], wp(t) and wlp1-mapped to rice chromosome 1-were located in the syntenic region of the barley mutant gene wh on chromosome 3H.

Isolation of the orthologous gene of wlp1 in wh and its wild type
Of the five white panicle genes in rice, wlp1 was only one cloned and its full sequence was 2402 bp, which included four exons and three introns. By the sequence blast, the barley contig MC_48768 shared 92% identity with the genomic DNA sequence of wlp1. MC_48768 contained a gene model (MLOC_64398.1) annotated as 50S ribosomal protein L13 (http://apex. ipk-gatersleben.de/apex/f?p=284:20:0::NO::P20_GENE_NAME:MLOC_64398.1). In order to establish whether this gene was wh, we designed two forward primers at 1781 bp and 119 bp upstream of the initiation codon and two reverse primers at 171 bp and 2364 bp downstream of the initiation codon to amplify, respectively, the promoter and ORF of the orthologous gene of wlp1 in wh and its wild type. The PCR products were 1953 bp and 2484 bp, respectively (S2

Fine-mapping the wh gene
To fine-map the wh gene, we expanded the F 2 population of wh/Yangnongpi 5 to 4657 individual plants and used 1115 individuals with white husk of them to test the five polymorphic markers in the preliminary mapping results. Meanwhile, to identify more molecular markers linked with this gene, we screened an additional 102 published SSR primer pairs and EST-SSR primers on chromosome 3HS. Two additional markers Bmac0067 and Bmag0361 were linked with the wh gene, and the wh gene was mapped between Bmac0067 and Bmag0508a markers with genetic distances from wh of 0.36 and 0.27 cM, respectively (S2 Table) (Fig 3A). However, the wlp1 orthologous gene was mapped 2.07 cM from the wh gene. The fine-mapping results demonstrated that the barley wh is not the rice wlp1 orthologous gene. Further analysis indicated that the rice genomic region around wp(t) between 7009 kb and 11720 kb was co-linear with the barley genomic region between MC_1529 and MC_114580, which is more than likely the orthologous gene for wh (Fig 3). However, no functional gene has been identified for the wp(t) locus.

Discussion
In this study, we reported a white husk mutant wh from the malting barley cultivar, Supi 3. By SEM and TEM observations, we found that the external and internal structures of the lemma  [13], and (E) genetic map of wlp1 according to [15]. in wh differed from the wild type. Thus, the white husk of wh may be due to abnormal chloroplast development. As compared to Supi 3, the white husk in wh mutant lead to the reduction of 1000-grain weight and malt extract, and the increase of the diastatic power and α-amino nitrogen contents. It was probably due to that the abnormal chloroplast metabolism in white husk influenced the photosynthesis of husks, which then had some impacts on the weight and malting qualities of grains. In addition, the plant height was significantly shorter and the maturation period was about 7 days earlier in wh. It was unknown whether these phenomena were caused by white husk. However, the similar cases have also reported in other mutants. For example, a barley stage green-revertible albino mutant whs18 was reported, in which the plant height was obviously shorter and its heading date was about 8 days later than its wild type [37].
We mapped the white husk gene wh between Bmac0067 and Bmag0508a on the barley chromosome 3H. A spontaneous mutant Russia 82 was reported with albino lemma (alm) gene in 1958 [6]. Costa et al. [10] used the albino lemma trait (alm) as an anchor marker to construct a molecular linkage map and found that the AFLP marker E38M62-81 co-segregated with the alm locus. In our study, we tested the polymorphism between the parents using the marker E38M62-81 but no polymorphism was found in our population. More research is needed to determine whether alm and wh are allelic.
In previous studies, many leaf color mutants have been reported, such as albino, viridis and xantha [38]. Of these, the white leaf is an extreme trait that is unable to trap light energy. Therefore, most white leaf mutants cannot survive. Although extensive attention has been paid to albinism, the mechanism of mutation and the responsible loci are not fully understood at the molecular level. In Arabidopsis, four genes-CLA1 [39,40], ALB3 [41], ALB4 [42] and CLB6 [43]-which all result in the albino phenotype were isolated; these genes encode 1-deoxyxylulose 5-phosphate synthase, a chloroplast protein, thylakoid membrane protein and 1-hydroxy-2-methyl-butneyl 4-diphosphate reductase, respectively. CLA1 and CLB6 play important roles in the methy-D-erythritol 4-phosphate pathway. ALB3 may have an important function in the assembly of a chloroplast enzyme complex while ALB4 is involved in proper chloroplast biogenesis. In rice, Gothandam et al. [44] transferred the antisense OsPPR1 gene and all transgenic plants were albino, suggesting that the PPR protein is required for chloroplast biogenesis. Su et al. [45] reported a ysa mutant which appeared albino before the threeleaf stage, but gradually turned green from the four-leaf stage. The ysa gene was isolated using map-based cloning and shown to encode a new member of the superfamily of PPR proteins. In barley, Qin et al. [37]reported that a stage green-revertible albino mutant whs18, which could be caused by the substitution from C to A in the third exon of fructokinase-1-like gene generated a premature stop codon in whs18. The husks are unique organs in the Poaceae that protect the inner floral organs and kernels from environmental stresses; they are also photosynthetic organs that supply carbohydrates and amino acids for grain growth and development [46,47]. Song et al. [15] reported a wlp1 (white leaf and panicles 1) mutant in rice with albino early seedling leaves, and albino immature panicles at the heading stage. These abnormal leaf and husk color was induced by low temperature. The wlp1 gene encodes a 50S ribosome L13 protein which is required for normal chloroplast development, and is the only cloned gene related to the white panicle trait in the Poaceae. The gene was located in the syntenic region of the wh gene from the initial comparative analysis. It would be logical to speculate that the wlp1 is the orthologous gene of barley wh, as a large deletion was identified in the promoter region between wh and Yangnongpi 5 (S2 Fig). However, fine-mapping excluded this gene as the functional gene for the wh mutant. Thus, it should be worth to revisit some genes identified through syntenic map. In some cases, fine-mapping may be required.
We mapped the wh gene into a 0.6 cM interval (Fig 3). There are 40 high confident genes (S3 Table) in this interval based on the POPSEQ [36]. The mutant wh showed abnormal chloroplast development (Fig 2), but no obvious gene has been identified in chloroplast development in this interval. The four genes encoding zinc finger family protein, histone H2B, nucleic acid-binding OB-fold-like protein, and ATP-dependent RNA helicase may be reasonable candidates for the wh gene. Further research is required to identify the functional gene for the wh.  Table. The genotyping information for recombinants in F 2 population of wh/Yangnongpi 5. (XLSX) S3 Table. All genes in the target region based on the POPSEQ. (XLSX)