Introduction

The genus Brassica includes different vegetables and oil crops that are grown in different countries. Chinese kale (Brassica oleracea var. Alboglabra) was originated in Southern China. This vegetable was cultivated on nearly 50000 ha in Guangdong Province of China during 2014 (Li et al. 2015). Commonly, three to four crops of Chinese kale are produced in a single year. Along with the domestic consumption, Chinese kale is also exported to Asia, Europe and USA.

Bolting (BT) and flowering (FT) time are important agronomic traits of crop quality, production (Zhang et al. 2018). Understanding genetic mechanisms governing bolting and flowering time is an essential factor in plant breeding (Oded et al. 2014). Plant flowering is a series of developmental events. Budding and bolting mainly controls flowering time (FT) of plants. Several genes control FT trait that is an important determinant for the evolution and domestication. Genes controlling FT trait have been explored in Arabidopsis thaliana (Koornneef et al. 1998). Nearly 80 genes have been shown to control the flowering process either by direct or indirect mechanisms (Levy and Dean 1998). During the entire growth period of Chinese kale, BT and FT traits directly influence the time of harvest, yield and quality.

Genetic map and QTL analysis are important approaches for marker assisted breeding of plants. High-density genetic map is an essential resource for mapping of the phenotypic traits of interest (Liu et al. 2016a, b). Next-generation sequencing techniques have leveraged the advantages of molecular markers. SLAF sequencing, when assisted by next-generation sequencing technologies, is a useful strategy for discovering single nucleotide polymorphisms (SNPs) (Li et al. 2019). SLAF-seq has significantly improved the resolution of QTL mapping in numerous crops by large-scale SNPs discovery (Sunet al. 2013b). Previously, different QTLs controlling flowering time in B. oleracea have been identified using genetic linkage maps developed by crossing the members of different sub-species and old molecular markers such as restriction fragment length polymorphisms (RFLP), amplified fragment length polymorphisms (AFLP) and simple sequence repeats (SSR) (Li et al. 2015; Honghao et al. 2016; Rahman et al. 2018). These maps cannot serve effectively for molecular assisted selection of traits for breeding Chinese kale as these were of low genetic density. Hence, there is a need to develop a high-density genetic linkage map that can be used to find loci governing BT and FT traits in Chines kale plants. The availability of a high-density genetic map of Chinese kale along with the presence of reference genome of Arabidopsis can further assist in the identification of novel genetic resources controlling the flowering time trait of Chinese kale.

Keeping in view the importance of high-density genetic resources, it deems necessary to build a genetic map of relatively high density to perform QTL mapping of flowering traits of Chinese kale. Hence, this study was carried out to identify the loci controlling BT and FT traits in Chinese kale plants, construction of a high density genetic map and candidate gene analysis.

Materials and Methods

Mapping population

An F₂ population was raised by crossing ‘Dasun’ and ‘LB07M’ lines of Chinese kale. The line ‘Dasun’ has long bolting (BT) and flowering (FT) times and was the male in the cross. ‘LB07M’ has BT and FT and acted as female during cross. An F₂ population consisting on 159 lines was consisted derived from single seed.

Phenotypic evaluation of flowering and bolting time

The two parents (30 plants of each), F₁ (30 plants) and the F₂ plants were sown on 26 September of 2018 year in the field at Research station of Guangdong Academy of Agricultural Sciences, Guangzhou (25.1N, 115.0E), China. The soil at the research station is a sandy loam with adequate fertility. Field layout consisted of two-row plots, 20 (L) × 25 (W) cm, with planting following a randomized block design. The flowering and bolting times were evaluated using the values of each line and parent. Two rows were planted in each plot for avoidance of errors caused by the border effects. BT was noted as the day when the main flower stalk had grown to 1.0 cm. FT was noted when the first flower was fully opened.

DNA extraction

Young leaves from F₂ individuals and two parental lines were collected. For SLAF library construction, DNA was extracted using a Sangon Dzup plant DNA extraction kit following the provided instructions. DNA quality was assessed by standard electrophoresis method.

SLAF library construction and data analysis

SLAF sequencing was performed in accordance to the methodology proposed by Sun et al. (2013a). Briefly, preliminary studies were performed to select suitable restriction enzymes for the production of restriction fragments and high-quality SLAFs. Different restriction enzymes were screened in a pilot study and finally the genomic DNA of Chinese kale was digested with “HaeIII” and “Hpy166II” enzymes. Afterwards, a single nucleotide overhung was joined with the obtained fragments and duplex tag-labeled sequencing adaptors were ligated with “A-tailed” fragments. SLAFs (DNA fragments) of 400 to 500 bp were separated and sequenced on an Illumina HiSeq 2500 system.

The identification and genotyping of SLAF markers were performed according to the procedure of Sun et al. (2013a). Raw reads with < Q30 scores were removed. The left-over reads were allotted to the plant samples according to the duplex barcode sequences and clustered on the basis of sequence homologies perceived by BLAST analysis.

Genetic linkage map construction

The genetic map was constructed in accordance to the Wu et al. (2018). The incomplete significant segregation distortion markers were removed, leaving 4,337 Marker loci showing MLOD values larger than 5.0 were used for linkage group (LG) creation (Zhang et al. 2015). By calculating the MLOD value between two tags, the labels with MLOD value less than 3 of other SLAF tags were filtered out; a total of 4,296 loci was generated in the 9 linkage groups. Efficient maximum likehood estimation method was used for each chromosome to construct a genetic map using HighMap software.

QTL analysis

QTL analysis was performed by rQTL using an interval mapping method (Broman et al. 2003). The walking speed of 1 cM was set during composite interval mapping (CIM) (Zhiwen et al. 2012). The statistical likelihood ratio (LOD score) was calculated to denote the significance of each QTL interval. Afterwards, the threshold of the LOD score for significance (P=0.05) was sorted using 10,000 permutations as suggested by Liu et al. (2016a, b).

Annotation of candidate genes

The candidate genes found in major QTL regions were annotated by performing sequence alignments with a reference genome of Brassica oleracea in the Ensembl plant genome database and the Bolbase database (Yu et al. 2013) using the Blastn function with default parameters. Information regarding the gene functional annotations and protein features were acquired from different public databases as described by (Liu et al. 2016a, b).

Results

Genetic analysis of the BT and FT traits

Table 1: Results summary of the SLAF sequencing

Parameter	Value
Enzyme digestion protocol	HaeIII + Hpy166II
Digestion normally	91.37%
Digestion partly	8.63%
Fragment length	264-314bp
Clean reads	423.21M
Average Q30	94.78%
Average GC content	40.53%
Develop the total number of SLAF	191,209
Paired-end mapped reads	74.01%
Singled-end mapped reads	5.46 %
Polymorphic SLAF	27,875
Number of linkage groups	9
Number of markers above	4,296
Genetic map total map distance	1,075.66cM
Genetic map average distance	0.25cM

The F₂ population of B. oleracea was created by crossing ‘Dasun’ and ‘LB07M’ lines. This population was used to investigate the inheritance of FT and BT. The values for FT and BT ranged 42 to 80 and 37 to 61 days respectively (Fig. 1). The phenotypic distribution patterns of both traits in the F₂ population (Fig. 1) showed that both the traits are inherited quantitatively.

Analysis of SLAF sequencing data and genotyping

DNA sequencing provided nearly 4.8 Gb raw data containing 423.21 M clean reads with fragments length ranging from 264 to 314 bp (Table 1). The Q30 score was 94.78%, and the guanine-cytosine proportion was 36.10%. The details of SLAF sequencing are provided in Table 1. A total of 11.5 M and 15.4 M high quality reads were obtained (Table 2). The aa × bb segregation patterns with 13,329 SLAFs were used for further analysis (Fig. 2).

High-density linkage map construction

A total of 4,296 markers were mapped onto nine chromosomes, designated Chr 1-Chr 9 (Table 3). The average mapped markers integrity was 99.77%, as indicative of the high quality of the genetic map. Ultimately, a genetic linkage map of 1,075.66 cM was obtained (Fig. 3). On an average basis, chromosomes contained 477.3 spanning at an average distance of 119.52 cM. The map lengths of the 9 chromosomes ranged from 83.86 cM (Chr 2) to 164.24 cM (Chr 6). Chromosome 9 and 2 contained maximum (592) and minimum (295) numbers of markers, respectively (Table 3).

QTL mapping of FT and BT traits

Phenotypic data of FT and BT traits is provided in Fig. 1. Three QTL were detected for both traits (Table 3). The primary QTL for BT (LOD=17.77) was identified as the map position of the chromosome 2 (Table 3 and Fig. 5). This accounted for 41.14% of the total phenotypic variance for the trait. Whereas, one minor QTL controlling the same trait was located at the chromosome 7 (LOD=6.56) and accounted for 15.00% of the total phenotypic variance for the BT trait (Table 4).

The primary QTL for the FT was also located at the chromosome 2, and accounted for 45.08% of the total phenotypic variance for the FT trait (Table 4; Fig. 4, 5). It is worth mentioning that both major QTLs controlling the FT and BT traits were located on chromosome 2 and have a total spanned distance of 83.86 cM. This indicated that pleiotropic or neighbor gene(s) are governing the FT and/or BT traits.

Association of SNP markers and functional genes

Based on the evidences from previously published literature (Table 5) three candidate genes (Bo2g089890, Bo2g009900 and Bo2g048220) were selected mainly governing BT and FT traits. The gene Bo2g089890 (SAM1) is homologous of Arabidopsis gene AT3G15170 which encodes the cup-shaped cotyledon protein-2 that regulates shoot apical meristem (Arús and Orton 1983; Takada et al. 2001; Vroemen et al. 2003). The second selected gene Bo2g009900 (WNK8) is homologous to Arabidopsis gene AT5G41990 which regulates the floral meristem and phyllotactic patterning (Chandler 2014; Zhang et al. 2016).

Discussion

In this study, SLAF sequencing was used to construct a high-density genetic map of B. oleracea using an F₂ segregating population of 159 individuals obtained by crossing plants of Chinese kale lines ‘Dasun’ and ‘LB07M’. Parents of the mapping population had different bolting time and flowering time profiles. Particularly, ‘Dasun’ has longer bolting and flowering times than ‘LB07M’. Our map spanned 1,075.66 cM in 9 linkage groups, designated Chr1–Chr9. There were a few maps available for Chinese kale despite of many linkage maps have been reported for B. oleracea. Compared with previously reported genetic map of B. oleracea (Babula et al. 2003; Huang et al. 2017), this newly constructed map had fewer gaps and high linearity between genetic and physical distance.

Flowering time correlates with production and quality of Chinese kale. In a previous study, loci controlling flowering time were mapped in F₂ populations using SSR and SRAP markers (Li et al. 2015). The two QTLs were found on LG5 at a distance of 1.0 cM, governing bolting and flowering times in Chinese kale (Li et al. 2015). In this study both major QTL of BT and FT (1.1 and 2.1) were located at chromosome 2 at the distance less than 1.0 cM, showing their presence at an identical locus (Table 4 and Fig. 5). The correlation coefficient between BT and FT time in F2 was 0.872, showing a very significant positive correlation. Some previous studies e.g., Chaim et al. (2001); Fazio et al. (2003) and Rao et al. (2003) have also reported that QTLs for closely related traits are likely to be located on identical or same regions of chromosomes.

Fig. 1: The phenotypic distribution of the bolting (A) and flowering time (B) traits in the F₂ population of B. oleoracea. P₁ shows the average bolting or flowering time of parent one (LB07M). P₂ shows the average bolting or flowering time of parent two Dasun. F₁ shows the average bolting or flowering time of F₁ generation

Fig. 2: Numbers of each marker segregation type on the linkage maps of Chinese kale. X-axis represents segregation patterns whereas y-axis shows number of SLAFs

Fig. 4: The collinearity of chromosomes with the B. oleoracea reference genome. The x-axis indicates the genetic distance of B. oleoracea chromosomes accordingly, and the y-axis represents the linearity order of the physical position in the soybean genome

Fig. 5: Distribution of QTLs and selected candidate genes of BT and FT traits on Chromosome-2 of Chinese Kale. (A) Candidate genomic region putatively controlling BT trait on chromosome-2. (B) Candidate genomic region putatively controlling FT trait on chromosome-2

Fig. 6: Possible involvement of selected candidate genes in floral pathway of Chinese kale plant. Selected genes ID are shown in red color

Fig. 3: Distribution of SLAF markers on nine chromosomes of Chinese kale. A black bar indicates a SLAF marker. The x-axis represents chromosome number and the y-axis indicates genetic distance in centi-morgan (cM)

Three candidate genes were selected based on the reference genome, previously published literature (Table 5) and their homologous genes related to same traits in other plants. The putative role of these candidate genes in floral pathway is shown in Fig. 6. Two candidate genes, Bo2g089890 and Bo2g009900 that were selected on the Table 2: SLAF-seq data summary for Chinese kale F₂ population

Samples	Clean reads	Q30 %	GC %	SLAF number	Total depth	Average depth
Dasun	11,506,750	95.05	40.12	146,077	5,487,967	37.57
LB07M	15,469,125	94.63	40.50	155,551	7,134,346	45.86
Offspring	2,492,061	94.77	40.53	122,703	1,174,674	9.57

Table 3: Basic characteristics of Chinese kale linkage groups

Chromosome	Marker no.	Map length	Maximum distance	Marker interval	Integrity (%)	Double crossover (%)	Missing (%)
Chr 1	286	84.71	3.27	0.30	99.82	0.01	0.18
Chr 2	295	83.86	3.59	0.29	99.80	0.00	0.20
Chr 3	760	154.81	3.34	0.20	99.75	0.01	0.25
Chr 4	390	87.12	2.58	0.22	100.0	0.01	0.00
Chr 5	460	130.89	3.91	0.29	99.79	0.02	0.21
Chr 6	368	164.24	2.64	0.45	99.51	0.07	0.49
Chr 7	527	101.59	3.21	0.19	99.73	0.01	0.27
Chr 8	618	129.33	3.65	0.21	99.79	0.02	0.21
Chr 9	592	139.11	4.22	0.24	99.74	0.02	0.26
Total	4296	1075.66	4.22	0.25	99.77

Table 4: An overview of QTLs related with FT and BT traits

QTL ID	Chromosome	Start (cM)	End (cM)	max LOD	PVE
1.1	2	6.983	7.298	17.77	41.14
1.2	7	48.81	49.12	6.56	15.00
2.1	2	6.04	6.35	18.80	48.08

PVE= Phenotypic contribution rate

Table 5: Selected candidate genes associated with BT and FT traits in B. oleracea found on major QTLs

Gene name	Gene ID	Function	Reference
SAM1	Bo2g089890	Regulation of shoot apical meristem.	(Takada et al. 2001; Vroemen et al. 2003)
WNK8	Bo2g009900	Regulating floral meristem, phyllotactic patterning	(Chandler 2014; Zhang et al. 2016)
GA20OX3	Bo2g048220	Initiation of flowering.	(Mutasa-Göttgens and Hedden 2009; Plackett et al. 2011; Rebers et al. 1999; Tenreira et al. 2017)

major QTL 1.2, which controls the traits for bolting time or first flowering bud. These genes encode the cup-shaped cotyledon protein-2 (Bo2g089890) and serine/threonine-protein kinase (Bo2g009900), which have been shown to regulate shoot apical and flower meristems, respectively, according to the GO annotations and previously published literature (Table 5). The gene Bo2g009900, is homologous to Arabidopsis gene AT5G41990, and encodes a WNK8 like serine/threonine-protein kinase. The same gene in Arabidopsis plants has been shown to regulate the flowering time by modulating the photoperiod pathway (Wang et al. 2010). The WNK8 protein also interacts with the EDM2 protein that in turn, modulates floral meristem and developmental processes (Tsuchiya and Eulgem 2010). The flower organ specification processes are centered within the flower meristem (FM) to generate floral organs (Chandler 2014), with the FM differentiation determined by the stem cells within the shoot apical meristem (Zhang et al. 2016). This coordinated role of SAM and FM highlights the importance of both of these meristems in bolting time or appearance of the first flower node. When considering the importance of the shoot apical and flower meristems in the development of first flower node (Zhang et al. 2018), the genes Bo2g024555 and Bo2g009900 are important candidates for bolting time or appearance of the first flower node trait in B. oleoracea. Chromosome 02 appears to be a strong candidate having major QTLs with the genes controlling bolting time.

The candidate gene for flowering time, Bo2g048220, is homologous to Arabidopsis AT5G07200, and encodes the gibberellic acid (GA) 3-oxidase (GA3ox) protein. The GA oxidases proteins primarily dictate early floral initiation (Rebers et al. 1999; Mutasa-Göttgens and Hedden 2009; Tenreira et al. 2017) and fruit development (Rebers et al. 1999) by regulating different floral networks (Mutasa-Göttgens and Hedden 2009). The GA3ox catalyzes the last step in bioactive GA biosynthesis. GA is a regulator of flowering initiation (Plackett et al. 2011). The exogenous application of GA₃ and GA₄ has shown early bud development in apple plants (Bertelsen et al. 2002).

As both major QTLs are residing on chromosome 02, this genomic region can be helpful for developing early maturing Chinese kale varieties using marker assisted breeding technology.

Conclusion

This is the first report highlighting the involvement GA3ox, WNK8 and SAM1 genes in the quantitative inheritance of early BT and FT traits in Chinese kale plants. Secondly, the high-density genetic map of Chinese kale constructed in this study will offer a suitable basis for further study of Chinese kale, such as gene mapping, map-based cloning of specific genes, quantitative trait locus mapping and marker-assisted selection.

Acknowledgements

This study was supported by the projects of Guangdong Agriculture Department Foundation (2019KJ122) and Guangdong Academy of Agricultural Sciences Foundation of the Dean Project (Grant No. 201816B).

References

Arús P, TJ Orton (1983). Inheritance and linkage relationships of isozyme loci in Brassica oleracea. J Hered 74:405‒412

Babula D, M Kaczmarek, A Barakat, M Delseny, CF Quiros, J Sadowski (2003). Chromosomal mapping of Brassica oleracea based on ESTs from Arabidopsis thaliana: Complexity of the comparative map. Mol Genet Genomics 268:656‒665

Bertelsen MG, DS Tustin, RP Waagepetersen (2002). Effects of GA₃ and GA₄₊₇ on early bud development of apple. J Hortic Sci Biotechnol 77:83‒90

Broman KW, W Hao, S Saunak, GA Churchill (2003). R/qtl: QTL mapping in experimental crosses. Bioinformatics 19:889‒890

Chaim AB, I Paran, RC Grube, M Jahn, RV Wijk, J Peleman (2001). QTL mapping of fruit-related traits in pepper (Capsicum annuum). Theor Appl Genet 102:1016‒1028

Chandler JW (2014). Patterns and polarity in floral meristem and floral organ initiation. Crit Rev Plant Sci 33:457‒469

Fazio G, JE Staub, MR Stevens (2003). Genetic mapping and QTL analysis of horticultural traits in cucumber (Cucumis sativus L.) using recombinant inbred lines. Theor Appl Genet 107:864‒874

Honghao L, W Qingbiao, L Xing, H Fengqing, F Zhiyuan, Y Limei, Z Mu, L Yumei, L Zhansheng, Z Yangyong (2016). Whole-genome mapping reveals novel QTL clusters associated with main agronomic traits of cabbage (Brassica oleracea L. var capitata). Front Plant Sci 7; Article 789

Huang L, Y Yang, F Zhang, J Cao (2017). A genome-wide SNP-based genetic map and QTL mapping for agronomic traits in Chinese cabbage. Sci Rep 7; Article 46305

Koornneef M, C Alonsoblanco, HB Vries, CJ Hanhart, AJM Peeters (1998). Genetic interactions among late-flowering mutants of Arabidopsis. Genetics 148:885‒892

Levy YY, C Dean (1998). The transition to flowering. Plant Cell 10:1973‒1990

Li G, G Zhang, Z Yan, L Kai, T Li, H Chen (2015). Identification of quantitative trait loci for bolting and flowering times in Chinese kale (Brassica oleracea var. alboglabra) based on SSR and SRAP markers. J Pomol Hortic Sci 90:728‒737

Li GH, HC Chen, JL Liu, WL Luo, DS Xie, SB Luo, TQ Wu, W Akram, YJ Zhong (2019). A high-density genetic map developed by specific-locus amplified fragment (SLAF) sequencing and identification of a locus controlling anthocyanin pigmentation in stalk of Zicaitai (Brassica rapa L. ssp. chinensis var. purpurea). BMC Genomics 20; Article 343

Liu C, B Fan, Z Cao, Q Su, Y Wang, Z Zhang, J Tian (2016a). Development of a high-density genetic linkage map and identification of flowering time QTLs in adzuki bean (Vigna angularis). Sci Rep 6; Article 39523

Liu T, L Guo, Y Pan, Z Qi, J Wang, Z Song (2016b). Construction of the first high-density genetic linkage map of Salvia miltiorrhizausing specific length amplified fragment (SLAF) sequencing. Sci Rep 6; Article 24070

Mutasa-Göttgens E, P Hedden (2009). Gibberellin as a factor in floral regulatory networks. J Exp Bot 60:1979‒1989

Oded C, B Yelena, DS Rakefet, P Ilan (2014). Capsicum annuum S (CaS) promotes reproductive transition and is required for flower formation in pepper (Capsicum annuum). New Phytol 202:1014‒1023

Plackett ARG, SG Thomas, ZA Wilson, H Peter (2011). Gibberellin control of stamen development: A fertile field. Trends Plant Sci 16:568‒578

Rao GU, AB Chaim, Y Borovsky, I Paran (2003). Mapping of yield-related QTLs in pepper in an interspecific cross of Capsicum annuum and C. frutescens. Theor Appl Genet 106:1457‒1466

Rebers M, T Kaneta, H Kawaide, S Yamaguchi, YY Yang, R Imai, H Sekimoto, Y Kamiya (1999). Regulation of gibberellin biosynthesis genes during flower and early fruit development of tomato. Plant J 17:241‒250

Rahman H, RA Bennett, B Kebede (2018). Molecular mapping of QTL alleles of Brassica oleracea affecting days to flowering and photosensitivity in spring Brassica napus. PLoS One 13; Article e0189723

Sun X, D Liu, X Zhang, W Li, H Liu, W Hong, C Jiang, N Guan, C Ma, H Zeng, C Xu, J Song, L Huang, C Wang, J Shi, R Wang, X Zheng, C Lu, X Wang, H Zheng (2013a). SLAF-seq: An efficient method of large-scale De Novo SNP discovery and genotyping using high-throughput sequencing. PLoS One 8; Artile e58700

Sun X, D Liu, X Zhang, W Li, H Liu, W Hong, C Jiang, N Guan, C Ma, H Zeng, C Xu, J Song, L Huang, C Wang, J Shi, R Wang, X Zheng, C Lu, X Wang, H Zheng (2013b). SLAF-seq: An efficient method of large-scale de novo SNP discovery and genotyping using high-throughput sequencing. PLoS One 8; Article e58700

Takada S, K Hibara, T Ishida, M Tasaka (2001). The CUP-SHAPED COTYLEDON1 gene of Arabidopsis regulates shoot apical meristem formation. Development 128:1127‒1134

Tenreira T, M JP Lange, T Lange, C Bres, M Labadie, A Monfort, M Hernould, C Rothan, B Denoyes (2017). A specific Gibberellin 20-Oxidase dictates the flowering-runnering decision in diploid strawberry. Plant Cell 29:2168‒2182

Tsuchiya T, T Eulgem (2010). Co-option of EDM2 to distinct regulatory modules in Arabidopsis thaliana development. BMC Plant Biol 10; Article 203

Vroemen CW, AP Mordhorst, A Cathy, MACJ Kwaaitaal, SC Vries (2003). The CUP-SHAPED COTYLEDON3 gene is required for boundary and shoot meristem formation in Arabidopsis. Plant Cell 15:1563‒1577

Wang Y, K Liu, H Liao, C Zhuang, H Ma, X Yan (2010). The plant WNK gene family and regulation of flowering time in Arabidopsis. Plant Biol 10:548‒562

Wu J, Q Zhao, L Zhang, S Li, Y Ma, L Pan, H Lin, G Wu, H Yuan, Y Yu (2018). QTL mapping of fiber-related traits based on a high-density genetic map in flax (Linum usitatissimum L.). Front Plant Sci 9; Article 885

Yu J, M Zhao, X Wang, C Tong, S Huang, S Tehrim, Y Liu, W Hua, S Liu (2013). Bolbase: A comprehensive genomics database for Brassica oleracea. BMC Genomics 14; Article 664

Zhang X, G Wang, B Chen, H Du, F Zhang, H Zhang, Q Wang, S Geng (2018). Candidate genes for first flower node identified in pepper using combined SLAF-seq and BSA. PLoS One 13; Article e0194071

Zhang Y, D Zhang, H Yu, B Lin, Y Fu, S Hua (2016). Floral initiation in response to planting date reveals the key role of floral meristem differentiation prior to budding in canola (Brassica napus L.). Front Plant Sci 7; Article 1369

Zhang J, Q Zhang, T Cheng, W Yang, H Pan, J Zhong, L Huang, E Liu (2015). High-density genetic map construction and identification of a locus controlling weeping trait in an ornamental woody plant (Prunus mume Sieb. et Zucc). DNA Res 22:183‒189

Zhiwen W, H Neil, G Leonardo, Z Shilin, S Daihu, MD Joshua, Y Linfeng, H Simon, N Godfrey, D Raju (2012). The genome of flax (Linum usitatissimum) assembled de novo from short shotgun sequence reads. Plant J 72:461‒473