HE Dao-fang , ZHAO Xiang , LlANG Cheng-zhen, ZHU Tao, Muhammad Ali Abid, CAl Yongping HE Jin-ling ZHANG Rui

1 School of Life Science, Anhui Agricultural University, Hefei 230036, P.R.China

2 Biotechnology Research Institute, Chinese Academy of Agricultural Sciences, Beijing 100081, P.R.China

Abstract Leaf is a essential part of the plants for photosynthetic activities which mainly economize the resources for boll heath.Significant variations of leaf shapes across the Gossypium sp. considerably influence the infiltration of sunlight for photosynthesis. To understand the genetic variants and molecular processes underlying for cotton leaf shape, we used F2 population derived from upland cotton genotype P30A (shallow-lobed leaf) and sea-island cotton genotype ISR (deeplobed leaf) to map leaf deep lobed phenotype controlling genes LBL1 and LBL2. Genetic analysis and localization results have unmasked the position and interaction between both loci of LBL1 and LBL2, and revealed the co-dominance impact of the genes in regulating depth of leaf blades lobes in cotton. LBL1 had been described as a main gene and member of transcription factor family leucine zipper (HD-ZIPI) from a class I homologous domain factor Gorai.002G244000. The qRT-PCR results elaborated the continuous change in expression level of LBL1 at different growth stages and leaf parts of cotton. Higher expression level was observed in mature large leaves followed by medium and young leaves respectively.For further confirmation, plants were tested from hormonal induction treatments, which explained that LBL1 expression was influenced by hormonal signaling. Moreover, the highest expression level was detected in brassinolides (BR) treatment as compared to other hormones, and this hormone plays an important role in the process of leaf blade lobed formation.

Keywords: leaf blades lobes, HD-ZIPI, LBL1, cotton, Gossypium barbadense, Gossypium hirsutum

1. lntroduction

Cotton (Gossypium sp.) is an important economic crop of the world and the fiber of cotton mainly used as a raw material for producing textiles. In Gossypium genus, the leaf lobe is a highly variable trait, cultivated cotton comprise of broad leaf (Chang et al. 2016). The leaf shape diversification of cotton is described by the degree of leaf margin into the following categories: ovate, cordate, normal, sub okra,laciaiate, okra, super okra; and leaf shape diversification is also beneficial for textile production because it improves the photosynthetic efficiency of the whole plant (Kidner 2010b;Nicotra et al. 2011). Cotton genotypes with deeper leaf foliage showed better environmental adaptability, reduced lint trash, better plant’s canopy structure, ventilation, and light transmittance. The deep leaf lobe phenotype of cotton is also associated with resistance to pests and diseases (Andres et al. 2013), drought resistance (Baker and Myhre 1969;Karami et al. 1980; Siso et al. 2001), early maturity (Karami and Weaver et al. 1980), fiber quality and yield (Andries et al.1970; William and Randy 1986; Wu and Sun 1987; Zhu et al.2009; Liu et al. 2015). The okra leaf phenotype is associated with better water use efficiency, CO2exchange rate, and a lower stomatal conductance as compared to broad leaf phenotype (Pettigrew et al. 1993). Despite the importance of leaf shape, the molecular and genetic control of leaf shape in cotton is not well characterized. Many studies confirmed that the okara shape leaf phenotype in upland cotton was controlled by a pair of imperfectly dominant genes (Peebles and Kearney 1928; Nawab et al. 2011). QTL analysis of upland cotton showed that the major gene was mapped on the short arm of chromosome 15 (Jiang et al. 2000; Paterson et al. 2012; Lacape et al. 2013). Ten candidate genes were also identified in the 112 kb region at 5.4 cM to the telomere of chr15 (Andres et al. 2014). Corresponding region in G. raimondii located on chr02 has a gene Gorai.002g244000,which is a homologous to the candidate gene in upland cotton,which is considered to control the trait of the deep leaf blades lobe (OKRA) (Zhu et al. 2015). Transcriptome analysis of this candidate gene showed that most of the pathways were related to photosynthesis. Most of differentially expressed photosynthesis-related genes were upregulated in okra leaf,this indicate that it may be play an important regulatory role in plant photosynthesis (Andres et al. 2016).

In this study, we used the F2population, which was obtained by crossing upland cotton cultivar P30A with normal leaf morphology and sea-island cotton ISR with deeplobed leaves for mapping of the LBL genes (Gossypium barbadense L., deep leaf lobes). The sea-island cotton ISR leaves are palmate, deeply lobed, 7-12 cm in diameter,lobes ovate or oblong, apex long acuminate, middle lobe longer, lateral lobes usually spreading, and base cordate.The upland cotton P30A leaves are broadly ovate, 5-12 cm in diameter, length, and width nearly equal or wide, base cordate or heart-shaped to frustum shaped, usually three lobed, rarely five lobed, lobes often parted the leaves in half,lobes broadly triangular ovate, apex abruptly acuminate and with a wide base. In this study, we have mapped and cloned the main gene LBL1 for deep leaf lobe shape in cotton which provide comprehensive understanding of the cotton leaf development and open new avenues for developing cotton cultivars with ideal leaf shape for better yield and quality production of cotton.

2. Materials and methods

2.1. Plant materials and leaf shape measurement

The F2population from the F1self pollinated progeny of a cross of P30A (Gossypium hirsutum) genotype with ISR(G. barbadense) genotype, was grown to map the leaf shape traits in the field of the Chinese Academy of Agricultural Sciences, Beijing, China in 2016 (Appendix A). Leaf shapes of individual plants in F2generation were observed. At the same time, the first lateral veins, the second lateral veins,the first lobes, and the second lobes were measured and analyzed. Part A represents the main vein, part B represents the first lateral vein, part C represents the first lateral vein,part D represents the second lobe, and part E represents the second lateral vein (Fig. 1).

For genetic analysis, two groups of 300 and 500 individual plants were analyzed for main veins of inverted the fourth true leaves, leaf blade on right side first, the second lateral veins, the first leaf lobe, and the second leaf lobe. By using these measurements calculated the following ratios: (A)The first leaf lobe to main vein; (B) the second leaf lobe to main vein; (C) the first leaf lobe to first lateral vein; (D)the second leaf lobe to second lateral vein. The ratio of A had no cross and could be separated from the character area of ISR (deep leaf lobe) and P30A (shallow leaf lobe)(Appendix B). There were some overlapping between ISR and P30A in ratio of B (ISR: 0.26-0.45, P30A: 0.35-0.54), C(ISR: 0.40-0.79, P30A: 0.63-0.97), and D (ISR: 0.30-0.50,P30A: 0.39-0.61), which could not distinguish the traits of the two parents. As these three overlapping traits cannot be used as standards of differentiation, therefore these traits were not used for further analysis (Fig. 2). Hence the ratio of A was selected as the standard character, the value of ISR leaves was found to be lower than 0.46 as a dominant trait, while P30A larger than 4.7 (Table 1). Based on above results, the phenotypic data of F2population were divided into two groups using 0.46 and 0.47 as a separation, the ratio lower than 0.46 were dominant trait, while larger than 0.47 were recessive trait.

Fig. 1 Different measuring parts of leaf. A, main vein; B, first lateral vein; C, second lateral vein; D, first leaf lobe; E, second leaf lobe.

Fig. 2 Comparison and analysis of the four groups of data, the depth of leaf blades lobes in ISR is significantly higher than that of P30A. A, first leaf lobe/main vein, the ratio of ISR data range of 0.30-0.46, the ratio of P30A to 0.46-0.61 interval data, no overlap between the two sets of data. B, second leaf lobe/main vein, the ratio of data interval ISR is 0.26-0.45, the ratio of P30A is 0.35-0.54, so there is overlap between the two groups of data. C, first leaf lobe/first lateral vein, the ratio of ISR is 0.40-0.79,the ratio of P30A is 0.63-0.97, and there is overlap between values. D, second leaf lobe/second lateral vein, the ratio of ISR was 0.30-0.50, the ratio of P30A was 0.39-0.61, and there was overlap between the two groups.

2.2. Statistics and analysis

Statistical data were analyzed by using Excel 2003 and SPASS 16.0 Software. Significant differences were analyzed by using Tukey’s and Duncan’s methods, and the “D/A, E/A, D/B, E/C” were statistically analyzed (Fig. 1).The criteria that best distinguished the trait difference between the two parents were screened as the phenotypes of F1and F2populations.

2.3. Primary mapping of the LBL1

For fine mapping, 1 000 recessive single plants were selected from 34 860 F2individuals. Approximately 720 pairs of insertion/deletion (InDel) markers were used to screen for polymorphisms between upland cotton and sea-island cotton, and partial segregation primers were screened to determine the location of the LBL1 gene on the chromosome. PCR reactions of 20 mL contained 10 mL mix (2×Taq PCR Master Mix, Biomed), 7.5 mL ddH2O,1 mL of each primer, and 1 mL DNA template. The PCR reaction program was as follows: an initial cycle of 94°C for 3 min, then 40 cycles of 94°C for 30 s, annealing temperature 56°C for 30 s, 72°C for 1 min, and a final extension at 72°C for 5 min. DNA bands were observed on 4% agarose gel.

2.4. Fine mapping of the LBL1

Genome-wide reprogramming (WGS) data from island cotton released by Nanjing Agricultural University of China was downloaded from NCBI’s SRA database (https://www.ncbi.nlm.nih.gov/sra) and the downloaded WGS data were filtered by using NGS QC Toolkit: A Toolkit for Quality Control of Next Generation Sequencing Data. Burrows-Wheeler Aligner (BWA, http://bio-bwa.sourceforge.net/)tool was used to map the reads of island cotton to the genome of upland cotton (GenBank published by Nanjing Agricultural University). InDel and sequence alignment/map format (SAM) tools software were used to find the primers designed on both sides of each polymorphic site for InDel sites with ＜80 bp difference, and TNT BLAST used to detect primer-to-genomic (NAU) matches. By using the designed polymorphic primers, 757 single plant lets selected from 1 632 recessive individuals in the population and were amplified by PCR and further fine-mapping. The first true leaves and the third, fourth, fifth, and sixth leaves of the cotton plants with the complete leaf-splitting phenotype were used to analyze the expression level of the candidate genes between the parents.

This time Mom got even with me. She put it under the base of our living?room lamp, knowing that as a mother of three little ones, housecleaning and moving lamps would not be everyday events. . . ~ .

2.5. Expression of LBL1 gene in different tissues

RNA was extracted from both ISR and P30A roots, stemsand leaves at different growth stages, such as, mature leaf(inverted fifth leaf), medium sized leaf (inverted fourth leaf)and young leaf (inverted third leaf). About 21 parts of single leaf, including: the leaf petiole (Fig. 3 A, 1; a,), the leaf margin between leaf petiole and the first lateral vein (Fig. 3 B, 2; b,), the tip of first lateral vein (Fig. 3 C, 3; c,), the first leaf lobe (Fig. 3 D, 4; d,) , the leaf margin between the first lobes and the second lateral vein (Fig. 3 E, 5; e,), the tip at the second lateral vein (Fig. 3 F, 6; f,), the second leaf lobe (Fig. 3 G, 7; g,), the leaf margin between the second leaf lobe and the main vein (Fig. 3 H, 8; h,), the main vein at the tip (Fig. 3 I, 9), the parts between first leaf lobe and leaf petiole (Fig. 3 J, 10; j,), the parts between second leaf lobe and leaf petiole (Fig. 3 K, 11; k,), and the main veins on both sides (Fig. 3 L, 12; l,). At the same times, the sampling positions on the right and left sides correspond to each other. RNA extracted was analyzed by real-time fluorescence quantitative PCR, respectively(Fig. 3). The EASY Spin Plant Ultra-Pure RNA Rapid Extraction Kit (Beijing Yuanping Hao Biological Technology Co., Ltd., China) was used to extract RNA, and TransScript One Step Removal and cDNA Super Mix Synthesis Reverse Transcription Kit (Beijing TransGen Biotech Co., Ltd., China)were used to obtain the expression of cDNA, according to the manufacturer’s instructions. The primers used in qRTPCR are shown in Appendix A.

2.6. Phytohormone induction

Different hormonal conditions: abscisic acid (ABA, 50 μm),jasmonic acid (JA, 50 μm), gibberellin A3 (GA3, 50 μm),6-benzylaminopurine (6BA, 50 μm), indole-3-acetic acid (IAA,50 μm), ethylene (50 μm), and salicylic acid (SA, 100 μm)were used for 3-wk-old cotton seedlings and treated for 4 h to observe the change in expression level of candidate gene.

3. Results

3.1. The trait of deep leaf blades lobes in lSR is dominant to shallow leaf lobes in P30A

The ratio of A had no overlapping values that can be distinguished this character between ISR (parted) and P30A (lobed). If these three sets of data were used as a standards of differentiation, the different leaf-splitting traits could not be divided into rational stratification, therefore the later three sets of data were not used (Fig. 2). The ratio of A was selected as the standard division characters, the value of ISR was found to be more than 0.30 and lower than 0.46 as a dominant trait (Table 1). Based on this result, the data can be divided to 0.46 and 0.47 as a separation criteria,lower than 0.46 were parted and dominant character, large than 0.47 were lobed and recessive trait.

Table 1 The first leaf lobe/main vein ratio as the best leaf classification standard

Fig. 3 Different parts of one leaf of various sampling sites corresponding to mature leaves, medium sized leaves and young leaves. Including leaf petiole (A, 1; a,), the leaf margin between leaf petiole and first lateral vein (B, 2; b,), the tip of first lateral vein (C, 3; c,), the first leaf lobe (D, 4; d,), the leaf margin between first lobe and the second lateral vein (E,5; e,), the tip at second lateral vein (F, 6; f,), the second leaf lobe (G, 7; g,), the leaf margin between second leaf lobe and main vein (H, 8; h,), main vein at the tip (I, 9), the parts between the first leaf lobe and leaf petiole (J, 10; j,), the parts between the second leaf lobe and leaf petiole (K, 11; k,), the main veins on both sides (L, 12; l,). The sampling positions on the right and left side correspond to each other.

3.2. Two interacting genes control the trait of leaf blades lobe through genetic analysis

In this experiment, the F2population was obtained from the F1self pollinated progeny of a cross of P30A (G. hirsutum)genotype with ISR (G. barbadense) genotype. The F2generations were contained 5 498 individuals of which 3 083 individuals had leaves with the ISR sea-island phenotype and 2 415 individuals had leaves with the P30A normal phenotype. The result was consistent by Chi-square test of leaf separation accord with Mendel genetic separation ratio (9:7, χ2=0.07) (Table 2), indicated that there were two interacting genes controlling the leaf lobe shape in cotton,and the sea-island genes were dominant.

3.3. The candidate gene selected in the location interval is homologous to Gorai.002G244000

According to the differences between the sequences of upland cotton and sea-island cotton, 720 polymorphic InDel markers were designed for gene mapping, and leaf lobe gene between two markers D01DEL25.11 and D01DEL61.45 on chrD01 (Appendices A-D) and the other gene between two markers D09DEL35.63 and D09DEL37.84 on chrD09 were discovered (Appendix C).All appeared to have the segregation phenomenon, which is consistent with previous genetic analyses showing that leaf traits were controlled by two sets of alleles. In order to reduce the positioning interval, above two markers were used to screen 3 648 individuals with shallow lobes from F2population. Based on the polymorphic markers between the two parents, a series of markers (D01DEL46.16,D01DEL58.27, D01DEL42.16, D01INS49.99, D01INS58.54,and D01INS60.02) were developed between the two markers D01DEL25.11 and D01DEL61.45 for fine positioning. Finally, 154 kb DNA fragment was mapped between D01INS59.343 and D01DEL59.497 markers containing nine candidate genes (Appendix D, Fig. 4), which are highly conserve in the two cotton subspecies genomes,G. barbadense and G. hirsutum.

Expression analysis of these nine genes in the two parental genotypes showed that Gorai.002G244000 (putative homeobox-leucine zipper protein ATHB-51) was not found to have an expression in the first true leaf. The expression of Gorai.002G244000 in the leaf with deep lobed phenotype in the bud period showed a trend of increasing from the tip to the middle, in the detection of the inverted third to sixth leaves, the inverted third leaves had the lowest expression level, then the inverted fourth and fifth leaves, the inverted sixth leaves had the highest expression level (Fig. 5-A).This result illustrated that the changes of the expression level of Gorai.002G244000 was along with the phenotypic characteristics of cotton leaf morphological development and the leaflet depth during different growth periods (Fig. 5-B),while the remaining other eight genes did not show the similar trend. Therefore, Gorai.002G244000 was determined as a key regulatory gene for deep-leaf blades in the sea-island cotton cultivar. Gorai.002G244000 code a transcription factor with the leucine zipper (HD-ZIPI) class I domain (Table 3).It’s homologous to LATE MERISTEM IDENTITY1 (LMI1) in Arabidopsis which was verified to be involved in the regulation of Arabidopsis thaliana margin traits.

3.4. The mutant gene lbl1 has two deletions and three SNPs

The LBL1/lbl1 genome sequence on Chr_D01 has cloned from the ISR and P30A, respectively. ISR/P30A multiplesequence alignment revealed that there were three SNPs and two deletions. At positions of 185, 268, and 882 bp,lbl1 gene in P30A has “A”, “A”, and “T”, while LBL1 gene in ISR has “G”, “C”, and “A”. At positions of 742 and 861 bp,lbl1 gene in P30A has a 8- and 1-bp deletion, respectively while compare to LBL1 gene in ISR. The 8-bp deletion oflbl1 gene in P30A is likely to lead to premature termination,may result in the low expression level and confer shallow lobe leaves appearance (Appendix E).

Table 2 The individual number of two leaf shapes and Chi-square test in (ISR×P30A) populations

Fig. 4 Fine mapping of LBL1, a major effect gene of the deep-lobe trait in the sea-island cultivar, arrows indicate the direction of transcription of the corresponding gene, candidate genes in the mapping region are highly conserved in the cotton genome. The red arrow represents the gene of interest, and the positions of the blue arrows indicate the relevant gene of Gh-Dt01. The position of the gray arrow indicates that the relevant gene on Gh_Dt01 corresponds to the position on Gorai.002. NAU, Nanjing Agricultural University, China; JGI, JOINT GENOME INSTITUTE.

Fig. 5 The difference between the expression levels of the candidate genes in accordance with the phenotypic changes is that of the gene Gorai.002G244000, which has been identified as the best candidate gene. A, nine genes in the region between D01DEL58.823-D01INS59.735 and relative expression. a, Gorai.002G243500; b, Gorai.002G243600; c, Gorai.002G243700; d,Gorai.002G243800; e, Gorai.002G243900; f, Gorai.002G243000; g, Gorai.002G244400; h, Gorai.002G244500; i, Gorai.002G244000.The leaves of the first leaf of the early stage of germination, the leaves of the cotton buds in the bud stage, the three leaves, the inverted four leaves, the inverted five leaves, and the internal reference was Histone 3. B, the first row of ISR, the second row of P30A. 1, the first leaf of germination; 2-4, bud stage cotton plant with complete leaf split table type of inverted three, four, and five leaves. Bars=1.0 cm.

Table 3 Genes in the region between D01DEL58.823-D01INS59.735, and information about candidate genes

3.5. The expression of LBL1 gene is tissue-specific and highly expressed in the deepest part of the leaf lobes

The expression levels in various parts of the mature leaf were investigated. The corresponding position on both sides of a leaf has consistent trends on relative expression level of LBL1 gene (Fig. 3). The result showed that the LBL1 gene expression level of the deep lobe site was the highest (Fig. 3 G, 7; g,), followed by the shallow lobe site (Fig. 3 D, 4;d,), then the portion between petiole and blade (Fig. 3 J, 10; j,/K, 11; k,). The apex (Fig. 3 C, 3; c,/F, 6;f,/I, 9) and margin (Fig. 3 B, 2; b,/E, 5; e,/H, 8; h,) sites were with the lowest level, and had no significant difference from each other (Fig. 6). The phenomenon that the expression level of LBL1 gene in leaf blades was higher than the other leaf parts, with an increasing trend along with the fissure deepened, inferred that the leaf lobes trait was controlled by LBL1 gene.

On the whole, the expression level of LBL1 gene also changed at different growth stages of cotton, and its expression level in mature large leaves was the highest,followed by medium leaves, and the lowest in the young leaves. However, at the seedling stage, the expression level of LBL1 gene is different. The first three to four true leaves have only shallow or no leaf lobe, and expression level of these true leaves is too low to detect. Then the leaf lobes gradually deepen until the seventh true leaf has the fully developed leaf lobes, and the expression level of LBL1 gene is increasing accordingly (Fig. 7).

3.6. The expression of LBL1 gene is highly induced after brassinolide (BR) treatment

Plant hormones play a very important regulatory role in cell differentiation, plant development, and metabolism. We used different phytohormones to spray on the 4-mon-old cotton plants. The results showed the expression level of LBL1 gene increased most after BR induction, followed by IAA, GA3, and 6-BA induction (Fig. 8). However, there were not significantly different after SA, ACC, and JA induction,and significant reduced after ABA induction (Fig. 8). BR is a plant endogenous hormone, and its effect on the growth is highly significant. However, the expression level of LBL1 was significantly decreased after ABA treatment,possibly due to ABA was a stress-related phytohormone, the development and differentiation of a plant usually stop while in a stress environment with a increased ABA content, which might result in the decrease for LBL1 expression (Fig. 8).

4. Discussion

4.1. Deep leaf blades lobes of the island cotton were regulated by two alleles of LBL1 and LBL2, and the foot leaf traits of upland cotton were regulated by a pair of alleles LBL1

Leaf shape differs considerably across plant evolution and in response to the environment. Critical understanding of leaf shape and its genetic architecture is important for harnessing its effects for better agronomic profitability and to change plant physiology (Nicotra et al. 2011; Chitwood et al. 2016).Using genetic and molecular tools, we have identified two key genes controlling deep leaf lobes differentiation which located on chr_D01 and chr_D09, respectively.

Fig. 6 LBL1 in different growth stages of cotton and the same leaf in different parts of the expression level. The abscissa indicates different sampling sites. A, LBL1 on the relative expression of mature leaf on the right side of the amount. B, LBL1 on the relative expression of mature leaf on the left side of the amount. C, LBL1 on the relative expression of right of medium size leaves. D,LBL1 on the relative expression of medium size on the left side of the blade. E, LBL1 gene in relative expression leaves on the right side of the amount. F, LBL1 in the young leaves on the left side of the relative expression quantity. Bars are SD.

At the first generated, an F2genetic population using the sea-island cotton cultivar ISR with deep-lobed leaves as the male parent, while the upland cotton cultivar P30A with shallow-lobed leaves as the female parent. The separation ratio of leaf shape in F2population was 9:7 in line with the Mendel genetic ratio, which showed the trait of deep-lobed leaves in sea-island cotton ISR was controlled by two interacting gene loci. However, this result was not consistent with Andres et al. (2016) who thought only one gene regulating the leaf lobes trait. This may be due to the different genetic populations used. We used a F2population of the upland cotton P30A crossed with the sea-island cotton ISR, while they used a interbreeding population of upland cotton. What’s more, we studied the formation of deep lobes in sea-island cotton leaf, while they concentrated on okra leaf shape in upland cotton.

The main gene controlling leaf lobes differentiation was located on chrD01 between two InDel markers D01DEL58.823 and D01INS59.735. The physical distance was 154 kb, including nine candidate genes. The result of the expression analysis for those nine genes showed Gorai.002G244000 was the candidate. It coded a leucine zipper (HD-ZIP I), class I homologous domain transcription factor and was homologous to Arabidopsis LATE MERISTEM IDENTITY1 (LMI1), which was consistent with previous studies (Andres et al. 2016; Chang et al. 2016).

Fig. 7 Phenotypic differences of 2-mon-old cotton leaves and showing a dynamic mode of expression. There was no significant difference in the first four leaves of Gossypium hirsutum or Gossypium barbadense. The leaves are different from the 5th leaf,and the stable leaves appeared in the 7th leaf. Bars=1.0 cm.

Fig. 8 In the induction of exogenous hormone, the response of LBL1 gene to brassinolide (BR) is the strongest. A, using Histone3 as the internal reference. B, using actin-7 as the internal reference. IAA, indole-3-acetic acid; GA3, gibberellin A3; 6-BA,6-benzylaminopurine; SA, salicylic acid; ACC, 1-aminocyclopropanecarboxylic acid; JA, jasmonic acid; ABA, abscisic acid. Bars are SD.

4.2. The Gohir.D01G199600 is a class of leucine zipper (HD-ZlP l) transcription factors that encode the l homologous domain

The homologous domain of leucine zipper transcription factor HD-ZIP can be divided into four families, HD-ZIP I-IV, all of which play an important role in the growth and development of Arabidopsis thaliana (Sessa et al. 1994). A number of homologous domain transcription factors HD-ZIP genes have been identified in the Arabidopsis genome, and most of them play a key role in leaf morphology. Therefore,it can be inferred that Gorai.002G244000 is a critical gene that controls the development of cotton leaf lobes. The other interaction gene was initially located on chrD09 between two InDel markers D09INDEL35.63 and D09INDEL37.84,with a physical distance of 2.21 Mb (Appendix C), and need to design more specific primers to narrow the distance for further fine mapping.

4.3. The expression of gene LBL1 is spatiotemporal in cotton leaf morphogenesis

As the large amount of energy needed in the vegetative growth stage of cotton, it is important for the whole leaf to receive the largest amount of light, which improve photosynthetic efficiency and increase the accumulation of photosynthetic products, and ultimately improve plant growth. With the growth and development of the plant,the number of leaves increased and the canopy structure became dense, so the phenomenon of overlapping leaves were observed. This involved the competition of the leaves to capture the sun, which was beneficial to the leaf-shaped spatial distribution of light absorption. The lower part of the plant greatly improve the efficiency of light utilization, leaf blade with good ventilation and light transmission, so form a more loose canopy structure, with the upper leaves on the lower leaves. Therefore, we concluded that the expression of LBL1 gene has the characters of both development stage specific and space specific, this is more conducive for a cotton plant to the use of light energy.

4.4. The expression of LBL1 can respond to hormonal signals, indicating that hormones play an important role in the process of leaf morphogenesis

The discrepancy of the expression level of LBL1 in different growth stages of cotton and the morphological changes of cotton leaves at different developmental stages can be comprehended that, the morphological formation of cotton leaves is a dynamic process. The expression of LBL1 in the BR treatment group is the highest compare to other groups, shows BR promotes the growth of cotton plants and accelerate leaf differentiation. IAA is an endogenous hormone that also promotes the growth and development of plants. Auxin is synthesized in the tip and transported to other parts of the plant, which leads to the apex advantage of plant growth, it also plays an important role in promoting leaf development and differentiation. GA3and 6-BA can also promote cell division, and effectively stimulate the growth of leaves and buds. The expression level in the ABA treatment group was significantly decreased, possibly due to ABA was a stress related phytohormone, the development and differentiation of a plant usually stop or delay while in a stress environment with a increased ABA content. The growth and development of plant leaves are determined by not only their genotypes but also environmental conditions.These factors regulate leaf morphogenesis through direct or indirect means, and interact each other, finally to form a complex effect on leaf differentiation and development.

5. Conclusion

The region harboring the deep lobed leaf shape gene in the cotton was accurately localized by using fine mapping approach. An HD-ZIP class I transcription factor gene was identified as the candidate gene controlling for deep lobed leaf shape in cotton with confirmatory analysis, sequence comparisons, and gene expression analyses. Our results provide an useful understanding of essential genetic and molecular processes that are responsible for the deep lobed leaf shape in cotton. This study will help in the breeding of selecting cotton genotypes with an ideal leaf shape, which may improve cotton production.

Acknowledgements

This work was supported by the Genetically Modified Organisms Breeding Major Projects, China (2016ZX0800 5 0 04, 2016ZX08009003-003-004), the National Natural Science Foundation of China (31601349), and the Innovation Program of Chinese Academy of Agricultural Sciences.

Appendicesassociated with this paper can be available on http://www.ChinaAgriSci.com/V2/En/appendix.htm