Rice Diversity Panel Mapping for Identifying Genetics Behind Leaf Vein Density Trait in Rice (Oryza sativa L.)

A screen of rice genotypes was carried out on a total of 327 genotypes of the rice diversity panel for leaf vein density. Identifying locations of genes that impart to leaf vein density in a quantitative way should capable the using of these genes in plant breeding and accelerate producing C4 rice plant. Finding of genotypes of rice that have less distances between veins is still the challenge of the plant breeders. Screen was shown significant variations in leaf vein density. Quantitative trait loci (QTLs) for the leaf vein density trait were identified by using Efficient Mixed Model Analysis (EMMA). QTLs were considered reportable if they had P values (below 0.0001). The most significant Single nucleotide polymorphisms (SNP) associations (EMMA 1.3, EMMA 1.8 and EMMA 10.1) were in each of the rice chromosomes 1 and 10 respectively. All genes positioned 200 kb around associations were selected. The candidacy of the most promising were NADP-dependent malic enzyme, chloroplast precursor (LOC_Os01g09320; ras-related protein, putative, expressed (LOC_Os01g51700); 60S acidic ribosomal protein (LOC_Os01g09510); Auxin-responsive Aux/IAA gene family member (LOC_Os01g09450; myb-related transcription activator, putative, expressed (LOC_Os01g09280); glycine-rich protein (LOC_Os01g09246); phosphofructokinase (LOC_Os01g09570); oxidoreductase, short chain dehydrogenase/reductase family (SDR) LOC_Os10g31780), which have been expressed in leaf tissue and requisite to be investigated further.


Introduction
To meet the ever growing world population, it is necessary to increase yield that can only be done by increasing the efficiency with which photosynthesis uses solar energy [1,2]. Plants are differing in canopy photosynthesis that leads to differences by about 50% between plants in the radiation use efficiencies [3,4]. C4 crops, grown in hot and dry environments, have higher yields, reduced water loss and increased nitrogen use efficiency compared to C3 crops such as rice. Plants use three photosynthetic pathways [5]. Plants that are called C3 produce a three-carbon compound by using photosynthetic pathway in which CO2 is fixed by Ribulose Bisphosphate Carboxylase/Oxygenase (RuBisCO) in the Calvin-Benson cycle. Whereas species, that evolved from C3 plants, use the C4 and Crassulacean Acid Metabolism (CAM) pathways, and in both cases, a four-carbon compound is initially formed from fixation of HCO3. It is suggested that C3 is the ancestral pathway. Thus, C4 and CAM are forms diverged from C3 plants in recent times. The majority of C4 plants, and certainly all known C4 grasses, compartmentalize photosynthetic reactions between two morphologically distinct cell types that are located in circles around veins [6]. There are enlarged bundle sheath (BS) cells around the veins that are surrounded by mesophyll (M) cells that called Kranz anatomy. This structure produces a consistent interveinal distance of four cells (vein-BS-M-M-BS-vein). Each of BS and M cells shared metabolic reactions such that C4 acids are generated in M cells and then diffuse to the BS where the Calvin-Benson cycle operates. Restricting the expression of a few numbers of genes to either the BS or M cells achieves this separation of metabolism. [7] reported that Carbonic anhydrase (CA), phosphoenolpyruvate carboxylase (PEPC), NADP-malate dehydrogenease (MDH), pyruvate orthophosphate dikinase (PPDK) and the proteins involved in their post-translational regulation accumulate in the M cells, while NADP-malic enzyme (ME) and RuBisCO are restricted to the BS. All of these enzymes encoded by genes that are present in C3 plants, but the levels of expression are much lower than in C4 species. According to the polyphyletic evolution of the C4 pathway, the transition from C3 to C4 seems to be relatively simple. the most obvious differences between leaf morphology in C3 and C4 plants is International Journal of Applied Agricultural Sciences 2016; 2(4): 49-55 50 leaf venation pattern indicating that veins play an important role in the differentiation of C4 leaf anatomy. Vein density measured in a range of C3 and C4 species demonstrated that veins are consistently more closely spaced in C4 species [8]. in addition, quantitative measurements of BS-to-M cell ratios in C3 and C4 leaves showed that in C4 plants the ratio approaches 1:1 [9,10,11]. This ratio equates to veins (V) being separated by only four photosynthetic cells in C4 leaves as opposed to up to 20 cells in C3 leaves. As such, the repeating V-BS-M-M-BS-V unit of Kranz anatomy is produced. One remarkable exception to this repeating model is found in Arundinella hirta, a C4 grass that exhibits an atypical anatomy where wreaths of so-called distinctive cells are found between V-BS-M-M-BS-V units [12,13]. The distinctive cells carry out the same function as BS cells but are not themselves associated with veins [14,15,16]. Notably, if the number of BS and distinctive cells is combined, the 1:1 ratio is also observed in Arundinella hirta (17) reported that a comparison of vascular development in C3 and C4 Flaveria species showed that both the major and minor veins were commenced at comparable stages in development but that a higher number of minor veins were commenced in the C4 species. An acquisition of a mechanism to induce procambium at more regular intervals across the leaf may be the first step in the evolution of Kranz anatomy. The role of auxin in vascular development is recognized, it is probably that such a mechanism was adapted from existing auxin pathways. A study that compared anatomical and biochemical differences between 16 Flaveria species that included C3, C3-C4 intermediate, C4-like, and C4 types gave more evidence for the suggestion that changed vein patterning was an early modification in the evolution of C4. To find accessions of rice with less interveinal distance is the challenge for plant breeder. To determine the extent to which rice leaf morphology can vary, it is important to screen rice genome wide association for leaf morphology to determine variations in interveinal distance.

Material and Methods
The layout of this experiment was randomized complete block design with four replications. A subset of 327 accessions was utilised for this study. Of those examined, 279 belonged to the rice subpopulations aromatic 9, aus 53, Photosynthetically active radiation (PAR) was supplied within temperature range from 28 -30°C and watered with Yoshida's full strength nutrient solution [18]. Vein counts were made off-site at green house -University of Aberdeen from images snapped of freshly collected leaves. Veins images in the second completely expanded newly harvested youngest leaf at 8 weeks after sowing were captured for screening the 327 rice diversity panel by a digital camera model canon (EOS 7D Mark II DSLR Camera, 20.2 MP). Veins were counted at the widest area of the leaf including the midrib or edges ( Figure  1). Genotyping of the Rice Diversity Panel has been done at Cornell University, New York on an Affymetrix genotyping array which includes of 44,100 SNPs distributed over the rice genome (380 Mb) [19]. For this SNP chip, the estimation was about 1 SNP/10 kb coverage [20]. The association mapping analysis was done by Dr. Alexander Douglas Statistician and Bioinformatician (University of Aberdeen). An Efficient Mixed Model Analysis (EMMA) accounting for population structure was performed on all the genotypes according to methods reported by [21] that modified from [22] who developed a novel mixed-model method to simultaneously take into account for many levels of relatedness identified by random genetic markers. In addition, this analysis with EMMA plus separate sub-populations is the same to the statistical method adopted by [19] in the first publication mapping traits using this SNP data on this population. The result presented by Dr Douglas provided were many files containing 4 pdf files (Histogram, Manhattan mixed, Manhattan naïve and QQ plot). In addition, the statistic for the minor allele frequency value that was taken from previous analysis of other data was performed, and has relation with the allele of the SNP that is of lowest frequency within the genotypes. The SNP was considered potentially not reliable, when this proportion was less than 0.05, Furthermore, analysis of the separate subspecies was done; the aromatic and admixtures were removed from the data set prior to association mapping analyses as there were insufficient numbers of those within these subspecies. As well as, the four subspecies of association populations were analysed for association mapping. It's important to mention that the literature proposes that there is no uniform threshold P value that can be considered in rice diversity panel [19]. In current study, QTLs were considered reportable if they had multiple close (within 200 kb) SNPs with low P values (below 0.0001) and where at least some of these SNPs did not have minor allele frequencies below 5%. Using the 51 Zaniab Statistical and bioinformatic analysis All the data were analysed using Minitab version 15. Twoway ANOVA with factors genotype and block was applied. The data were corrected for the block effect and also for normality by using base log 10 . Using EMMA, the association mapping analysis was made. The reliability of Trait-marker associations was with P values below 0.0001. The significant SNPs were tested for minor allele frequencies, with values above 0.05 considered to be reliable. All genes located with 200 kb of the QTL identified were selected as positional candidates. To distinguish between genes and likely psuedogenes, all the candidate genes have been tested for full-length cDNA (fl, cDNA) and expressed sequence tag (EST) [23]. In order to test gene expression in plant leaves, the database of RiceXPro at http://ricexpro.dna.affrc.go.jp/ was used [24]. Furthermore, the candidate genes with significant expression were examined for whether they are linked to leaf vein density or even to cell development through literature studying to decrease the number of candidate genes.

Results and Discussion
Leaf vein density was assessed at eight weeks after sowing. The output of One-way ANOVA showed that there were high significant differences in leaf vein density between the genotypes (F = 2.71, P < 0.001, R 2 = 30.03).   Figure 3. Use of EMMA revealed that three SNP are significantly associated with leaf vein density using a threshold value from the association analysis of under 0.0001 (-Log 10 P = 4) (Figure 4, Table 1). The most significant SNP association ware EMMA 1.3, EMMA 1.8 and EMMA 10.1 with minor allele frequency are 0.40, 0.37 and 0.46 respectively which indicates that this association is reliable. From association analysis for individual rice groups aus, indica, tropical japonica and temperate japonica with leaf vein density were detected ( Figure 5). In the present study only those SNPs detected in the mixed model have been taken forward for listing candidate genes.   Assessing of vein density of leaf in 327 rice accessions considered as a long term aim of introducing the mechanism of C4 into C3 plants. It was possible to show that there was significant variation across genotypes perhaps reflecting differences in the degree of the close relation between development of the vein and leaf width is associated to fracture proliferation in stretched material that leads to auxinlike signals in the new spaces in which new veins can be formed. Consequently, the link connecting the formation of auxin and vascular might explain the link between vein density and width of leaf [25]. In addition, the substantial variation in leaf vein density among genotypes was indentifying in current study which is confirmed by [25] who reported that there is variation in vein density, also by examining two rice mutants, they found that the trait of narrow leaf width was related to a high vein density trait but there was incomplete linkage. However more studies need to be done in order to exploit data above to set up future investigation. In addition, all genes situated 200 kb around associations, that include candidate gene lists detected by using the EMMA approach, were chosen ( Table 1). The RiceXPro database shows that many candidate genes are expressed in leaf tissue. From these, eight stand out after seeking gene function in the literature (Table 2).    NADP-dependent malic enzyme, chloroplast precursor (LOC_Os01g09320). NADP-malic enzyme is belongs to the family of oxidoreductases. In addition, the systematic name of this enzyme class is (S)-malate: NAD + oxidoreductase (decarboxylating) [26,27]. Most importantly, a study done by [28] confirmed that NADP-malic enzyme involved in C4 photosynthetic pathway in seven rice genotypes during grain filling stage. Furthermore, this gene is considered to be excellent candidate because it is linked to function in pyruvate metabolism and carbon fixation and has experssion intensity in leaf tissue of approximately 12000 Cy3.
Ras-related protein, putative, expressed (LOC_Os01g51700). The rgp1 is a gene that encodes rasrelated GTP-binding protein. [29] reported that transgenic tobacco plants expressing rgp1, show different morphological characteristics. For example, some R1 progenies of selfpollinating plants developed abnormal leaf phenotypes, such as the abnormal branching of the main vein leading to the formation of twin or even triple leaves [29]. This may nominate this gene as a good candidate gene. The database RiceXPro reported that this gene had expression intensity in leaf tissue just about 4200 Cy3.
60S acidic ribosomal protein (LOC_Os01g09510). The GTPase large subunit GTPase 1 (LSG1) is responsible for the steps through which the 60S ribosome is been maturated. An Arabidopsis mutant dig6 exhibited multiple auxin-related phenotypes such as altered leaf veins. Furthermore, the expression of this gene was found to be high in regions where the auxin can be accumulated and have role in the development of ribosome [30]. The database RiceXPro showed that this gene had expression intensity in leaf tissue reaching nearly 100000 Cy3.
Auxin-responsive Aux/IAA gene family member (LOC_Os01g09450). Gene expression regulates partially the acts of Auxin. Genes regulated by the Auxin can be classified into many gene families, such as the Aux/IAA, the SAUR and the GH3 family [31]. In addition, different growth and development stages of plant are mediated by auxin and are transported all through the plant [32]. The AXR2 gene was cloned by [33] from Arabidopsis plant showing that it is like to IAA7 gene, which is one of gene family that is responsible for making morphological responses to normal light.
RiceXPro database showed that this gene had an experssion intensity in leaf tissue of roughly 10000 Cy3.
Myb-related transcription activator, putative, expressed (LOC_Os01g09280). MYB proteins constitute a large family of transcription factors [34]. [35] confirm the important role of the MYB family in the metabolism and development of plant. Moreover, RiceXPro database suggested this gene had expression intensity in leaf tissue that reach of about 1250 Cy3.
Glycine-rich protein (LOC_Os01g09246). Glycine-rich protein is a group of proteins that responsible for cell wall structure. This group of proteins occurs in several higher plant species and its expression in vascular tissue of rice was reported by [36]. RiceXPro database showed that this gene had expression intensity in leaf tissue that about 7000 Cy3.
6-phosphofructokinase (LOC_Os01g09570). Isozymes of 6-phosphofructokinase are found in the cytosol and plastid compartments and this gene have role for developing leaf tissue [37]. In addition, RiceXPro database revelled that this gene had expression intensity in leaf tissue around 30000 Cy3.
Oxidoreductase, short chain dehydrogenase /reductase family (SDR) LOC_Os10g31780). This gene has very large family consists of at least 140 different enzymes. The SDR family comprise ABA2 [38,39]. The ABA2 gene, which involved in ABA biosynthesis, encodes OST1 gene. This gene encodes a protein, which controlled by ABA is responsible for stomatal closure [40]. In addition, RiceXPro database showed that this gene had an expression intensity in leaf tissue namely about 150000 Cy3.
Improving high yielding rice genotype is a target of rice breeders. Therefore, understanding the mechanisms affecting leaf vein density is an important matter for rice breeding. However, the mechanisms underlying leaf vein density are complicated.

Conclusion
Detecting quantitative trait loci (QTLs) which confer leaf vein density promises to speed up the aim of this experiment. In this study differences in leaf vein density between rice genotypes has been exposed and QTL identified. These data point out areas of the rice genome containing genes of potential value in breeding leaf vein density rice. These candidate genes are worthy further investigation.