Intervarietal Hybridization and Genetic Diversity of Rice by Molecular Markers

Cross ability of various accessions of Oryza L . on the basis of pollen development and seed set was studied along with genetic diversity by RAPD and SSR markers. The variety Kalanamak had maximum number of fertile pollen (82.25%) whereas the variety Narendra-359 had the least number (61.03%). The cross of CAU R-1 with Hansraj (83.67%) gave the highest germination % after 1 hour. The cross of CAU R-1 with Narendra-359 gave the maximum seed set (68.96%). The pollen tube growth at 1 hour after pollination had a positive correlation with seed set. In general, selfing showed more pollen germination and pollen tube growth as compared to inter-varietal crosses with respect to seed set. Fourteen RAPD and eight SSR primers were used to assess the genetic diversity of 17 rice varieties. A total of 78 RAPD and 16 SSR amp icons were generated. The value of Jacaard’s similarity coefficient for RAPDs ranged from 0.284 to 0.766 with an average value of 0.526 whereas for SSRs it ranged from 0.071 to 1 with an average value of 0.396. The UPGMA cluster analysis grouped the 17 rice varieties in four clusters in case of RAPDs and three clusters in case of SSRs. The results of principal component analysis were comparable to the cluster analysis. In the Principal Component Analysis (PCA), for RAPD the first three components explained 65.86% of the total variation, with 53.21% explained by the first component and 7.23% by the second component, whereas for SSR the first three components explained 68.98% of the total variation, with 41.27% explained by the first component and 17.71% by the second component. The correlation coefficient and the significance of the correlation of the matrices based on RAPD and SSR data tested by the Mantel test showed that non-significant correlation (r=0.45) existed between both matrices.


Introduction
Hybridization is one of the most commonly used breeding methods for improvement, mainly of open pollinated and often cross pollinated crops. The genetic improvement through hybridization and seed selection has been the main objective in the breeding programmes in rice. Inter-Varietal crosses have been used almost exclusively in the development of varieties by hybridization. The process of pollination and pollen germination leading to fertilization determines seed setting, and these processes are negatively affected by high temperature stress coinciding with a thesis [1]. Shi Qiang [2] reported that the rice pollen grain starts to germinate at 2min after pollination and the pollen tube penetrated stigma into style in 5-10 minutes, 30 minutes later the end of pollen tube reached the bottom of ovary, and only some pollen tubes arrived at embryo sac at 40 minutes after pollination. Variation in pollen tube growth rates in the pistil is frequently cited as a phenotypic manifestation of differences in gametophytic quality, leading to differences in the reproductive success among male gametes [3,4]. Jaitly, Khanna [5] reported that inadequate pollination, low % pollen germination and to some extent slow pollen tube growth and pollen tube abnormalities affect crossability in inter varietal crosses and the backcrosses in Oryza sativa accessions. RAPD marker was first described by Williams, et al. (1990).
Usually decamer primers are used to amplify the homologous sites of the target genomes. Polymorphisms are detected as presence or absence of bands. These markers have been used widely in many species because of their applications without prior sequence information. However, RAPD is very sensitive to reaction conditions and dominant in nature. Simple sequence repeat is an important species with closer genetic relationship. Kibria et al. [11] studied molecular marker based genetic diversity analysis in aromatic rice genotypes by both SSR and RAPD markers through Marker Assisted Selection (MAS). They found that SSR markers are more effective in getting higher genetic diversity; however OPA 02 and 67 AB10G7 (RAPD) primers gave 100% polymorphism.  The gel was visualized and photographed in a Gel Documentation system. Molecular weights of bands were estimated by using 1Kb ladder for RAPD and 100bp for SSR. The homology of bands was based on distance of migration in the gel. RAPD and SSR amplicons obtained from each entry were resolved as multiple and a single band on the agarose system, respectively and the data set were used to calculate pair wise similarity coefficients following Jaccard [14]. The similarity matrices constructed were subjected to cluster analysis by unweighted pair group method of arithmetic average (UPGMA) analysis to generate dendrogram. These computations were performed using NTSYS-PC ver. 2.02 j, Exeter Software [15].

Materials and Methods
Mantel's correlation test was performed by calculating correlations between Jaccard's similarity coefficients and cophenetic values for each pair of comparisons [16]. The raw data matrix was used to calculate correlations between variables. The correlation matrix was subjected to "Eigen" vectors analyses, following which the principle components were extracted using the "Projection" module in NTSYS-PC. The first three most important PCA were used to construct a three dimensional plot of the accessions.

Pollen fertility
Pollen fertility was recorded in all the six parents. The fertile pollens picked up the acetocarmine stain and were stained red.    (Figures 1-9).

Correlation
The correlation studies were done for the pre fertilization factors in the inter varietal crosses performed for the seed set. A positive correlation of seed set with pollen tube growth after 60 min was established. A positive correlation with pollen germination was established which was nonsignificant at 5% level of significance (Table 4).

RAPD Analysis
All the 17 cultivars were subjected to PCR amplifications using 14 RAPD primers. A total of 78 amplification products were scored in the 17 cultivars with different primers, which exhibited an overall 92% polymorphism ( Table 5). The average numbers of amplification products formed were 5.57 with a maximum of 8 (OPB 1; OPD-06; OPF 14) and a minimum of 2 (OPK 10). The size of the amplification products varied in case of each primer and the range was 0.10 kb to 1.85kb. In general, the extent of polymorphism found was high. Eight out of 14 primers showed 100% polymorphism.
Three primers (OPB 1; OPD-06; OPF14) were found to be most polymorphic whereas OPK 04 was the least polymorphic primer ( Table 5). The data obtained from RAPD analysis were subjected to UPGMA analysis to find out the relationship between the cultivars analyzed. The value of Jacaard's similarity coefficient ranged from 0.284 to 0.766 with the average value of 0.526 ( Figure 10).    generated with RAPD data were tested by 2-Way Mantel test [16].
High support for clustering patterns was observed for the cluster with Matrix correlation (r) as 0.82.
In the Principal Component Analysis (PCA) (Figure 12) (Table 6). The data obtained from SSR analysis were subjected to UPGMA analysis to find out the relationship between the cultivars analyzed. The value of Jacaard's similarity coefficient ranged from 0.071 to 1 with the average value of 0.396 ( Figure 13).     The Goodness-of-Fit of the UPGMA dendrogram generated with RAPD and SSR data were tested by 2-Way Mantel test [16].
High support for clustering patterns was observed for the cluster with Matrix correlation (r) as 0.92. In the Principal Component