Introduction
Rice varieties grown in Sri Lanka from ancient to middle of the last century are known as traditional varieties. High yielding new improved varieties have been produced from crosses between traditional and exotic genotypes. Almost all rice varieties grown today are new improved varieties. However, traditional genotypes are still conserved in situ and/or ex situ in Genebanks. As these locally adapted genotypes may contain traits/genes of economic importance, their characterization is essential to use the available genetic diversity. AFLP technique is a robust, reliable and highly informative DNA fingerprinting method used to assess a large number of traits without prior sequence knowledge. Therefore, AFLP is widely accepted as an effective tool for identifying genomic differences (Loh et al., Reference Loh, Kiew, Kee, Gan and Gan1999). Furthermore, fluorescent labelling in fluorescent AFLP (FAFLP) replaces the radioactive labelling and increases the throughput by enabling automated detection and scoring of fragments generated in AFLP.
The objective of this study was to assess the genetic diversity among different rice accessions available in the Genebank of Sri Lanka by FAFLP markers.
Materials and methods
DNA extraction
Seeds of rice accessions (Table 1) were obtained from the Genebank of the Plant genetic Resources Center, Sri Lanka. They were planted in pots filled with soil collected from a paddy field and were grown in a green house. Tender leaves were harvested after 2 weeks and stored at − 80°C. DNA was extracted from leaves according to the method described by Chen and Ronald (Reference Chen and Ronald1999). Concentrations of DNA were estimated, and concentrations of all DNA samples were adjusted approximately to 300 ng/μl.
List of traditional rice varieties (TR), cultivated varieties (CV), wild rice species (WS) and hybrid variety (HV) used in this study
AFLP analysis
AFLP analysis was carried out according to Vos et al. (Reference Vos, Hogers, Bleeker, Reijans, van de Lee, Hornes, Fritjers, Pot, Peleman, Kuiper and Zabeau1995) with modifications. DNA was digested with enzymes EcoR1 and Mse1 at 37°C for 3 h and 30 min. Oligonucleotide adapters were ligated to digested DNA by incubating DNA, and adapters with T4 DNA ligase at 37°C for overnight. Then, digested/ligated DNA was pre-amplified with pre-amplification primers. The pre-amplified products were diluted 20 times with sterile distilled water and used for selective amplification.
Selective amplification reactions were performed using pre-amplified DNA and fluorescently labelled (HEX, TMR or FAM) EcoR1 primers and unlabelled Mse1 primers. Ten different primer combinations were used for selective amplification. The products were purified by ethanol precipitation followed by washing with 70% ethanol. The dried pellets were re-suspended in 5 μl of water. Finally, 2.5 μl of re-suspended sample was mixed with ET 550-ROX size standard (GE Healthcare Life Sciences, Amersham Place, Little Chalfont, Buckinghamshire, HP7 9NA, UK) and deionized formamide according to the manufacturer's instructions and denatured at 95°C for 2 min. Fragments were resolved using capillary electrophoresis on MegaBACE 1000 automated DNA sequencer (GE Healthcare Life Sciences). AFLP fragment analysis was performed with Genetic Profiler software 2.2 (GE Healthcare Life Sciences).
Data analysis
The electropherograms in the range of 30–550 bp were analyzed by Genetic Profiler software version 2.2 (GE Healthcare Life Sciences). Each fragment size was treated as a unique character and converted to binary data (present, 1; absent, 0). Jaccard's (Reference Jaccard1901) similarity coefficients (J) were calculated, and similarity coefficients' matrix was constructed. The matrix was used for cluster analysis by unweighted pair group method with arithmetic mean (UPGMA) (Sneath and Sokal, Reference Sneath and Sokal1973), and the dendrogram was constructed by Multivariate Statistical Package (MSVP 3.1; Kovach, Reference Kovach1998). The confidence of the UPGMA clusters was assessed by Mantel (Reference Mantel1967) test to calculate the Cophenetic correlation coefficient (r) as described by Abhijit et al. (Reference Abhijit, Kamlesh, José, Vyankatesh, Milind, Dilip and Yogesh2004). Cophenetic correlation coefficient is a comparison between the dendrogram and the similarity matrix.
Results and discussion
Ten primer combinations generated 784 fragments. Of which 772 were polymorphic (98.4%) and 12 (1.6%) were monomorphic. The number of amplified products generated by each primer pair ranged from 41 to 171 with an average of 78.4 fragments.
J showed that the genetic similarity varied from 0.083 to 0.565. The traditional rice Dikwee and At353 showed the lowest similarity (0.073), while Mudukiriel and Bw267-3 showed highest genetic similarity (0.565). The UPGMA dendrogram (Fig. 1) separated the accessions into four main clusters at the similarity coefficient of 0.186.
The UPGMA dendrogram showing genetic diversity among Sri Lankan rice varieties.
Clusters I and IV contained only one accession, while clusters II and III comprised 37 and 38 accessions, respectively. A modern rice variety (Taichung Native 1 (TN-1)), Hygroryza aristata and five wild species were also included in this study to find out the reliability of data and analysis. H. aristata previously classified under genus Oryza but now separated as a different genus was introduced in the analysis as an outlier. H. aristata was found to be the most divergent line (cluster I) by separating from the rest at the similarity coefficient of 0.130. All accessions in other clusters belong to genus Oryza. Cluster IV encloses only TN-1, which is the first Indica variety carrying the Dee-geo-woo-gen gene (semi-dwarfing gene), and TN-1 is the most susceptible variety for pest and diseases and photoperiod-insensitive and known to be different from rest of the varieties used. These results confirmed the reliability of data and methods of analyses.
Wild species Oryza nivara, Oryza rufipogon, Oryza rhizomatis, Oryza granulata and Oryza eichingeri showed clear separation from other accessions representing cluster II. Most of the traditional rice varieties known to be introduced from India (Molaga Samba, Pachchaiperumal, Vellaiperunel, Vellai Illankalayan, Peillianel and Murunga) were in cluster III. These plants had been cultivated in northern part of Sri Lanka during 1930s.
The Cophenetic correlation coefficient (r) from the comparison between the dendrogram and the similarity matrix was 0.781. The high value of Cophenetic correlation coefficient (r) indicates that the UPGMA dendrogram represents similarity data accurately. Principal component analysis of those 81 rice accessions based on the Gower's general similarity coefficient (Gower and Legendre, Reference Gower and Legendre1986) also confirmed the distribution of clusters in UPGMA analysis (data not shown). Scatter diagram of first three coordinates (PCo1, PCo2 and PCo3) revealed three well-defined groups. All wild species were found within the same group. Rice varieties with Indian origin were only found in two other groups, while all other verities were scattered among these groups.
The genetic diversity of rice varieties revealed by this study is useful in categorizing the accessions and preventing duplications in core collection in Genebanks. In addition, this information at molecular level can be integrated to devise strategies for ex situ and in situ genetic conservation, their utilization and exchange of genetic material.
Conclusion
AFLP analysis of the rice varieties revealed significantly high genetic diversity among the Sri Lankan rice germplasm used in this study and degree of genetic resemblance/distance of each one of those varieties. This genetic diversity data will provide more direct and reliable genetic information for selecting suitable parents in rice breeding programmes. Furthermore, the information given here may be useful in the management of in situ and ex situ preservations of rice germplasm.
Acknowledgements
The authors greatly appreciate the financial support given by the National Research Council of Sri Lanka grant no. 05-61. Technical assistance of Ms Anoma Jayasoma is gratefully acknowledged.
