Abstract: In this study the genetic diversity among 27 rapeseed (B. napus) cultivars and one cultivar of B. rapa was investigated using Random Amplified Polymorphic DNA (RAPD) marker. A set of 24 arbitrary primers was used which produced 133 reliable polymorphic DNA bands ranging in size from 440 to 2988 bp. RAPD data were used to compute genetic similarities based on Jaccrad (J) and Simple Matching (SM) indexes. Similarity coefficients values ranged from 0.22 to 0.87 with a mean of 0.48 for J and from 0.13 to 0.61, with a mean of 0.31 for SM coefficient, for all genotypes. Similarity matrices were used to contrast dendrograms using Unweighted Paired Group of Arithmetic Mean (UPMGA) algorithm. In cluster analysis, based on both similarity coefficients, B. napus cultivars were grouped together and Hysyn110, which belongs to B. rapa was classified as an outlier. Winter and spring B. napus cultivars were clustered into two separate groups. Results of Principal Component Analysis (PCA) supported cluster analysis results strongly. RAPD profiling was found efficient to reveal DNA polymorphism among rapeseed cultivars. RAPD data showed relatively low level of genetic diversity among studied cultivars, but also revealed some genetically distinct cultivars. This finding should be useful to breeding programs for selection of the genetically divergent parents in order to obtain the highest level of heterosis.
INTRODUCTION
Oilseed rape (Brassica napus L.) is one of the most important oilseed crops worldwide. B. napus is an amphidiploid (2n = 38, AACC) that has been arisen by spontaneous interspecific hybridization between two diploid species, B. rapa (syn. campestris) (2n = 20, AA) and B. oleracea (2n = 18, CC) (U, 1935). B. napus has evolved through different combinations of the diploid morphotypes (Song et al., 1988). RFLP analysis on organelle and nuclear genome strongly support the concept of multiple origins of B. napus (Song and Osborn, 1992).
B. napus is predominantly self-pollinated with about 20% out-crossing (Luhs et al., 1994). Most oilseed rape cultivars are open-pollinated populations derived from inbred plants (Ripley and Beversdorf, 2003). There is also interest in developing hybrid oilseed rape cultivars because high levels of heterosis have been reported (Grant and Beveresdorf, 1985). Increasing oil content and development of varieties with enhanced oil quality is one of the most important breeding goals for rapeseed (Hu et al., 1999). There is intensive attempt in developing rapeseed cultivars for different growing regions.
Characterization of genetic diversity in plant varieties and breeding lines is essential for effective conservation and utilization of plant genetic resources. These information are essential for implementation of plant breeding strategies. Molecular markers, especially DNA markers, have been employed for analysis of genetic diversity, because of their advantages. Their expression is independent of environmental conditions and potential number of polymorphic molecular markers is nearly unlimited. One of the most extensively used molecular markers is Random Amplified Polymorphic DNA (RAPD). The main advantages of RAPD technology are its simplicity and low cost. It does not require any digestion by restriction enzymes, any radioactive probes or previous knowledge about the sequences and require small amount of DNA (Williams et al., 1990). RAPD provide a level of resolution equivalent to RFLPs for determination of genetic relationships between genotypes at the intraspecific level (dos Santos et al., 1994; Graham et al., 1996).
Several molecular markers such as RFLPs, AFLPs, microsatellites and isozymes have been used for genetic analysis in B. napus germplasm (Diers and Osborn, 1994; Lombard et al., 2000; Pliesk and Struss, 2001; Zhao-Jian et al., 1998). Several studies have confirmed the significance of RAPD marker as a powerful tool in evaluation of genetic diversity in B. napus cultivars and its wild relatives (Dulson et al., 1998; Lanner et al., 1996; Mailer et al., 1997). RAPD markers are useful for investigation of genetic relationships among cultivars in other plant species (Shiran et al., 2004; Diaby and Casler, 2003; Fu et al., 2003; Shiran and Raina, 2001; Stallen et al., 2000). RAPD also has been used for constructing genetic linkage map in B. napus (Foisset et al., 1996; Lombard and Delourme, 2001). RAPD and SCAR markers associated with low linolenic acid and other important traits in rapeseed have been identified (Hansen et al., 1997; Rajcan et al., 1999; Zhang et al., 2003).
The objectives of the present study are to study the extent of genetic diversity among B. napus cultivars and to identify the genetic relationships among them.
MATERIALS AND METHODS
Plant material: Twenty-seven B. napus cultivars including both spring and winter types with canola quality were studied (Table 1). Most of cultivars were chosen to represent the genetic diversity in rapeseed germplasm. Also, these double-Zero cultivars represented breeding lines currently used in conventional breeding systems in order to introduce the most proper cultivars for commercial cultivation in Iran. This collection included open pollinated and hybrid rapeseed cultivars from different origins. Finally, one cultivar of B. rapa (Hysyn110) were included for comparison to B. napus cultivars. Seeds were supplied with Seed and Plant Improvement Institute-Oilseed Crop Research Department, Karaj, Iran.
Total DNA isolation and RAPD fingerprinting: Total cellular DNA was isolated from freshly germinated young leaves following a modification of the CTAB method of Murray and Thompson (1980): about 1-1.5 g of fresh leaf tissue was crushed to powder in liquid nitrogen and transferred to pre-warmed (60°C) 2xCTAB buffer. The DNA was purified by phenol-chloroform extraction and ethanol precipitated. The final DNA pellet was dissolved in TE buffer, were stored in -20°C until use.
An initial screening of 40 random primers (Genset, France) was performed in order to evaluate their capacity to produce polymorphic, reproducible and reliable RAPD patterns.
Table 1: | Characteristics of accessions used in this study |
Table 2: | Primers included in the study |
Twenty-four primers selected (Table 2). Polymerase chain reaction was performed based on the protocol of Williams et al. (1990), with minor modification. Amplification were carried out in 25 μL of reaction mixture containing 25 ng of genomic DNA, 0.5 unit of Taq DNA polymerase (Amersham biosciences), 2.5 μL of PCR buffer, 240 μM each of dNTPs (Amersham biosciences), 15 ng of the primer. DNA amplification was performed in an Eppendorf Mastercycler Gradient programmed as follows: pre-denaturation 3 min at 92°C followed by 45 cycles, each of 1 min at 92°C, 1 min at 35°C and 2 min at 72°C, followed by one final extension cycle of 10 min at 72°C. The amplification products were size-separated by gel electrophoresis in 1.2% agarose gels with 0.5xTBE and stained with ethidium bromide.
Fig. 1: | Random amplified polymorphic DNA banding patterns produced using OPG10 (a) and OPG19 (b) on a subset of 28 rapeseed cultivars |
EcoR1/HindIII digested lambda DNA served as molecular size marker. PCR reactions were repeated at least twice to establish reproducibility of results. After AGE, the gel was photographed in UV light (Fig. 1). The molecular size of each fragment was calculated by using UVI bandmap software 10.02. Special care was taken to eliminate variation in DNA and any other reaction reagent concentration and to ensure consistent reaction conditions.
Data analysis: All polymorphic bands were scored as (1) for present and (0) for absence. Only reproducible and reliable bands were scored. Bands of identified size amplified with the same primer were considered to be the same DNA marker. Two similarity coefficients were computed, the Jaccard's coefficient (Jaccard, 1908) and Simple matching coefficient (Sokal and Michener, 1958). Cluster analysis was done and dendrogram were constructed using unweighted paired group of arithmetic mean (UPGMA) algorithm.
The structure of the genetic diversity was further analyzed by Principal Component Analysis (PCA) from the correlation matrix among the markers. Both cluster analysis and principal component analysis were performed using NTSYS 2.02 statistical software.
RESULTS
Twenty four primers were selected to amplify total DNA from samples of all genotypes among which amplification products of 22 primers (91.67%) showed polymorphism. A total of 173 reproducible bands amplified against 28 genotypes among which 133 products (76.88%) were polymorphic. Among total scored bands 164 fragments observed only in B. napus cultivars.
The number of amplicons generated varied from 5 (OPG1, OPG2, OPG7, OPG15, OPG16) to 18 (OPG8) and presented molecular weight between 440 to 2988 bp. The ability of primers made a wide range of polymorphism varied from 20% (OPG15) to 100% (OPG7, OPG11). OPG21 and OPG24 failed to reveal any polymorphism. RAPD profile produced using OPG11 against Pronto was clearly different from other ones. OPG1 and OPG16 produced uniform patterns for all B. napus cultivars. One fragment (OPG5-1419 bp) was produced only in Option500 and Option501. OPG5 also produced a band (1788 bp) in the profiles of two cultivars (PF7045/91 and Foseto). OPG11 produced a fragment of 2863 bp in the RAPD pattern of Hysyn110 and Fusia and another one (OPG11-663 bp) at Eurol. These polymorphic bands were not presented into the patterns of other cultivars.
The distance matrices based on the two similarity coefficients, Jaccard and SM, were highly correlated. Genetic distances ranged from 0.22 to 0.87 with a mean of 0.48 for J and from 0.13 to 0.61, with a mean of 0.31 for SM. If we ignore Hysyn110, the average of GDs will reduced to 0.45 for J and 0.29 for SM. Whatever the estimator considered, Kristina and SLM-046 were the closet B. napus cultivars and LG3310 (a Canadian spring cultivar) and Licord (a German winter cultivar) had the largest GD.
Fig. 2: | Dendrogram illustrating genetic relationship among 27 cultivars of Brassica napus and one cultivar belongs to Brassica rapa, generated by unweighted paired group of arithmetic mean (UPGMA) cluster analysis based on Jaccard coefficient |
The Canadian spring cultivar, Hylight201, had the largest genetic distance with Hysyn110, 0.87 for J and 0.61 for SM.
Cluster analysis was done using UPGMA method and dendrograms were constructed (Fig. 2). Dendrogram based on the J distance matrix defined one major group, consisted of all B. napus cultivars and Hysyn110, that belongs to B. rapa, classified as one outlier. The B. napus cultivars were distinctly separated into two subgroups and Pronto clustered as an outlier. The first subgroup included winter rapeseed cultivars with the exception of Sponsor, one spring type cultivar and the second one included spring cultivars with the exception of SLM-046 and Regent x Cobra, two winter type cultivars. At lower hierarchy, Foseto (a Mexician spring cultivar) and then Chinese (a Chinese spring cultivar) were separated from all other cultivars in their subgroups as outliers. Dendrogram constructed based on SM coefficient was the same; with this exception that Regent x Cobra clustered with winter type B. napus cultivars.
RAPD data also analyzed using PCA. Associations among the cultivars revealed by PCA were represented in Fig. 3. The first principal component explained about 54.4% of total variation among cultivars, this PC separated B. napus cultivars from Hysyn110, an cultivar belongs to B. rapa. The second PC explained about 6.7% of total variation among cultivars and separated winter and spring B. napus cultivars into two major groups.
Fig. 3: | Scatter plot based on the first and the second principal components from a principal component analysis of RAPD data, based on Jaccard coefficient |
DISCUSSION
The present study demonstrated that RAPD profiling is efficient enough to reveal usable level of DNA polymorphism among rapeseed cultivars. This supports the earlier observations with regard to the RAPD profiling in B. napus (Hallden et al., 1994; Mailer et al., 1994, 1997). In this study we used 24 decamer random primers in RAPD analysis of 28 genotypes among which 91.67% produced polymorphic fragments. OPG1, OPG16, OPG21 and OPG24 failed to reveal any polymorphism among B. napus cultivars. Therefore, they would not be recommended to use in B. napus germplasm studies. OPG5, OPG11 and OPG19 had the most efficiency in rapeseed genotype discrimination.
Several variety specific RAPD markers were produced for Option500- Option501 (OPG5-1419 bp), PF7045/91- Foseto (OPG5-1788 bp), Hysyn110- Fusia (OPG11- 2863 bp) and Eurol (OPG11-663 bp). These two primers can be employed in varietal registration studies. Specially, about unique band which produced in Eurol, this band can be used for identification of this cultivar from other cultivars if this locus can be proven to be homozygous by examining several plants of this cultivar. Also, if association between these RAPD markers and important traits or special genes can be demonstrated through bulk segregant analysis (Michelmore et al., 1991), it will be possible to convert them into codominant SCAR markers (Roy, 2000).
The distance matrices based on two similarity coefficients were highly correlated. Lombard et al. (2000) demonstrated that J and SM similarity coefficients were highly correlated among B. napus cultivars.
The choice of a genetic distance is a crucial issue for estimating the level of relatedness between cultivars. GDJ takes into consideration only matches between bands-alleles that are presented and ignores pairs in which a band-allele is absent in both individuals. For dominant markers such as RAPD, the expected GDJ of related pairs of lines is a linear function of their coancestry coefficient (Link et al., 1995). GDSM, a Euclidian measure of distances, take into account mismatches and matches; it has metric properties that allow its usage in hierarchical strategies such as the minimum variance method within a group (Ward, 1963). However, many researchers do not prefer using GDSM as it gives equal weight to both 0-0 and 1-1 matches in case of binary data (Mohammadi and Parsanna, 2003).
The average GDs among investigated rapeseed cultivars, 0.48 for J coefficient and 0.31 for SM, demonstrated that the level of genetic diversity among them is relatively low. However, according to its autogamous nature, it was predictable. The relatively low level of genetic diversity among rapeseed cultivar is unavoidable since the breeding system is an influential factor of variability within plants and populations of self pollinated species are less polymorphic than out breeding species (Campos-de-Quiroz and Ortega-Klose, 2001). Correlation between the level of intraspecific variation and the reproductive system has been observed for RAPD markers in other species (De Bustos et al., 1998; Kazan et al., 1993). Also, we should consider that low level of polymorphism in cultivars could be attributed to a narrow genetic base and the frequent inbreeding involved in breeding programs (Pujar et al., 1999).
In this study, B. napus cultivars and Hysyn110 which belong to B. rapa were separately clustered. Cluster analysis also separated winter and spring type B. napus cultivars into two separated clusters. PCA revealed the same structure as the cluster analysis. On the basis of the results of PCA, B. napus cultivars and Hysysn110 were distinctly separated. PCA also separated winter type B. napus cultivars from spring ones. Diers and Osborn (1994) reported similar result using RFLP markers. In their study most of winter and spring B. napus cultivars were separated into two different groups. Other similar results have been demonstrated by Plieske and Struss (2001) using microsatellite markers and by Lombard et al. (2000) using Amplified Fragment Length Polymorphism (AFLP) markers. Winter and spring B. napus cultivars are fairly distinct groups and there has been little intercrossing between these groups in breeding programs (Sernyk and Steffansson, 1983).
RegentxCobra, an Iranian winter cultivar grouped with spring types in the phenetic analysis based on J coefficient. In principal component analysis, based on both coefficients RegentxCobra, positioned between spring and winter cultivars. This is evidence that although this cultivar grouped with spring cultivars, its association is not strong. Therefore, further studies must be done in order to identify cultivation type of this cultivar, exactually. Sponsor (a spring type cultivar) clustered with winter types. These results demonstrated that Sponsor has more genetic similarity with winter type cultivars. Unexpected, SLM-046 one winter type cultivar, clustered with spring type cultivars. It should be considered that accuracy of genetic similarity estimates base on molecular markers depends on several variable factors such as the number of markers analyzed, their distribution over the genome and the accuracy in scoring of markers (Schut et al., 1997). Moreover, any seed contamination, both mechanical and genetically contamination, might be accrued.
In conclusion, this study demonstrated that RAPD markers provide an effective tool in assessment of genetic variation among rapeseed cultivars and in cultivar identification. This information could be used for selection of most divergent pairs of parents, which could be used for genetic linkage map construction in this species. Information on the genetic diversity in B. napus and germplasm also can help breeders to select the best combinations, which produce high level of heterosis. The observed genetic variation among investigated rapeseed cultivars found in analyzed loci should alert the breeders for introducing new genetic sources in hybridization programs. Investigated collection includes rapeseed cultivars with different cultivation type and different origins. However, direct effort to diversify the genetic base of breeding material and to seek new sources of favorable alleles would be still necessary. Accessions belong to B. rapa species and other wild relatives of B. napus could be useful to broad the genetic basis of rapeseed germplasm.
ACKNOWLEDGEMENTS
The authors would like to thank Dr. S.E. Tabatabai, from Isfahan university of Technology. The authors would also like to thank Mr. Falahati for his helpful advice. This research was supported by Shahrekord University.