Phylogenetic Relationships Amongst 10 Durio Species Based on Pcr-rflp Analysis of Two Chloroplast Genes

Twenty seven species of Durio have been identified in Sabah and Sarawak, Malaysia, but their relationships have not been studied. This study was conducted to analyse phylogenetic relationships amongst 10 Durio species in Malaysia using PCR-RFLP on two chloroplast DNA genes, i.e. ndhC-trnV and rbcL. DNAs were extracted from young leaves of 11 accessions from 10 Durio species collected from the Tenom Agriculture Research Station, Sabah, and University Agriculture Park, Universiti Putra Malaysia. Two pairs of oligonucleotide primers, N1-N2 and rbcL1-rbcL2, were used to flank the target regions ndhC-trnV and rbcL. Eight restriction enzymes, HindIII, BsuRI, PstI, TaqI, MspI, SmaI, BshNI, and EcoR130I, were used to digest the amplicons. Based on the results of PCR-RFLP on ndhC-trnV gene, the 10 Durio species were grouped into five distinct clusters, and the accessions generally showed high variations. However, based on the results of PCR-RFLP on the rbcL gene, the species were grouped into three distinct clusters, and generally showed low variations. This means that ndhC-trnV gene is more reliable for phylogenetic analysis in lower taxonomic level of Durio species or for diversity analysis, while rbcL gene is reliable marker for phylogenetic analysis at higher taxonomic level. PCR-RFLP on the ndhC-trnV and rbcL genes could therefore be considered as useful markers to phylogenetic analysis amongst Durio species. These finding might be used for further molecular marker assisted in Durio breeding program.


INTRODUCTION
Malaysia is one of the centers of origin of durian (Durio spp.). There are 27 species of the genus Durio in the country, mainly grow in Malaysian Borneo (Sabah and Sarawak) (Popenoe 1974;Soon and Lum 1984;Nanthachai 1994). Twenty-three species of Durio have been recorded, while 10 species had already been characterized . In Sarawak, 16 species have been recorded, which seven of those are considered as bearing edible fruits (Gadug and Voon 2000). The most economically important species is Durio zibethinus Murr. (Tinggal et al. 1994).
Studies on population structure and genetic diversity are essential for the conservation and management strategies of the genetic resources (Karp et al. 1997). Phylogenetic analysis is a population genetic study attempting to define the evolutionary history of plants. Current plant populations are grouped to explain their development from their ancestor (Henry 1997). According Bretting and Widrlechner (1995), plant genetic resource management comprises of germplasm conservation and encouraging utilization. Effective conservation is required to maximize and ensure that its genetic potential will be fully available for use in the future (Esquinas-Alcazar 1993).
The analysis of genetic relationships among samples is started with the construction of a matrix specifying the characters-state of each marker for each sample (Karp et al. 1997). The result is often presented in the form of a matrix of similarities or dissimilarities (Henry 1997) and illustrated by means of a phylogenetic tree (Li and Graur 1991) or dendrogram tree (Karp et al. 1997), where aggregation of samples linking together in clusters are more genetically similar each other than to samples in other clusters.
Polymerase chain reaction-based restriction fragment length polymorphism (PCR-based RFLP) or shortly called PCR-RFLP combines both the techniques of PCR, which has inherent ability to utilize small amount of relatively crude genomic DNA, and RFLP, which has the ability to discriminate between genotypes based upon the presence or absence of restriction sites within the amplified DNA (Karp et al. 1997). Genetic and breeding efforts have applied this technique in various plant groups such as Imperata cylindrica (Chou and Tsai 1999), Dipterocarpaceae (Tsumura et al. 1996), conifers (Tsumura et al. 1995), maize (Kaufman et al. 1996), rice (Cordesse et al. 1990), Brassica (Bhatia et al. 1996), and pepper (Livneh et al. 1992). The advantages of PCR-RFLP approach are in the quality of the data and the information obtained from it. The polymorphisms studied are of known identity and reveal information on phylogenetic relationships. This technique is also called cleaved amplified polymorphic sequence (CAPS) (Karp et al. 1997).
Chloroplast DNA (cpDNA) is a genome frequently used for population structure studies (Comes and Abbott 1999), and has been concentrated on relatively conservative genes such as rbcL (Doyle and Doyle 1999). This method has successfully analysed the population structure and phylogeographic studies of several species (Ziegenhagen and Fladung 1997). Beside rbcL, ndhC-trnV is also the type of cpDNA that has successfully been used in genetic studies of Coffea spp. (Lashermes et al. 1996). The objective of this study was to analyse phylogenetic relationships amongst 10 Durio species in Malaysia using PCR-RFLP on two cpDNA genes, ndhC-trnV and rbcL.

MATERIALS AND METHODS
The study was conducted at the Genetics and Plant Breeding Laboratory, Department of Crops Science, Faculty of Agriculture, Universiti Putra Malaysia from December 2002 to May 2003. Plant samples were 11 accessions from 10 Durio species collected from the Tenom Agriculture Research Station and Sabah Agriculture Park, Sabah, and University Agriculture Park, Universiti Putra Malaysia (Table 1).
DNAs were extracted from 300 mg of young leaves of durian accessions using modified CTAB-based protocol of Mathius and Hutabarat (1997). Leaf samples taken from Sabah were preserved in plastic bag containing 1% mercaptoethanol-enriched CTAB buffer (Santoso et al. 2003), while leaf samples from University Agriculture Park were used in fresh. Two pairs of oligonucleotides N 1 -N 2 and rbcL 1 -rbcL 2 primers (Table 2) obtained from Research Instrument were used to flank the target regions. Eight restriction enzymes (HindIII,BsuRI,PstI,TaqI,MspI,SmaI,BshNI,and EcoR130I) were used to digest the amplicons.
Amplification of ndhC-trnV and rbcL regions used the modified method of Edwards (1998). Amplicon was performed in a volume of 50 µl containing 27.2 µl distilled water, 1x PCR buffer, 2 mM MgCl 2 , 200 µM dNTPs, 0.6 mM of each primer, 60 ng genomic DNA, and 1 unit Taq polymerase. The PCR mixture was put into the 0.5 ml thin-walled microcentrifuge tube, placed in the thermal-cycler model PTC-200 (MJ Research), and amplified for 35 cycles of denaturation process at 95°C for 20 seconds, annealing process at 55°C for 60 seconds, extension process at 72°C for 60 seconds, and final extension at 72°C for 10 minutes.
First gel electrophoresis was performed in a volume of 12 µl containing 8 µl of distilled water, 2 µl of PCR products, and 2 µl of gel loading buffer. The samples were loaded onto a 2.0% agarose gel in 1x TAE along with DNA marker. The gel was run in 1x TAE buffer at 80 volts for 1 hour. After that, the gel was stained in the ethidium bromide solution for 15 minutes and then destained in distilled water for 30 minutes. The image of the stained gel was captured using the photo document system (model Chemilmager 5500-Alpha Innotech, USA).
Digestion of the PCR products was performed in a volume of 25 µl containing 15 µl distilled water, 2.5 µl 10x RE buffer, 5.0 µl amplicon, and 2.5 µl restriction enzyme. The reaction mixture was placed into the 0.5 µl thinwalled microcentrifuge tube. The mixture was mixed by pipetting up and down several times, and incubated at 37 o C for 1.5-3.0 hours. The second electrophoresis Variations of the restricted PCR products were scored based on the presence and absence of bands on the images taken from the gels. The NTSys-PC software (Rohlf 2000) was used to compute Jacquard's coefficients of similarity, and the dendrogram was then constructed using Unweighted Pair Group Method with Arithmetic Mean (UPGMA).

Phylogeny of Ten Durio Species based on PCR-RFLP of ndhC-trnV Gene
Primers N 1 -N 2 were able to amplify DNA sequences of 10 Durio species that ranged in size of approximately 900-1100 bp (Fig. 1). Following digestion with eight restriction enzymes, PCR-RFLP products of the ndhC-trnV gene produced restricted fragments with sizes ranging from 250 to 1100 bp (Fig. 2). Number of frag-ments produced from DNA of each accession ranged from 1 to 6; average number was 3.25. A total of 286 fragments were produced, of which 275 (96.15%) were polymorphic, while 11 (3.85%) were monomorphic. The monomorphic bands were observed by digestion using SmaI enzyme. Generally, the highest number of fragments was produced using SmaI (48 fragments), and the lowest was using TaqI (16 fragments). Genetic distance matrices constructed based on restriction products of all fragments analysed are presented in Table 3. These matrices show the genetic distances ranging from 2.2% to 96.3%. The widest was between D. sabahensis and D. testudinarum, having similarity value of 2.2%. Two accesions of D. graveolens (orange fruit and green fruit) are closely related, having similarity value of 96.3%. This means both D. graveolens are identic. The wide range in genetic distances (2.2% to 96.3%) shows that there were high variations amongst the 11 accessions of Durio species.
The dendrogram tree indicating the phylogenetic relationships amongst species based on RFLPs on   ndhC-trnV region is presented in Fig. 3. The dendrogram reveals the formation of five clusters. Three of those, the first, fourth and fifth clusters, were scattered off and occupied by single species of D. kinabaluensis, D. grandiflorus and D. testudinarum in each cluster, respectively. This reveals their low genetic similarities to other species, indicating that independent evolutionary processes involving ndhC-trnV region might have occurred on those three species. The second cluster consists of four accessions belonging to three species, i.e. D. oxleyanus, two accesions of D. graveolens (orange fruit and green fruit), and D. dulcis. This reveals high genetic similarities amongst the three species, indicating the presence of some associations in Table 3. Genetic distance matrices involving 10 Durio species, using Jacquard's coefficients of similarity, based on PCR-RFLP on ndhC-trnV region. the evolutionary processes involving ndhC-trnV region found in the three species within the cluster. A similar trend was observed in the third cluster, which consisted of four species, i.e. D. kotejensis, D. sabahensis, D. purureus and D. zibethinus. This cluster reveals high genetic similarities amongst the four species, indicating the presence of some associations in the evolutionary processes involving ndhC-trnV region in the four species.

Phylogeny of Ten Durio Species based on PCR-RFLP of rbcL Gene
Primers rbcL 1 -rbcL 2 were able to amplify targeted DNA sequences of 10 Durio species using and produced an aligning band approximately 1400 bp (Fig. 4). Digestion of the PCR product using eight restriction enzymes revealed restricted fragments sized between 150 and 1400 bp (Fig. 5). Number of fragments produced from DNA of each accession ranged from 1 to 4; average number was 2.13.
A total of 188 fragments were produced, of which, 34 (18.09%) were polymorphic and 154 (81.91%) were monomorphic. Generally, the highest number of fragments was produced using TaqI (41 fragments), and the lowest were using PstI, MspI and SmaI (each with 11 fragments).
Genetic distance matrices constructed based on restriction products of all fragments analysed are presented in Table 4. These matrices show that the genetic distances ranged from 76.1% to 100%. The widest distance was between D. sabahensis and D. testudinarum, having similarity value of 76.1%, while the closest distances were between the two accessions of D. graveolens, and amongst D. oxleyanus, D. dulcis and D purureus, having similarity value of 100%. This result also confirmed that two accession of D. graveolens are identic (Fig. 6).

Species
These results indicated the difference in the level of alteration in the DNA sequences of the two genes, where the sequences in the ndhC-trnV gene have altered more intensively than those in the rbcL gene. The ndhC-trnV region is a noncoding region of Cluster chloroplast DNA, which corresponds to intergenic spacer between ndhC and trnV genes (Lashermes et al. 1996). Noncoding region tends to rapidly in sequence alteration (Clegg and Zurawski 1992), makes it more reliable to be used in plant diversity analysis, and might be used to analyse the phylogenetic relationship in lower taxa. The evidence shown by the two accessions of D. graviolens, both have little mutation from the same length gene, differed from other species. Meanwhile, the rbcL region is a coding region of chloroplast DNA, which corresponds to RuBisCo large subunit (Petit et al. 1998). The gene is highly conserved among various plants and slow in sequence evolution, hence rbcL is a powerful marker to be used for phylogenetic analysis (Ziegenhagen and Fladung 1997), mainly at higher taxonomic level (Doyle and Doyle 1999). Although there was a different level of variations between the two genes studied, the technique involving PCR-RFLP on the ndhC-trnV and rbcL genes, have succeeded distinguish the 10 Durio species into distinct clusters. This technique can therefore be considered as useful markers for phylogenetic analysis of Durio species.
The results also showed that two cpDNA primers could be used in molecular analysis of Durio species, besides the intergenic spacer of rbcL-ORF106 gene (Kanzaki et al. 1997) and ndhF gene (Nyffeler and Baum 2000). This finding might be useful for further molecular marker assisted selection in Durio breeding program in both Malaysia and Indonesia, since these countries have the shared area of Durio species origin, and so the diversity. Discovering any other markers are, however, required to maximize recovery of Durio genome map.

CONCLUSION
Based on PCR-RFLP on the ndhC-trnV gene, the 10 Durio species were grouped into five distinct clusters, where the accessions generally showed high variations. In contrast, based on PCR-RFLP on the rbcL gene, the accessions were grouped into three distinct clusters, and generally showed low variations.
NdhC-trnV gene is more reliable for phylogeny analysis in lower taxonomic level of Durio or to be used for diversity analysis, while rbcL gene is reliable marker to be used for phylogeny analysis at higher taxonomic level.
The technique, PCR-RFLP on the ndhC-trnV and rbcL genes therefore, could be considered as useful markers for phylogeny analysis amongst Durio species. The finding might be useful for further molecular marker assisted selection in Durio breeding program.