VERIFICATION AND GENETIC DIVERSITY ANALYSIS OF INDONESIAN BANANA CULTIVARS CONTAINING

III. VERIFICATION AND GENETIC DIVERSITY ANALYSIS OF INDONESIAN BANANA CULTIVARS CONTAINING

B GENOME BY MICROSATELLITE ABSTRACT Genetic diversity of bananas in Indonesia is considered to be high. Phenotypic identification of banana accessions is sometimes incorrect due to instability of morphological characters. The aims of this study were to verify banana accessions containing the B genome previously identified based on morphological characters, to construct molecular determination key for the genomic groups, and to reveal genetic relationships among the accessions based on the microsatellite markers. The DNA of 92 banana accessions containing the B genome was analyzed using 8 microsatellite primers. The results showed that 9 accessions were incorrectly grouped into banana with the B genome. Those accessions should be placed into Musa acuminata due to the absence of the MaCIR108 alleles which longer than 270 bp. Four alleles of the MaCIR108 were considered as the diagnostic characters for the B genome, and 1 allele of the Ma- 3-90 could be used to differentiate AAB from ABB genomic groups. Alleles of the MaCIR108 and Ma-3-90 with the size of 295 bp and 152 bp were only found in BB and ABB and never been detected in AAB. The molecular determination key divided 83 accessions containing the B genome into 3 genomic groups; which are 8 accessions of BB, 33 accessions of AAB, and 42 accessions of ABB. Twenty four of the 83 accessions formed 8 identical genotypes, therefore this study detected only 67 genotypes. These accessions clustered according to their genomic groups. The ABB formed fewer clusters than the AAB genomic groups indicating that the contribution of the B genome is less diverse than that of the A genome. Key words: allele, genomic group, genetic relationships, molecular determination key INTRODUCTION For over 50 years, genomic constitution in banana has traditionally been determined based on morphological characters following the scoring system as described by Simmonds and Shepherd 1955. The modified version proposed by Silayoi and Chomchalow 1987 suggested that 15 morphological characters be used for identifying genomic group of AA; AAA; AAB; AB or AABB; ABB or ABBB; and BB or BBB. The phenotypic identification of banana accessions are sometimes incorrect due to instability of morphological characters Vicente et al. 2005, mainly in identifying accessions which have intermediate characteristics between Musa acuminata and M. balbisiana. Both species are designated as AA and BB genomes Stover and Simmonds 1987. Incorrect identification tended to increase among closely related cultivars Vicente et al. 2005. Jumari 2000 reported several banana accessions of AAB genomic group were detected as the ABB or vice versa in the study of Pudjoarinto et al. 1994 and Hadisunarso et al. 1995. Genomic groups have an important role in cultivated banana classification. However, genomic composition among banana accessions was hardly interpreted due to the occurrences of hybridization and polyploidy Pillay et al. 2004. Currently, several genomic groups were questionable due to the restricted variability on the morphological characters, so that it is necessary to verify the groups using high accurate investigations. Cultivar characteristics could be easily detected through nucleotide distinction of the DNA. Resolution of molecular markers is commonly higher than that of morphological appearance Ford-Lloyd et al . 1997. Vicente et al. 2005 suggested that DNA nucleotides are better than morphological markers in identifying cultivars. Microsatellites have been employed to investigate the genetic diversity of Musa species Creste et al. 2003, 2004. The markers were powerful for estimating an intraspecific genetic distance and effective for analyzing polyploidy Bruvo et al. 2004 especially for allopolipoidy. Microsatellites are chosen as marker in the genetics analysis since they produce high polymorphic alleles even in the closely related lines Semagn et al. 2006. Therefore, microsatellite markers were employed to characterize and identify cultivars of fruit crops, such as grape Arnold et al. 2002, strawberry Ashley et al. 2003, sweet orange Novelli et al. 2005, apple Galli et al. 2005 rasberry and blackberry Stafne et al. 2005, and mango Ukoskit 2007. The first study in M. acuminata cultivars Chapter II showed that microsatellites could be used for clarifying taxonomic status of banana containing the A genome alone. The absence of diagnostic characters of the B genome determined taxonomic status of the accessions. That microsatellite markers were able to be used for genomic fingerprint Kaemmer et al. 1997; Creste et al. 2003, and banana germplasms wealth in Indonesia are considered to be high Daniells et al. 2001 support the opportunity for constructing a determination key using molecular data. The goals of the study were to verify cultivars of banana containing the B genome previously identified based on morphological characters, to construct the molecular determination key for banana genomic groups and to investigate the genetic relationships among the accessions using microsatellite markers. MATERIALS AND METHODS Plant materials This work used genomic DNA from fresh leaves of 92 banana accessions previously morphologically identified and classified into BB; BB or BBB; AAB; and ABB genomic groups Jumari 2000; Jumari and Pudjoarinto 2000; Edison et al . 2004; INIBAP 2002a; Siddiqah 2002. The accessions originated from several regions in Indonesia Table 3.1. Table 3.1 List of 92 accessions used in study of banana containing the B genome Accession Subgroup a Genomic group Region Source b Batu Klutuk BB Medan Diperta Batu Krawang Klutuk BB Unknown Diperta Klutuk Awu - BB Uknown RIF Klutuk Batu - BB Uknown RIF Klutuk Hitam - BB Jasinga Bogor Bogor Klutuk Warangan Klutuk BB Kulon Progo Diperta Klutuk Watu Klutuk BB Uknown Diperta Klutuk Wulung Klutuk BB Cilongok Banyumas Diperta Boi - BB or BBB Genyem Jayapura RIF Angleng - AAB Jasinga Bogor Bogor Austroli Plantain AAB Dongkelan Yogyakarta Diperta Bangkahulu - AAB Unknown Diperta Beleum - AAB Jasinga Bogor Bogor Burlangge - AAB Sentani Jayapura RIF Haseum - AAB Jasinga Bogor Bogor Jambe Saat Triolin AAB Cipaku Bogor Diperta Kapas - AAB Jasinga Bogor Bogor Ketan - AAB East Java Diperta Ketip Gunungsari - AAB Cipaku Bogor Diperta Koja Susu - AAB Unknown Diperta Longong - AAB Cipaku Bogor Diperta Nangka - AAB Jasinga Bogor Bogor Nangka - AAB Unknown RIF Pisang Seribu Pisang Seribu AAB Unknown Diperta Pulut Pulut AAB Sukoharjo Diperta Raja - AAB Uknown RIF Table 3.1 List of 92 accessions used in study of banana containing the B genome continued Accession Subgroup a Genomic group Region Source b Raja Bagus Raja AAB Sleman Diperta Raja Bandung - AAB Bantul Diperta Raja Bulu - AAB Jasinga Bogor Bogor Raja Kasman Plantain AAB Kulon Progo Diperta Raja Lini Raja AAB Guning Kidul Diperta Raja Marto Plantain AAB Tembarak Temanggung Diperta Raja Nangka Plantain AAB Giwangan Yogyakarta Diperta Raja Puser Plantain AAB Gunung Kidul Diperta Raja Sabrang - AAB Kulon Progo Diperta Raja Sereh - AAB Jasinga Bogor Bogor Raja Sereh - AAB Uknown RIF Raja Sereh - AAB Purworejo Diperta Raja Talun - AAB Kotagede Yogyakarta Diperta Roid - AAB Cipaku Bogor Diperta Rojomolo Plantain AAB Malang Diperta Susu - AAB Jasinga Bogor Bogor Tanduk - AAB Jasinga Bogor Bogor Tanduk ex Kedondong - AAB Cipaku Bogor Diperta Tanduk Hijau Plantain AAB Surakarta Diperta Triolin Triolin AAB Bantul Diperta Abu Awak Awak ABB Unknown Diperta Apu - ABB Jasinga Bogor Bogor Awak Awak ABB Samarinda Diperta Awak Rawa Awak ABB Unknown Diperta Bawean Kepok ABB Banyumas Diperta Brentel Kepok ABB Tlekung Malang Diperta Byar - ABB Purworejo Diperta Comot Awak ABB Gunung Kidul Diperta Gablok Sobo ABB Unknown Diperta Gandul Sobo ABB Giri Banyuwangi Diperta Gedah Awak ABB Cipaku Bogor Diperta Kapas - ABB Tirto Pekalongan Diperta Kaso Awak ABB Cipaku Bogor Diperta Kates Kepok- ABB Giri Banyuwangi Diperta Kepok - ABB Jasinga Bogor Bogor Kepok Asam - ABB Bantul Diperta Kepok Awu - ABB Gunung Kidul Diperta Kepok Brot Kepok ABB Purworejo Diperta Kepok Byar Kepok ABB Unknown Diperta Kepok Gandul Kepok ABB Gunung Kidul Diperta Kepok Gajih Kepok ABB Unknown Diperta Kepok Klutuk - ABB Bantul Diperta Kepok Kuning Kepok ABB Dongkelan Yogyakarta Diperta Kepok Kuning - ABB Unknown RIF Kepok Kuningan Kepok ABB Tegalsari Girimulyo Diperta Kepok Ladrang Kepok ABB Unknown Diperta Kepok Lumut - ABB Bantul Diperta Kepok Ungu Sobo ABB Pendowoharjo Bantul Diperta Table 3.1 List of 92 accessions used in study of banana containing the B genome continued Accession Subgroup a Genomic group Region Source b Kepok Urang Kepok ABB Gunung Kidul Diperta Klutuk Susu Awak ABB Umbulharjo Yogyakarta Diperta Lempeneng Sobo ABB Cilongok Banyumas Diperta Raja Bali Awak ABB Bantul Diperta Raja Entog Awak ABB Purworejo Diperta Raja Kul Kepok ABB Malang Diperta Raja Pendopo Sobo ABB Purworejo Diperta Raja Siem - ABB Unknown RIF Raja Utri Awak ABB Gunung Kidul Diperta Raja Wesi Awak ABB Tawangsari Sukoharjo Diperta Selayar - ABB Sentani Jayapura RIF Siem - ABB Jasinga Bogor Bogor Sobo Kapuk Sobo ABB Giri Banyuwangi Diperta Sobo Kepok - ABB Unknown Diperta Sobo Kerik - ABB Tlekung Malang Diperta Sobo Londoijo Sobo ABB Unknown Diperta Sobo Londoputih Sobo ABB Sukoharjo Diperta Sobo Madura Sobo ABB Giri Banyuwangi Diperta a. following the subgroup determination key constructed by Jumari 2000 and Jumari and Pudjoarinto 2000 b1 Diperta, Dinas Pertanian dan Kehewanan, Yogyakarta b2 RIF, Research Institute of Fruits, Solok West Sumatera Microsatellite analysis Eight primers producing discrete and repeatable alleles in the first study Chapter II were used in this study Table 3.2. The MaCIR108 alleles were confirmed as fingerprint which could differentiate banana cultivar containing the B genome from banana with the A genome alone Kaemmers et al. 1997; Creste et al. 2003. A genomic DNA was extracted following Dixit 1998 procedure with modification. Leaves were crushed in liquid nitrogen and homogenized in the extraction buffer containing 100 mM Tris pH 8.0, 50 mM EDTA pH 8.0, 500 mM NaCl, 10 mM beta-mercaptoethanol, and 20 SDS. After purification using 10 mgml of RNAse free at 37ºC for 1 hour, precipitation of DNA was carried out without PEG solution treatment. DNA concentration was measured by spectrophotometer and then the DNA was diluted in TE to a suitable concentration. Microsatellites were amplified using PCR in a 10 µl reaction solution containing 15-20 ng of genomic DNA, 1 μl 10 x PCR buffer with 20 mM MgCl 2 , 0.2 μl 10 mM dNTPs, 0.2 μl 10 µM of each primer, and 0.06 μl 5 u μl -1 Taq DNA Polymerase Native GenScript Corporation®. The amplification was performed for 35 cycles consisting of denaturation at 95ºC for 30 s, annealing at 50, 53, 55 and 56ºC depending on the primers for 30 s, and extention at 72ºC for 30 s. These cycles were preceded by a 4 min denaturation at 95ºC, and ended by a 10 min final extention at 72ºC Kaemmer et al. 1997. PCR reactions were carried out in a Perkin Elmer 2400 thermocycler Applied Biosystems, Foster City, CA, USA®. The PCR products were then separated on a gel containing 6 wv polyacrylamide and 7 M Urea Sigma-Aldrich Chemie, Germany® at 45-60 W constant power for 2 to 3 hours. Then, the gel was stained using a modified silver staining method described by Creste et al. 2001. All solutions were prepared using distilled water. Table 3.2 Primers used for verification and genetic diversity analysis of banana containing the B genome Primers Forward primer sequence Reverse primer sequence, 5’-3’ Optimized annealing temperature ºC MaCIR108 a F: TAAAGGTGGGTTAGCATTAGG R: TTTGATGTCACAATGGTGTTCC 55 Ma-1-132 a F: GGAAAACGCGAATGTGTG R: AGCCATATACCGAGCACTTG 53 MaCIR327b a F: AAGTTAGTCAAGATAGTGGGATTT R: CTTTTGCACCAGTTGTTAGGG 50 MaCIR332a a F: TCCCAACCCCTGCAACCACT R: ATGACCTGTCGAACATCCTTT 53 Ma-1-17 a F: AGGCGGGGAATCGGTAGA R: GGCGGGAGACAGATGGAGT 56 Ma-1-27 a F: TGAATCCCAAGTTTGGTCAAG R: CAAAACACTGTCCCCATCTC 56 Ma-3-90 b F: GCACGAAGAGGCATCAC R: GGCCAAATTTGATGGACT 56 Ma-3-139 b F: ACTGCTGCTCTCCACCTCAAC R: GTCCCCCAAGAACCATATGATT 56 a Primers designed by Kaemmer et al. 1997 b Primers designed by Crouch et al. 1998 Data analysis Banana accessions containing the B genome were distinguished from the A genome alone based on the fragment size of MaCIR108 amplified products. Each variant fragment of microsatellite was assumed as an allele. For each allele was scored as present 1 or absent 0. The allele size was estimated using 100-bp DNA ladder Invitrogen®. Designation of allele submitted to the relative size of allele followed CYMMYT 2004. The loci are named alphabetically as a for the largest fragment to h for the smallest fragment. Any band below the fragment is named according to the letter designation followed by number, for instance a 1 for the largest of a allele, etc. A similarity matrix was calculated based on Jaccard similarity index using Similarity of Qualitative Data SYMQUAL procedures. The matrix was then used for clustering analysis to examine genetic relationships among the accessions. Clustering analysis was performed using Sequential, Agglomerative, Hierarchical and Nested SAHN of Unweighted Pair-grouping Method with Arithmatic Average UPGMA by the Numerical Taxonomy and Multivariate Analysis System NTSYSpc version 2.02 Rohlf 1998. RESULTS AND DISCUSSION Verification of banana accessions containing the B genome All primers produced well-defined alleles exhibiting a polymorphic banding pattern. Total 75 alleles were obtained in this study, with the range of allele size from 110 to 436 bp Table 3.3. Nine of the 92 accessions examined did not have the MaCIR108 alleles with the size of more than 270 bp that were confirmed as the diagnostic characters for banana containing the B genome Chapter II. It means that those 9 accessions had the A genome alone. The absence of the MaCIR108 alleles which larger than 270 bp indicated that they should be classified into M. acuminata as reported in the first study Chapter II. Banana genomic groups were incorrectly identified perhaps due to the morphological expressions which were overlap between M. acuminata and M. balbisiana Jumari 2000. However in this study, incorrect classification of those 9 accessions was considered mostly due to mislabeled Table 3.4. Table 3.3 Microsatellite alleles revealed from analysis of banana containing the B genome using the 8 microsatellite primers Primer Alleles size range bp Number of alleles Alleles of each primer MaCIR327b 388-436 5 a 1 ,a 2 ,a 3 ,a 4 ,a 5 Ma-1-132 330-378 8 b 1 ,b 2 ,b 3 ,b 4 ,b 5 ,b 6 ,b 7 ,b 8 MaCIR332a 260-296 10 c 1 ,c 2 ,c 3 ,c 4 ,c 5 ,c 6 ,c 7 ,c 8 ,c 9 ,c 10 MaCIR108 220-295 13 d 1 ,d 2 ,d 3 ,d 4 ,d 5 ,d 6 ,d 7 ,d 8 ,d 9 ,d 10 ,d 11 ,d 12 ,d 13 Ma-3-139 132-177 11 e 1 ,e 2 ,e 3 ,e 4 ,e 5 ,e 6 ,e 7 ,e 8 ,e 9 ,e 10 ,e 11 Ma-3-90 132-172 12 f 1 ,f 2 ,f 3 ,f 4 ,f 5 ,f 6 ,f 7 ,f 8 ,f 9 ,f 10 ,f 11 ,f 12 Ma-1-27 122-142 6 g 1 ,g 2 ,g 3 ,g 4 ,g 5 ,g 6 Ma-1-17 110-154 10 h 1 ,h 2 ,h 3 ,h 4 ,h 5 ,h 6, h 7, h 8, h 9, h 10 Total 75 ‘Bawean’, ‘Kepok Klutuk’, ‘Kepok Ungu’, and ‘Sobo Madura’ are the specific names for banana containing the B genome. The inaccurate identification was human errors that incorrectly labeled the accessions in the field. Mislabeled was common phenomenon in the ex situ germplasm collections because morphological appearances were sometimes difficult to be interpreted due to the influence of environmental and cultivation factors Hernandez et al. 2001. The occurrence of mislabeled in olive, barley and banana germplasm collections were reported by Lopes and Mendonca 2004, Hamza et al. 2004, and Karamura et al. 2000 respectively. This misidentification could have great influence on efficiency of the breeding improvement Lopes and Mendonga 2004. Table 3.4 List of the 9 banana accessions incorrectly classified into banana containing the B genome Accession Genomic group Based on morphology Based on microsatellites Angleng AAB AAA Bawean ABB AAA Beleum AAB AAA Kapas ABB AAA Kepok Klutuk ABB AAA Kepok Ungu ABB AAA Ketip Gunungsari AAB AAA Roid AAB AAA Sobo Madura ABB AAA Diagnostic characters and molecular key for determination of banana genomic groups Four alleles of the MaCIR108 with the size of more than 270 bp d 1 to d 4 ; alleles of the MaCIR108 with the size of 270 bp or less d 5 to d 13 ; and 1 allele of the Ma-3-90 with the size of 152 bp f 7 were diagnostic characters for banana genomic groups Table 3.5. The nine accessions: 4 accessions of AAB and 5 accessions of ABB only had 3 alleles of the MaCIR108 with the size of 270 bp or less, thus they should be classified into the triploid AAA. The remaining 83 accessions were bananas containing the B genome because 1 or 2 alleles of the MaCIR108 with the size of more than 270 bp were detected in those accessions. Table 3.5 Classification of banana accessions into genomic groups based on the primers MaCIR108 and Ma-3-90 Alleles MaCIR108 Ma-3-90 d 1 d 2 d 3 d 4 d 5 -d 13 f 7 Genomic group based on morphology 295 bp 289 bp 287 bp 275 bp ≤270 bp 152 bp Genomic group based on microsatellites 8 BB + - - - - + 8 BB 1 BB or BBB + - + - + + 1 ABB 37 AAB - - - - +++ - 4 AAA - - + - +++ - 23 AAB - - - + +++ - 7 AAB + - - - + + 2 ABB + + - - + + 1 ABB 46 ABB - - - - +++ - 5 AAA - - + - ++ - 3 AAB + - + - + + 5 ABB - - + - + + 4 ABB + - - - + + 16 ABB + + - - + + 13 ABB + = allele present - = allele absent ++++++ = 1, 2, or 3 alleles Allele d 1 , d 2 , d 3 , or d 4 of the MaCIR108 and allele f 7 of the Ma-3-90 were consistently presence or absence in certain genomic group. Thirty accessions of AAB and 3 accessions of ABB did not have allele f 7 of the Ma-3-90. This allele was detected only in the other 50 accessions. Allele d 1 of the MaCIR108 was observed in all BB accessions. Allele d 1 , d 2 , or d 3 were detected in 41 accessions of ABB. Twenty three of 33 AAB accessions had allele d 3 ; 2 accessions had allele d 1 ; 1 accession had allele d 1 and d 2 ; and the remaining 7 accessions had allele d 4 . The size estimation of allele d 1 ; d 2 ; d 3 ; d 4; and f 7 were 295 bp; 289 bp; 287 bp; 275 bp; and 152 bp, respectively Based on primer MaCIR108, only 1 accession of BB or BBB studied was known to have heterozygous genotype consisting of 3 different alleles. The two alleles fell into the range of allele length of the B genome and 1 allele fell into the A genome. Therefore based on microsatellite marker, this accession should be classified into ABB genomic group. Allele f 7 of the Ma-3-90 was detected in all BB 8 accessions, in 93 38 accessions of ABB, only in 9 3 accessions of AAB and none was observed in the 9 accessions of AAA. Table 3.5 showed the 3 accessions of AAB were classified into ABB genomic group due to existence of 1 allele of the Ma-3-90 with the size of 152 bp and 1 allele of the MaCIR108 with the size of 270 bp or less. Those accessions were ‘Ketan’ and ‘Raja Bandung’ from Diperta collection, and ‘Susu’ from Bogor. One allele of the MaCIR108 with the size of more than 270 bp was also detected in these accessions. On the other hands, the 3 accessions of ABB ‘Byar’, ‘Brentel’, and ‘Kates’ from Diperta collection were included into AAB genomic group because allele of the Ma-3-90 152 bp was absent. In these accessions also found 2 alleles of the MaCIR108 with the size of 270 bp or less and 1 allele with the size of more than 270 bp. This study exhibited that all alleles of the MaCIR108 could be considered as diagnostic characters for banana genomic groups and only one allele of the Ma-3- 90 which could be used to differentiate AAB from ABB, especially for accessions having allele of the MaCIR108 with the size of 287 bp and 1 allele of the MaCIR108 for the A genome. Four alleles of the MaCIR108 were thought as diagnostic character for the B genome, and the remaining 9 alleles for the A genome. Based on those alleles, a molecular determination key for genomic groups was constructed as the key created by Guzow-Krzeminska et al. 2001. Molecular key for determination of banana genomic groups 1a. The MaCIR108 alleles with the size of 270 bp or less……………..…………2 2a. The number of alleles, one to two ..…………AA or AAA genomic group 2b. The number of alleles, three..…….………………....AAA genomic group 1b. The MaCIR108 alleles with the size of more than 270 bp……………………3 3a. Allele with the size of 295 bp, only…………………...BB genomic group 3b. Alleles combination: 295 bp; 295 bp and 289 bp; 295 bp and 287 bp; and 1 allele of the MaCIR108 with the size of 270 bp or less ……...……………………………………………….ABB genomic group 3c. Alleles combination of 287 bp and alleles of the MaCIR108 with the size of 270 bp or less.………………………………………………………….4 4a. Alleles combination of 287 bp and 2 alleles of 270 bp or less …………………….……………………………AAB genomic group 4b. Alleles combination of 287 bp and 1 allele of 270 bp or less ……….5 5a. The Ma-3-90 allele of 152 bp, present .…….ABB genomic group 5b. The Ma-3-90 allele of 152 bp, absent .……..AAB genomic group 3d. Allele combination of 275 bp and alleles of the MaCIR108 with the size of 270 bp or less ............................................................AAB genomic group A fascinating characteristic of the B genome markers was that the MaCIR108 allele with the size of 295 bp and the Ma-3-90 allele with the size of 152 bp were always present in accessions with two B genomes BB and ABB, and never detected in banana with one B genome AAB. These results confirm that genomic groups of banana accessions particularly derived from hybrid between M. acuminata and M. balbisiana could be easily and rapidly determined by microsatellite marker. The result of determination of genomic group is summarized in Table 3.6. The keys demonstrated that the genomic group determination based on morphological characters is always not accurate. Genomic group of several hybrids and accession “BB or BBB” studied have been incorrectly identified. Actually, the existence of pure natural BBB accession remains questionable Pillay et al. 2004. Based on microsatellite markers, all accessions of BB were correctly classified into its genomic group, but only 81 accessions of AAB and 83 accessions of ABB which have correctly been classified into their genomic groups. Table 3.6 Number of accessions classified into the genomic groups based on the primers MaCIR108 and Ma-3-90 Genomic group based on microsatellites Genomic group based on morphology Number of accessions AAA BB BBB AAB ABB BB 8 8 BB or BBB 1 1 AAB 37 4 0 30 3 ABB 46 5 0 3 38 Total 92 9 8 33 42 The key in the present study could be applied only for banana cultivars derived from section Eumusa Stover and Simmonds 1987. This section is the most geographically widespread, with species ranging from India to Pacific INIBAP 2004. It contains at least 11 species The Australian Government Office of the Gene Technology Regulator 2008, but only 2 wild species, M. acuminata and M. balbisiana, involving in development of most edible bananas Cheesman 1948; Simmonds and Shepherd 1955; Sharrock 1998. A number of distinct cultivars of major group are derived from M. acuminata and M. balbisiana either alone or in various hybrid combinations. Molecular as well as morphological determination key also may have limited ability in their application. The key in the study could not be used for identifying bananas cultivars derived from the other section, such as sect. Callimusa . According to Wong et al. 2002, Callimusa and Eumusa were genetically distinct because the basic number of chromosome of n = x = 10- and n = x = =11- grouping based on AFLP analysis are robust and justified to separate Musa species into different section. The similarity between conserved regions of the 2 organisms from different taxonomical groups would be less significant Guzow-Krzeminska et al. 2001. Therefore, different primers should be designed for constructing determination keys of different taxonomical groups. The MaCIR108 alleles were a fingerprinting for distinguishing banana cultivars containing the B genome from those containing the A genome alone. The sequence of MaCIR108 flanking region in both genomes has been known precisely Kaemmer et al. 1997. The Ma-3-90 microsatellite allele could be only used to differentiate AAB from ABB genomic groups. Besides the primers MaCIR108 and Ma-3-90 , Creste et al. 2005 also had successfully designed the other of microsatellite loci which were also confirmed as fingerprint for genomic groups, such as MaOCEN03, MaOCE12, MaOCE13 and MaOCE14. In order to avoid incorrect identification, it is possible for identifying banana cultivars using molecular marker before they were commercially cultivated, particularly for the cultivars which were not easily classified using morphological characters. Molecular determination key has the potentially applied in identifying a small fragment of organisms or organisms which is difficult to be identified using traditionally keys Guzow-Krzeminska et al. 2001. The molecular determination key also provides suitable tools for researchers who are familiar with molecular techniques. Analysis of relationships among banana accessions containing the B genome Microsatellite has provided important information concerning the genetic relationships among accessions of banana containing the B genome. Clustering analysis based on the 8 primers Figure 3.1 detected only 67 genotypes. It is because 24 accessions clustered within 8 identical genotypes. At the coefficient similarity of 0.22, the dendrogram showed 2 main clusters. In the first main cluster I, the subcluster Ia, 8 accessions of BB were grouped together with 42 accessions of ABB and separated from 15 accessions of AAB. In the subcluster Ib it was observed 11 accessions of AAB. The second main cluster II consisted of 7 accessions of AAB. The accessions of AAB in this cluster separated from those in the first main cluster due to the presence of allele MaCIR108 with the size of 275 bp. The results suggested that the AAB genomic group in the main cluster II was showed to be the most distance from the other genomic groups. According to Jumari and Pudjoarinto 2000, based on morphological characters, 2 of the 7 accessions in the main cluster II namely ‘Triolin’ and ‘Jambe Saat’ belong to the AB genomic group. The accession characteristics tended to indicate the transition between cultivar groups of AA, AAA, and AAB; and groups of BB, ABB, and ABBB. Naturally, the banana accessions of the AB genomic groups are rare Valmayor et al. 2000 so that the existence of hybrid AB especially in Indonesia still debatable. Figure 3.1 Dendrogram from analysis of the 83 banana containing the B genome based on the 8 microsatellite primers. ● genomic group based on morphological characters, ●● genomic group based on the primers MaCIR108 and Ma-3-90. 1.00 ● ●● Accession Coefficient 0.22 0.41 0.61 0.80 1.00 5 s51 17w 98 2w s52 32w 160 9 10 99 180 s55 s104 45w 61 45 47w 100 41w 1w 4w 51 7w 81 55 88 30 153 64 79 82 25w 84 96 6 31 41 52 56 27 101 43w 161 31w 158 s60 s89 46 87 7 s106 147 49 23w 24w 21w 154 103 39w 95 155 s68 26 69 22 97 149 78 s33 25 74 24 23 29w 33 8 157 11 77 152 s62 83 ABB ABB Kepok Kuning ABB ABB Kepok Kuning ABB ABB Kepok Urang ABB ABB Kepok Lumut ABB ABB Raja Wesi BBB ABB Boi ABB ABB Raja Kul ABB ABB Kepok BB BB Klutuk Watu BB BB Klutuk Wulung BB BB Batu BB BB Klutuk Hitam BB BB Klutuk Batu BB BB Klutuk Awu BB BB Batu Krawang BB BB Klutuk Warangan ABB ABB Kepok Asam ABB ABB Gablok ABB ABB Kepok Awu ABB ABB Gandul ABB ABB Lempeneng ABB ABB Sobo Kepok ABB ABB Sobo Kerik ABB ABB Kepok Ladrang AAB ABB Ketan ABB ABB Kepok Gandul ABB ABB Sobo Londoijo ABB ABB Raja Pendopo AAB ABB Susu ABB ABB Kepok Gajih ABB ABB Sobo Kapuk ABB ABB Kepok Brot ABB ABB Kepok Byar ABB ABB Sobo Londoputih ABB ABB Kepok Kuningan ABB ABB Awak Rawa ABB ABB Awak ABB ABB Raja Entog ABB ABB Gedah ABB ABB Abu Awak ABB ABB Klutuk Susu AAB ABB Raja Bandung ABB ABB Raja Utri ABB ABB Apu ABB ABB Kaso ABB ABB Siem ABB ABB Raja Siem ABB ABB Selayar ABB ABB Raja Bali ABB ABB Comot AAB AAB Raja Nangka AAB AAB Nangka AAB AAB Nangka AAB AAB Raja Puser AAB AAB Raja Marto AAB AAB Raja Kasman AAB AAB Austoli AAB AAB Kapas AAB AAB Tanduk Hijau ABB AAB Byar AAB AAB Tanduk Ex Kedondong AAB AAB Tanduk AAB AAB Burlangge ABB AAB Brentel ABB AAB Kates AAB AAB Rojomolo AAB AAB Raja Bagus AAB AAB Raja Bulu AAB AAB Raja Sereh AAB AAB Raja AAB AAB Raja Sabrang AAB AAB Raja Talun AAB AAB Raja Lini AAB AAB Longong AAB AAB Pulut AAB AAB Pisang Seribu AAB AAB Jambe Saat AAB AAB Haseum AAB AAB Triolin AAB AAB Koja Susu AAB AAB Raja Sereh AAB AAB Raja Sereh AAB AAB Bangkahulu Ib Ia I II The microsatellite study demonstrated that 1 allele of the B genome and 2 alleles of the A genome were detected in the accessions ‘Triolin’ and ‘Jambe Saat’. Therefore, based on microsatellite marker they should be placed into the AAB genomic group. In this case, variations of morphological characters were not sufficient to precisely determine genomic group. The AAB accessions in the main cluster I were separated from BB and ABB. The BB accessions were genetically closely related to ABB due to the existence of the MaCIR108 allele with the size of 295 bp and allele of the Ma-3- 90 with the size of 152 bp in both genomic groups. It supported the results of previous study using the AFLP technique Ude et al. 2002. Most AAB accessions were grouped in the main cluster I due to the existence of the MaCIR108 allele with the size of 287 bp. The differences among these AAB accessions were determined by alleles of the other primers. The accessions of ABB, ‘Kates’ and ‘Brentel’ in the subcluster Ia based on the primers MaCIR108 and Ma-3-90 were classified into AAB genomic group. They were genetically distinct from the others AAB. Both ‘Kates’ and ‘Brentel’ had 1 allele unique in each locus MaCIR332a and Ma-3-90 which were not be found in the others AAB. This result supported Valmayor et al. 2000 which reported that Kates’ was recognized as a unique cultivar originating from Indonesia. Morphologically, ‘Kates’ produced large - solitary fruits per hand and the fingers that were similar to small papaya. Based on the 8 markers, 7 of the 8 accessions of BB were known to be identical. Seedless and parthenocarpy were considered as characteristics of the cultivated bananas Stover and Simmods 1987; Krishnamoorthy et al. 2004. Within the diploid of BB, the occurrence of parthenocarpy was not detected Heslop-Harisson and Schwarzacher 2007, therefore it has allowed the production of pulp with full seeds. Mutations within the cultivated bananas correspond with human intervention Carreel et al. 2002. Besides the BB genomic group which is less variable than diploid the A genome Karamura 1998, the selection pressure within BB genomic group was also smaller than that the other genomic groups due to the presence of seeds in their fruits. There were no different genotype of microsatellites among BB genomic groups; they were genetically similar. However, ‘Klutuk Warangan’ was the exception. Based on their phenotypic features, the cultivars of BB were usually not morphologically distinct to the wild type of M. balbisiana. Genetic relationship analysis among the 83 banana accessions containing the B genome showed that the accessions clustered according to their genomic groups. Similarity coefficient at 0.61 producing 11 subclusters of AAB, 4 subclusters of ABB, and 1 subcluster of BB indicated that the diversity contribution of the B genome is less than that of the A genome. This result supported cytometry analysis which revealed that the B genome was less diverse than the A genome Lysak et al. 1999; Kamate et al. 2001; Dolezel et al. 2004. The ABB accessions appear closer to the BB than AAB genomic groups. It suggested that AAB genomic group was more divergence from the original ancestors than ABB genomic group Pillay et al. 2004. Molecular markers were needed to identify the presence of multiple genotypes with a single name or a single genotype with multiple names Vicente et al . 2005; Semagn et al. 2006. This information is important to provide accurate identification for conservation activities. In the clonally propagated crops such as banana, the conservation activities demand a complex and expensive procedures. Molecular data could be used for verifying the collections efficiently and ensured the minimum repetition. The number of accessions could be reduced significantly without reducing genetic variations within germplasm collections Vicente et al. 2005. In general, allelic composition of microsatellites was relatively stable Kaemmer et al. 1997. It is because the genotypes arising from cross breeding that occurred thousands years ago were maintained through vegetative propagation Carreel et al. 2002. Actually, cultivated bananas were rather complex hybrids because they have been modified by various mutations and exploited by human for food for a long time Stover and Simmonds 1987. CONCLUSION The MaCIR108 alleles are convincingly confirmed as diagnostic characters for distinguishing banana containing the B genome from banana containing the A genome alone. Nine accessions were incorrectly classified into banana with the B genome. Those accessions should be placed into M. acuminata due to the absence of the MaCIR108 alleles which longer than 270 bp. Four of 13 alleles of the MaCIR108 with the size of more than 270 bp as the diagnostic characters for the B genome, whereas the remaining 9 alleles of the MaCIR108 with the size of 270 bp or less are considered as the diagnostic characters of the A genome. One allele of the Ma-3-90 with the size of 152 bp can be used for separating AAB from ABB genomic groups. Based on molecular determination key, the 83 accessions containing the B genome were divided into 3 genomic groups, which are 8 accessions of BB, 33 accessions of AAB, and 42 accessions of ABB. Analysis of these accessions using the 8 primers detected only 67 genotypes due to 24 accessions formed 8 identical genotypes. The genetic relationship analysis showed that banana accessions clustered according to their genomic groups. The ABB formed fewer clusters than the AAB genomic groups indicating that the contribution of the B genome is less diverse than that of the A genome. Similarity coefficient at 0.61 produced 11 subclusters of AAB, 4 subclusters of ABB, and 1 subcluster of BB.

IV. MICROSATELLITE MARKERS FOR CLASSIFIYING AND ANALYZING GENETIC RELATIONSHIP OF