5. Analysis of induced mutants of Philippine bananas with molecular markers - Hautea, D.M., G.C. Molina, C.H. Balatero, N.B. Coronado, E.B. Perez, M.T.H. Alvarez, A.O. Canama, R.H. Akuba, R.B. Quilloy, R.B. Frankie, C.S. Caspillo

Institute of Plant Breeding
College of Agriculture
University of the Philippines Los Banos
4031 College
Laguna,
The Philippines
E-mail: [email protected]; [email protected]

Abstract

Mutants were induced in the two most popular Philippine banana cultivars by gamma and fast neutron irradiation of in vitro shoot-tip cultures. Most of the clones produced were morphologically similar to the non-irradiated cultivars. Promising clones were selected and evaluated further using molecular markers. The RAPD, SSR and AFLP techniques were successfully established under local conditions and found useful for characterizing banana and other Musa species. RAPD and AFLP markers showed sufficient polymorphism for genotype discrimination, but SSR and AFLP markers are more highly reproducible. Only AFLP markers possess both high reproducibility and discriminatory capability. AFLP can distinguish between the two cultivars used with a minimum of one selective primer pair. It can also detect variation in DNA profiles of induced mutant clones which are otherwise morphologically indistinguishable, and detect variation between the induced mutant parent clones and their derived suckers. A non-radioactive silver staining technique for AFLP markers suitable for local laboratory conditions has been established. The results obtained from both morphological and molecular analyses show that more point mutations were generated by irradiation with fast neutrons than with gamma rays. However, the number of vegetative propagation cycles for the shoot-tip technique used may not be sufficient to eliminate chimeras completely in the mutated populations. The results obtained could provide a sound basis for the successful application of mutation and molecular marker techniques to improve bananas in the Philippines.

1. INTRODUCTION

The banana is one of the most important fruit crops of the Philippines. It is important both as a source of local and international revenues for farmers, and an important component of the daily diet of all Filipinos. Two Musa acuminata cultivars, 'Lakatan' (AAA) and 'Latundan' (AAB), are among the most popular dessert type of bananas produced primarily by small-scale banana farmers and consumed by the local market. Their production and trade potentials are severely constrained by low yield, susceptibility to diseases and undesirable post-harvest qualities. Breeding strategies to improve these traits are difficult to implement because of the asexual behaviour, sterility and polyploidy of edible bananas [1].

Induced mutation (either by chemicals or irradiation), coupled with in vitro propagation technique such as shoot-tip culture, has been established as a tool to generate variation in a number of vegetatively propagated crops [2-4]. Induced mutants are usually differentiated from each other and from the original cultivar by phenotypic analysis based on agromorphological and physiological traits. This approach could be severely limited by the large size of mutagenized populations, particularly in the case of banana, and by dominance and developmental and environmental effects in the traits used in characterizing the population. Chimeras obtained after irradiation are usually difficult to detect. However, advances in molecular marker analysis may overcome these limitations.

In banana, several DNA marker techniques have been used to investigate genetic relationships between Musa accessions, and to determine differences in somaclonal variants and radiation-induced mutants. These are restriction fragment length polymorphism (RFLP) [5-7], random amplified polymorphic DNA (RAPD) [8, 9], microsatellite or simple sequence repeats (SSR) [10-13], and amplified fragment length polymorphism (AFLP) [14]. This report summarizes the results of our efforts to generate useful induced mutants of the Philippine banana cultivars, 'Latundan' and 'Lakatan' through irradiation, and to evaluate the usefulness of DNA marker techniques, such as RAPD, microsatellites or SSR, and AFLP, to characterize the genomic alterations in induced mutants of the two Philippine banana cultivars.

2. GENERATION OF INDUCED MUTANTS OF PHILIPPINE BANANA CULTIVARS BY GAMMA AND FAST NEUTRON IRRADIATION

Induced mutation (either by chemicals or irradiation) has been established as a tool to generate variation in a number of seed- and vegetatively propagated crops [2-4]. The project successfully adapted established protocols for mutation induction and subsequent propagation of in vitro cultured shoot tips [15] in two Philippine banana cultivars.

2.1. In vitro culture, radiosensitivity and mutation induction

In vitro cultures of the cultivar 'Lakatan' were provided by the IPB in vitro germplasm collection for banana in 1995. Cultures were maintained on modified MS medium [16] supplemented with 20 mM BAP, 40 mg/l cysteine hydrochloride and 40 g/l sucrose at pH 5.6-5.8. Cultures were placed on a horizontal shaker (60 r.p.m.) and continuous proliferation was maintained for 30 days sub-culture intervals in a controlled environment at 28ºC and 16 h illumination. Meristem tips containing a meristematic dome of two to three pairs of leaf primordia excised from in vitro growing shoots were used for irradiation experiment.

Longitudinally dissected explants were transferred to sterile Petri dishes with a few drops of sterile distilled water and sealed with paraffin. The radiosensitivity assay was conducted using split shoots of the Lakatan irradiated with 10, 20, 30, 40, 50 or 60 Gy gamma rays from a ⁶⁰Co source in the FAO/IAEA laboratory in Seibersdorf, Austria. There were 12 shoots per treatment. After treatment, split shoot tips were immediately transferred to fresh medium (modified MS, described above) and multiplied to generation M₁V₄ in proliferating MS medium. Radiosensitivity for gamma radiation of Lakatan was established at 40 Gy. Other workers in this IAEA CRP on banana have obtained similar values (35-45 Gy) using their preferred cultivars such as Embul in Sri Lanka, Pisang Mas in Malaysia, and Paracido al Rey, Gran Enano and Burro CEMSA in Cuba [17].

After the radiosensitivity test, bulk irradiation of 500 meristems of Lakatan with 40 Gy gamma rays (LK-40) was carried out. Additional irradiation with fast neutron at a dose of 3 Gy was also carried out in 500 and 200 meristems of Lakatan (LK-3) and Latundan (LT-3), respectively. The shoots were sub-cultured to the M₁V₅ as described above. Roots were induced on half-strength MS medium supplemented with 40 mg/l cysteine hydrochloride and 20 g/l sucrose solidified with 7 g/l agar. Plantlets were transferred onto coir dust-soil mixture (1:1) in the greenhouse. After six months, the acclimatized plants were transferred directly to the field for further selection and characterization.

The results of mutation induction in the two Philippine cultivars used are summarized in Table 1. Irradiation with gamma rays or fast neutrons induced the production of agromorphological variant phenotypes in Lakatan and Latundan cultivars, as described in the following section, and produced clones which gave rise to suckers that remain negative (resistant) to the banana bunchy top virus (BBTV) test. Nevertheless, large number of clones was lost due to BBTV infection, particularly in the mutated populations of the cultivar Lakatan.

Table 1 Summary of mutation induction in the two Philippine cultivars, "Lakatan" (LK) and "Latundan" (LT) using 40 Gy gamma rays (LK-40) or 3 Gy fast neutrons (LK-3, LT-3)

2.2. Morphological characteristics of mutated populations

Morphological characterization of individual clones in each mutated populations (LK-40, LK-3 and LT-3) was carried out using the IPGRI descriptor for banana [18]. Each clone was characterized using characters taken at the vegetative, reproductive and fruiting stages.

Figure 1 Principal ordination of mutated populations for 27 vegetative characters. Mutant clones of (clockwise from top left) Lakatan irradiated with 40 Gy gamma rays show two main clusters and a few distant clones; Lakatan irradiated with 3 Gy fast neutrons shows one main cluster and a number of more distant clones; Latundan clones irradiated with 3 Gy fast neutrons also formed one main cluster with a larger number of more distant clones.

Morphological variants were observed for almost all of the traits measured in the three irradiated populations. These include dwarf habit and large fruit size. In general, only a small proportion of variant phenotypes was observed for each trait. Principal component analysis of 27 vegetative traits was conducted and the results are presented in Figure 1. Most of the clones (depicted as large clusters) in each mutated population were morphologically similar to each other. These clones exhibited traits that are typical for non-irradiated Latundan and Lakatan. More variants were observed in populations irradiated with fast neutrons (LK-3 and LT-3) than with 40 Gy gamma rays. These results support previous observations that fast neutrons result in a higher frequency of induced mutations than gamma rays. Fast neutrons, a particle radiation, are more energetic than gamma rays, a form of electromagnetic radiation [4].

3. ESTABLISHMENT OF PROTOCOLS AND IDENTIFICATION OF MOLECULAR MARKERS FOR GENOME ANALYSIS OF INDUCED MUTATIONS OF BANANA

During the past two decades, various DNA marker techniques have been developed. These techniques can be classified primarily as restriction enzyme-based, polymerase chain reaction (PCR)-based, or a combination of both. Of these techniques, RFLP, RAPD, SSR and AFLP have been used in various aspects of genome analysis of Musa species. This section will present the results obtained by the project in developing protocols for RAPD, SSR and AFLP analysis of induced mutant populations of two Philippine banana cultivars.

3.1. DNA extraction

The micro extraction procedure of Fulton et al. [19] was adopted for all molecular marker analyses described in the following sections because it is fast, simple and yields relatively clean DNA in sufficient quantity for PCR amplification. This procedure was used to extract DNA from young leaf spindles of normal and irradiated banana clones.

3.2. Random amplified polymorphic DNA (RAPD)

Random amplified polymorphic DNA (RAPD) is a novel technique developed by Williams et al. [20] and Welsh and McCleland [21] based on the amplification of random DNA segments with single primers of arbitrary nucleotide sequence. The technique allows detection of extensive polymorphism because of the unlimited number of arbitrary primers. The assay is fast, non-radioactive, and requires only nanogram quantities of DNA. The literature is full of reports on the successful applications of RAPD analysis for various genome studies, including genotype identification, in a broad range of species.

The results obtained in this project indicate that the RAPD technique can be readily adapted for banana. A RAPD protocol was adapted from the procedure of Williams et al. [20] with the following modifications in the concentrations of MgCl₂ (1.5-2.0 mM), RAPD primer (0.15-0.18 µM), and template DNA (3-5 ng). RAPD primers were purchased from Operon Technology, USA. Amplification was performed using an MJ Research PTC-100 thermal cycler with a temperature cycle profile consisting of the following: 93ºC (1 min), 45 cycles of 93ºC (30 sec), 36ºC (1 min), 72ºC (1.5 min), and final extension at 72ºC (2 min). Amplification products were separated in 2% agarose gel in 1 × TAE buffer and stained with ethidium bromide. Figure 2 illustrates the RAPD patterns obtained following the procedure described above.

Preliminary RAPD analysis was conducted in four banana clones (one normal clone of Lakatan, and two irradiated with 3 Gy fast neutrons and 40 Gy gamma rays; and one clone of Latundan irradiated with 3 Gy fast neutrons). Twelve random primers (OPQ-01 to -03, OPR-01 to -03, OPS-01 to -03 and OPT-01 to -03) were tested. All of the 12 primers tested generated well resolved amplification products in all the four clones tested. Four primers revealed variation among the clones. These primers are OPR-01, OPR-02, OPR-03 and OPT-01. Four RAPD markers (OPR-01₃₈₃₂, OPR-03₇₄₁ and OPT-01₂₆₀₀) were detected only in the Lakatan clone irradiated with 3 Gy fast neutrons. The results showed that all 12 primers were able to differentiate the cultivar Lakatan from Latundan. The amplification patterns produced by Lakatan clones were found to be distinct from the Latundan clone. Lakatan belongs to the acuminata cultivars (AAA genome) while Latundan belongs to the acuminata × balbisiana hybrids (AAB genome).

A similar experiment using 60 RAPD primers yielded similar results: 33% of the primers tested detected variation between irradiated clones, and more differences in the RAPD bands were detected with the clones irradiated with fast neutrons. The results indicated that more point mutations were generated by the fast neutron treatment than by the gamma irradiation in these materials. Despite the ease and discriminating capability exhibited by RAPD markers, problems of reproducibility and transferability were often encountered during the experiments. Hence other marker techniques were considered, such as SSR and AFLP, described in the following sections.

Figure 2 RAPD banding patterns of DNA samples derived from induced mutants (irradiated, 40 Gy gamma source), a somaclonal variant and a non-irradiated clone of banana cv. Lakatan. The figure above shows banding patterns that were generated by 18 RAPD primers. Most of the primers gave positive amplification; some primers (e.g. OPE-08, OPC-08) gave amplification products in only two or three clones, and a few primers gave no amplification products. Of those that yielded amplification products, some products were found to be monomorphic (e.g. OPE-07); others were polymorphic (e.g. OPB-08, OPC-07, and OPB-01). A = control (non-irradiated), B and C = induced mutant LK-40, D = somaclonal variant (Lakatan-Davao), M = molecular marker (1 kb ladder). Primers used are designated as OPA, OPB, OPC, OPD, OPE, OPQ or OPR followed by a number.

3.3. Simple sequence repeats (SSRs) or microsatellites

Microsatellites are another type of PCR-based marker that consist of short DNA sequences (usually 1-6 bp in length) that are tandemly repeated from two to several thousand times [22-25]. They are also known as simple sequence repeats (SSRs), short tandem repeats (STR), simple sequence length polymorphisms (SSLP), or sequence-tagged microsatellite sites (STMS). SSRs are an important class of DNA marker because they are abundant, uniformly dispersed in the genome, PCR-based, and are highly informative because they are co-dominant markers [26]. SSR polymorphism results from differences in the number of repeat units between individuals at a particular SSR locus. The variation in the number of repeat units is thought to be due to unequal crossing-over or slippage of DNA polymerase during replication of repeat tracts [27, 28]. The microsatellite loci were found to be abundant and highly variable in many eukaryotic DNA examined [23, 25]. Since the DNA sequences that flank the SSR loci are conserved, suitable primers to amplify the SSR loci were designed, resulting in highly specific PCR amplification and therefore high reproducibility. The technology is also readily transferable since information can be communicated as primer sequences.

Our participation in the banana CRP allowed us access to STMS primers and protocols from the University of Frankfurt, Germany. The results obtained show that the non-radioactive SSR marker protocol has been successfully adapted for use with banana. This SSR protocol is highly reproducible and easy to transfer to less sophisticated laboratories such as those in the Philippines.

Genomic DNA was isolated from 1-3 g of fresh leaf tissue from field-grown plants, following the protocol of Fulton et al. [19]. PCR amplification was carried out following the conditions described by Kaemmer and co-workers [13] with some modifications. Each 20 µl reaction contained 30 ng of template DNA, 1 × PCR buffer (10 mM Tris-HCl, pH 8.3, 50 mM KCl), 1.5 mM MgCl₂, 0.15 mM dNTPs, and 0.2 µM each of forward and reverse primers at 0.4 U of Taq DNA polymerase. Amplification was carried out using an MJ Research thermal cycler using the following profile: initial denaturation, 94°C for 4 min, 35 cycles of 94°C for 30 s, 50-60°C for 30 s and 72°C for 30 s, and a final extension at 72°C for 10 min. Soaking temperature was 4°C. Modification of the annealing temperature for each primer resulted in improved band resolution. Each reaction was overlaid with mineral oil. Bands were resolved using 5% denaturing polyacrylamide gels. A non-radioactive silver staining protocol [29, 30] was used to detect the AFLP fingerprints instead of the recommended g-³³P or g-³²P radioisotopes for radioactive detection. Figure 3 illustrates the typical SSR banding patterns obtained following the procedure described above.

Figure 3 SSR banding patterns of DNA samples derived from induced mutants and non-irradiated clones of banana cv. Lakatan and Latundan. The figures above show banding patterns that were generated by four SSR primer pairs.

The SSR polymorphism survey was conducted using DNA extracted from representative clones of one non-irradiated and three irradiated banana populations (LK-40, LK-3, LT-3). The laboratory of Dr. Günter KAHL, University of Frankfurt, Germany, kindly provided a total of 12 Musa STMS primer pairs. Twenty-four additional Musa SSR primer pairs were purchased from Research Genetics, USA. The results of the polymorphism survey are summarized in Table 2. Of the 36 primer pairs tested, 72.2% (26/36) yielded amplification products from all clones used. All the SSR primers that yielded amplification products could discriminate between the two cultivars tested. However, none of the SSR primer pairs tested, except for Ma1/3, were able to detect variation between the non-irradiated and irradiated clones, or between the clones irradiated with 3 Gy fast neutrons and 40 Gy gamma rays.

Table 2 Polymorphic SSR markers in irradiated and non-irradiated Lakatan and Latundan clones

^a 1 = University of Frankfurt; 2= Research Genetics, USA

3.4. Amplified fragment length polymorphism (AFLP)

Amplified fragment length polymorphism (AFLP) is the newest marker technology developed for DNA fingerprinting. AFLP combines the reliability of classical restriction-based fingerprinting methods such as RFLP with the speed and convenience of PCR-based marker techniques. However, there are some major limitations to the adoption of this promising marker technology. First, AFLP is a proprietary technology and is therefore more expensive than other PCR-based marker techniques. Secondly, the AFLP protocol developed by Keygene is based on radioactive detection using radioisotopes, and as such can be used only in laboratories where facilities for radioisotope handling and containment are available.

A method was developed for non-radioactive detection of AFLP fingerprints in banana, suitable for laboratories in developing countries. Ten AFLP primer pairs (+3/+3) were initially tested in five representative samples of irradiated clones. Optimization was based on the AFLP protocol of Vos et al. [31], AFLP Analysis System I kit (Life Technologies, Cat. No.10544-013) and Promega silver staining kit [29]. The following modifications and additional steps were taken, which cut the cost and ensured high gel/band resolution in a non-radioactive detection system:

• High quality plant genomic DNA was obtained with an additional DNA purification step using phenol instead of the recommended GIBCO-BRL Plant DNAzol. The isolation of good quality intact DNA ensures complete digestion of the genomic DNA.

• Prior to restriction digestion using Eco RI/Mse I restriction enzymes, digestibility of the genomic DNA samples was tested using cheaper restriction enzymes such as Eco R1 or Hind III.

• The reaction volumes for the restriction enzyme digestion, ligation, pre-amplification and selective amplification were reduced to 50% of the recommended total reaction volume for each step without reducing the quality of bands or the efficiency of the reaction. This modification reduces the cost and doubles the number of samples that can be analyzed per kit.

• Dilution of the pre-amplification product was varied from 1:10 to 1:50 depending on the intensity of the pre-amplification product obtained using standard gel electrophoresis on representative samples.

• A non-radioactive silver staining protocol based on Promega was used to detect the AFLP fingerprints instead of the recommended g-³³P or g-³²P radioisotopes for radioactive detection

AFLP fingerprints of representative induced mutated clones of banana and accessions of Musa textilis are shown in Figure 4. The results of the AFLP screening are summarized in Table 3. Eight of the 12 AFLP primer pairs (+3/+3) used could differentiate between the cultivars Lakatan and Latundan, as well as among the different irradiated Latundan clones. The number of bands generated ranged from 18 to 40 (mean, 34 bands) and percentage polymorphism ranged from 31.4 to 77.8%, with a mean of 51.1%.

The results clearly indicated that AFLP is ideal and more useful for fingerprinting purposes than other marker systems because of its high multiplex ratio, i.e. more bands (loci) per gel can be resolved. While more primer combinations need to be tested, these preliminary results suggest the potential value of this technique for detecting genome variation between cultivars, and for detecting genome alterations in induced mutants of banana, including those showing very similar phenotypes.

Table 3 Summary of polymorphic AFLP markers in five irradiated Lakatan and Latundan clones

Figure 4 AFLP fingerprints of Lakatan (LK) and Latundan (LT) clones. The figure shows the banding patterns (lower half of the gel only) generated using four AFLP primer pair combinations. Note the distinct fingerprints between Lakatan and Latundan cultivars. M, Gibco/BRL Marker VIII; Lane 1, LK-40-289; lane 2, LT-3-32; lane 3, LT-3-36; lane 4,LT-3-75; lane 5, LK-40-284.

4. AFLP ANALYSIS OF PARENT CLONES AND SUCKERS OF SELECTED INDUCED MUTANTS

The mutation technique, in combination with in vitro micropropagation, is a powerful technique which may well produce desired variation in well established clones. Repeated in vitro propagation is incorporated in the procedure to generate the large number of samples required and to dissociate chimeras [1, 32]. In general, mutated cells are difficult to monitor. Induced mutants are usually differentiated from each other and from the original cultivar by phenotypic analysis, which can be severely limited by the large size of the mutagenized population and by dominance, and developmental and environmental effects in the traits used in characterizing the population. The use of molecular markers may overcome this limitation. Based on the results obtained in this study, and those reported by other workers [14], AFLP has proved to have the highest potential for detection of genome alterations in somaclonal variants and radiation-induced mutants.

The materials used in this study were generated as described in Section 2. Selected parent clones consisted of those which were morphologically similar to and those with obvious differences from the non-irradiated clones of each cultivar. From 30 selected parent clones of the three mutated populations, LK-40, LK-3, and LT-3 (Table 1), 78 suckers were derived and established in the field. Suckers of non-irradiated Lakatan and Latundan cultivars were planted alongside as controls.

AFLP analysis was carried out following the procedures described in the preceding sections. Six out of 24 primer pair combinations tested namely, E-AGC/M-CAA, E-AGC/M-CAG, E-AGC/M-CAT, E-AAC/M-CAC, E-AAC/M-CAG, and E-AAC/M-CAT, were employed. AFLP bands were scored manually as present (1) or absent (0). Jaccard's similarity coefficients were computed between genotypes in each clone group, namely primary and secondary suckers of irradiated and non-irradiated Latundan, Lakatan 40 and Lakatan 3, using SAS version 6.12 [34]. Cluster analysis was applied on the similarity matrix with the unweighted pair group method with arithmetic averages (UPGMA) strategy. Dendrograms were constructed using genetics software MEGA version 2.1.

4.1. AFLP analysis among parental clones

The six AFLP primer combinations generated consistent amplified products, with an average number of 20 polymorphic bands per primer pair. The analysis of 30 parental clones, plus the non-irradiated control, using the six AFLP primer pairs allowed us to calculate the matrix for all pairwise comparisons between the parent clones. The relationships among the clones in each mutated population, LT-3, LK-40 and LK-3, are illustrated by the dendrograms in Figure 5. The analysis showed that all mutated clones differed from the non-irradiated genotype and between each other in varying degrees. Of the three mutated populations generated, the clones from the 40 Gy (LK-40) gamma-irradiated populations showed the least variation, while both populations irradiated with 3 Gy fast neutrons were more variable. The results indicate that fast neutron irradiation gives rise to more mutations and that four cycles of subculture were not sufficient to eliminate chimeras.

Figure 5 Dendrogram based on the Jaccard coefficient, showing the relationships among induced mutant parent clones generated by fast neutron (LK-3 and LT-3) and gamma irradiation (LK-40).

4.2. AFLP analysis between parent clones and derived suckers

The relationships were also calculated between parent and suckers in each mutated populations, LT-3, LK- 40 and LK-3, as illustrated by the dendrograms and electrophoretograms in Figure 6. The analysis showed that the majority in suckers were different from the parental clones validating further the results that fast neutron irradiation gives rise to more mutations and that the shoot-tip culture technique and the number of propagation cycles employed were not sufficient to eliminate chimeras.

Figure 6 Dendrogram based on the Jaccard coefficient and AFLP fingerprints showing relationships among induced mutant parent clones and their suckers. Lane 1, LT-5; lane 2, LT-5B; lane 3, LT control; lane 4, LK-456; lane 5, LK-456A; lane 6, LK control.

6. CONCLUSION

The study has produced promising induced mutants of the two most important Philippine banana cultivars, and has established molecular marker techniques for the more efficient characterization and analysis of mutation-induced changes in the Musa genome. All the methods developed can be easily adopted in less developed countries where bananas are widely grown. The non-radioactive silver staining technique for SSR and AFLP markers are suitable for local laboratory conditions. The results obtained from both morphological and molecular analyses show that more point mutations were generated by irradiation with fast neutrons than with gamma rays. However, the results indicate that the number of vegetative propagation cycles for the shoot-tip technique used may not be sufficient to eliminate chimeras completely from the mutated populations. Hence, new mutation breeding strategies need to be developed. The results obtained represent significant steps that could provide a sound basis for the successful application of mutation and molecular marker techniques to the improvement of the banana in the Philippines.

ACKNOWLEDGEMENTS

We thank the Plant Breeding Unit, FAO/IAEA Agriculture and Biotechnology Laboratory, Seibersdorf, Austria, for the irradiation treatment of the plant material used and the training grant for Ms. Eden Perez. We also thank Dr. Dieter Kaemmer and Prof. Günter KAHL of the University of Frankfurt for kindly providing the Musa STMS primer pairs. We are indebted to the IPB FIC division and Ms. Marilyn Latiza for excellent technical assistance in the field and laboratory work, respectively. This work was supported by the Research Contract No. 302-D2-PHI 8146/RB from the International Atomic Energy Agency and matching funds from the University of the Philippines Los Banos.

REFERENCES

[1] NOVAK, F.J. 1992. "Musa (Bananas and Plantains)", Biotechnology of Perennial Fruit Crops (HAMMERSCHLAG F.A, LITZ, R.E., Eds),. CAB International, Wallingford, UK, 449-488.

[2] MICKE, A., et al., Induced mutation for crop improvement - a review, Trop. Agr. (Trinidad) 64 (1987) 259-278.

[3] MICKE, A.. "Induced mutations for crop improvement", Gamma Field Symp. No. 30 Institute of Radiation Breeding, NIAR, MAFF, Japan (1991) 1-21.

[4] DONINI, B., SINNINO,. "Somaclonal variants and induced mutations in crop improvement", (JAIN, S.M., et al., Eds), Kluwer Academic Press, UK (1998).

[5] JARRET, R., et al., RFLP-based phylogeny on Musa species in Papua New Guinea, Theor. Appl. Genet. 84 (1992) 579-584.

[6] GAWEL, N., et al.,. Restriction fragment length polymorphism (RFLP)-based phylogenetic analysis of Musa, Theor. Appl. Genet. 84 (1992) 286-290.

[7] BHAT, K.V., et al., RFLP characterization of Indian Musa germplasm for clonal identification and classification, Euphytica 80 (1994) 95-103.

[8] BHAT, K.V., JARRET, R.L., Random amplified polymorphic DNA and genetic diversity in Indian Musa germplasm, Genet. Resources Crop Evol. 42 (1995) 107-118.

[9] DAMASCO, O.P., et al., Random amplified polymorphic DNA (RAPD) detection of dwarf off-types in micropropagated Cavendish (Musa spp. AAA) bananas, Plant Cell Rep. 16 (1996) 118-123.

[10] JARRET, R.L., et al., Isolation of microsatellite DNA markers in Musa. InfoMusa 3 (2) (1994) 3-4.

[11] CROUCH, J.H., et al., Microsatellite markers for molecular breeding of Musa. InfoMusa 6 (1) (1997) 5-6.

[12] LAGODA, P., et al., "Sequence tagged microsatellite sites in the Musaseae", International symposium on biotechnology of tropical and subtropical species. Conference handbooks and abstracts. International Symposium on Biotechnology of Tropical and Subtropical Species, Brisbane 94 (1997) 1.

[13] KAEMMER, D., et al., Molecular breeding in the genus Musa: a strong case for STMS marker technology, Euphytica 96 (1997) 49-63.

[14] ENGELBORGHS, I., et al., The potential of AFLP to detect genetic differences and somaclonal variants in Musa spp., InfoMusa 17 (2) (1998) 3-6.

[15] NOVAK, F.J., et al., Mutation induction by gamma irradiation of in vitro cultured shoot-tips of banana and plantain (Musa cvs), Trop. Agr. (Trinidad) 67 (1989) 21-28.

[16] MURASHIGE, T., SKOOG, F., A revised medium for rapid growth and bioassays with tobacco tissue cultures, Physiol. Plant. 15 (1962) 486-492.

[17] IAEA-312.D2.RC.579.3. WORKING MATERIAL, Cellular biology and biotechnology including mutation techniques for creation of new useful banana genotypes. Report of the Third FAO/IAEA Research Coordination Meeting. Colombo, Sri Lanka, 4-8 October 1999 (2001).

[18] IPGRI-INIBAP/CIRAD, "Descriptor for banana (Musa spp.)", International Plant Genetic Resources Institute, Rome, Italy/International Network for the Improvement of Banana and Plantain, Montpellier, France/Centre de Cooperation Internationale en Recherche Agronomique pour le Development, Montpellier, France (1996).

[19] FULTON, T.M., et al., Microprep protocol for extraction of DNA from tomato and other herbaceous plants. Plant Mol. Biol. Rep. 13 (1995) 207-209.

[20] WILLIAMS, J.G.K., et al., DNA polymorphisms amplified by arbitrary primers are useful as genetic markers, Nucl. Acids Res. 18 (1990) 6531.

[21] WELSH, J., McCLELAND, M., Fingerprinting genomes using PCR with arbitrary primers, Nucl. Acids Res. 18 (1990) 7212-7218.

[22] HAMADA, H., et al., A novel repeated element with Z-DNA forming potential is widely found in evolutionarily diverse eukaryotic genomes, Proc. Natl. Acad. Sci. USA 79 (1982) 6465-6469.

[23] LITT, M., LUTY, J.A., A hypervariable microsatellite revealed by in vitro amplification of a dinucleotide repeat within the cardiac muscle actin gene, Am. J. Hum. Genet. 44 (1989) 397-401.

[24] TAUTZ, D., et al., Cryptic simplicity in DNA is a major source of genetic variation, Nature 322 (1986) 652-656.

[25] STALLINGS, R.L., et al., Evolution and distribution of (GT)_n repetitive sequences in human genomes, Genomics 10 (1991) 807-815.

[26] BROWN, S.M., et al., "Development and application of simple sequence repeat (SSR) loci for plant genome analysis", Methods of Genome Analysis in Plants. (JAUHAR, P.P., Ed.), CRC Press (1996) 147-159.

[27] LEVINSON, G., GUTMAN, G.A., Slipped-strand mispairing: a major mechanism for DNA sequence evolution, Mol. Biol. Evol. 4 (1987) 203-221.

[28] COGGINS, L.W., O'PREY, M., DNA tertiary structures formed in vitro by misaligned hybridization of multiple tandem repeat sequences, Nucl. Acids Res. 17 (1989) 7417-7426.

[29] PROMEGA, Technical manual: silver sequence DNA sequencing system (1996).

[30] BALATERO, C.H.B., et al., "Optimization of non-radioactive AFLP (amplified polymorphic fragment length polymorphism protocol for genome analysis of important Philippine crops", Annual Scientific Conference of the Federation of Crop Science Societies of the Philippines. Gen. Santos City, South Cotabato (Philippines) 10-15 May 1999, Philippine J. Crop Science 24 (Supplement No.1) (1999) 66.

[31] VOS, P., et al., AFLP: a new technique for DNA fingerprinting, Nucl. Acids Res. 23 (1995) 4007-4414.

[32] ROUX, N., et al., Effectiveness of three micropropagation techniques to dissociate chimeras in Musa spp., Plant Cell Tissue Organ Culture 66 (2001) 189-197.

[33] SAS INSTITUTE INC. 1996. SAS/STAT Users Guide release 6 (12th ed.), SAS Institute, Cary, NC (1996).

[34] KUMAR, S., et al., Molecular Evolutionary Genetics analysis (MEGA) version 2.1 (1993). http://evolgen.biol.metro-u.ac.jp/MEGA/default.html