Introduction

Banana belongs to the family Musaceae is one of the principal food for about 400 million people across the globe [34]. Plant parts are used as alternative sources of aesthetics, fiber, and medicine [30] besides, ornamental species like Musa laterita, M. ornata, fiber yielding species like M. textilis. The production and marketing of bananas largely contribute to the livelihood of many peoples in tropical and sub-tropical countries. India is the largest producer of bananas and has widespread biodiversity [21]. Genetically important wild and cultivated Musa varieties found across the country are useful in heterosis breeding programs. The seed sterility of cultivated clones remains a concern in producing new and resistant cultivars through breeding [40]. The challenges faced in banana production due to the biotic and abiotic stresses can be dealt with by conserving the vast genetic resources and its genetic characterization. Moreover, the management of properly characterized Musa germplasm is important because of the high expense and time consumption in maintaining field gene banks for their vegetative mode of propagation. The Indian banana genome is not completely understood and remains a domain of exploration and deep digging.

The genera of Musa and Ensete belong to the family Musaceae. Out of this the wild and domesticated plantains and bananas formed the Musa group represented by five sections namely Eumusa (x = 11), Callimusa (x = 9, 10), Australimusa (x = 10), Rhodochlamys (x = 11), and Ingentimusa (x = 14) represented by only species M. ingens [2, 6]. Eumusa and Rhodochlamys have merged into new sections Musa in the recent classification, and Australimusa which were placed under Callimusa [25]. The primitive genus Ensete is present in tropical Africa and Asia, represented by a few numbers of existing species. Most of the edible bananas having chromosome number x = 11 belong to Eumusa that consists of about eleven species. Two wild species i.e. M. acuminata having AA genome and M. balbisiana having BB genome of section Eumusa crossed naturally and the majority of domesticated varieties thus originated. It resulted in hybrids with different ploidy levels and genomic constituents like diploid (AA, BB, AB), triploid (AAA, AAB, ABB) and also tetraploid (AAAB, AABB, ABBB) [49]. Cultivated banana (Musa spp.) can be divided into either plantain (starchy even after ripening and generally made palatable by boiling or frying and roasting) or tabletop banana (sweet in taste, easily digestible, and consumed raw when ripened). Karyotype variations among varieties provide useful information regarding their apparent origin and evolution. Due to random and repeated translocation, insertion, deletion, and transposition events, the chromosomal characteristics are subsequently altered which can be detected in their metaphase chromosome morphology and karyotype. The evolution of important variations can be analyzed by statistical indices that measure the degree of deviation from a standard karyotype that has all metacentric chromosomes of equal length [51]. A number of intra-chromosomal asymmetries (asymmetry increases with increase in chromosomes with the sub-telomeric or telomeric position of the centromere) and inter-chromosomal asymmetries (asymmetry increases with increase in variation of chromosome length in a chromosome complement) indices have been proposed since Lewitsky [33] to Peruzzi et al. [42]. Afterward, notable studies on karyotype asymmetry in various plant species were carried out [33, 54] including banana [18].

DNA markers or molecular markers like AFLP (Amplified Fragment Length Polymorphism), CAPS (Cleaved Amplified Polymorphic Sequence), RAPD (Random Amplified Polymorphic DNA), RFLP (Restriction Fragment Length Polymorphism), retrotransposons, SCAR (Sequence Characterized Amplified Region), and SSRs (Simple-Sequence Repeats) are being extensively used in the last two decades due to reliability and effectiveness in characterizing various plant species [22]. Due to advancement in sequencing technologies, identification of novel sequence-based DNA markers such as GBS (Genotyping-by-sequencing), NGS (Next-Generation Sequencing) based marker system, diversity array technology (DArTSeq), and SNP (Single Nucleotide Polymorphism) markers have facilitated genetic marker development and characterization of more complex genomes [37]. Simple sequence repeats or microsatellites, one type of repetitive DNA sequence (repeats of motif sequence of 1–6 bp length), are quite abundant in plant genomes. Among other available markers, the SSR marker was widely used in molecular breeding because of its high polymorphism, codominant inheritance, efficiency, ability to detect allelic variation [48]. SSR markers still remain efficient in revealing the genetic diversity in M. acuminata [7, 10, 12, 31, 36, 55] and in M. balbisiana [4, 43, 45]. The availability of developed Musa SSR is still far behind that of other crop plants. Christelová et al. [8] standardized a genotyping procedure with a set of 22 SSR markers for Musa [26, 31]. Later on, Christelová et al. [7] used this set of 22SSR markers to characterize 630 Musa accessions of the ITC (International Transit Centre), Leuven, Belgium, and efficiently detected genetic variation that reinforced the morphological and cytogenetic basis of classifications. These markers were used efficiently in marker-assisted selection on banana genotypes besides genetic diversity study [16, 38]. However, Backiyarani et al. [3] recently developed a database (MusatransSSRDB) having the information on transcriptome-derived SSRs. But information on SSR markers of Indian banana varieties is very scanty which can be explored for the genome-specific marker development. Moreover, the species diversity in India needs to be cytogenetically explored for its productive utilization in the breeding and systematic study. Thus, this study is focused on a combined approach of karyotype analysis and SSR marker study using 22 SSR markers of Christelová et al. [8] in twelve varieties of plantains and tabletop Musa available in India to find out genetic diversity, ploidy level, and evolutionary relationships, if any, for utilization in genome characterization and banana breeding program.

Material and methods

Materials

A living collection of twelve varieties of Musa L. of section Eumusa from different places of Odisha have been kept in the net house of the Botany Department, Utkal University, Bhubaneswar, Odisha (Table 1).

Table 1 List of the twelve of dessert and cooking banana (Eumusa) germplasm collected from different parts of Odisha
Full size table

Chromosome preparations

Young fresh root tips having 0.5 cm–1.0 cm length of each Musa variety were collected at10.00 am. The root-tips from five individuals of each variety were pre-treated in para-dichloro benzene (pDB): aesculin (1:1) solution for 3½ h at 18 °C followed by overnight fixation in glacial acetic acid: ethanol (1:3) at room temperature. Root-tip squash preparations were done with 45% glacial acetic acid after cold hydrolysis and 2% orcein staining [18]. Scattered chromosome plates were micro-photographed under Olympus BX-53 microscope with 100 × magnification. Measurements were taken with Q-Capture Pro 7 (Canada) software. At least five well-scattered metaphase chromosome plates were chosen out of 10–15 preparations for each variety for getting measurement accuracy of the karyotypes.

Chromosome measurements and statistical analysis

The total chromosomes length and volume of the karyotype were calculated as per the formulae used in our earlier paper [18]. The total form percentage (TF %) of a variety calculated from the individual Form percentage (F %) which is the average of all F% of individual chromosomes of a karyotype [32]. The Interphase Nuclear Volume (INV) was calculated applying formula 4/3πr3, where ‘r’ is the radius of the chromosome [15].

Idiograms based on mean values of the relative lengths and arm ratios of the chromosomes were analyzed and presented in bar form. Chromosomes of karyotype were classified in terms of the position of its constrictions and length of the chromosome. Type A = median constricted chromosomes with nearly median to sub-median secondary constriction (1.66–3.93 µm); Type B = sub-median constricted chromosomes having nearly sub-median secondary constriction (2.33–2.98 µm); Type C = nearly median primary constricted chromosomes (0.91–3.18 µm); Type D = nearly sub-median primary constricted chromosomes (0.74–3.69 µm). Besides the karyotype asymmetry, other karyotype parameters were calculated from five pairs of karyotype parameters to obtain scatter plots [19, 41, 54]. The dendrogram was constructed with the help of NTSysPC software version 2.20 [44] using five intra-chromosomal asymmetry indices like degree of asymmetry (A), intrachromosomal asymmetry index (A1), karyotype asymmetry index (AsK%), index of karyotype symmetry (Syi), and total form percentage (TF %).

DNA isolation and purification

Genomic DNA of the 12 Musa varieties was isolated following CTAB (cetyl trimethyl ammonium bromide) method with some modifications [47]. Young cigar leaf (2.0 g) was taken in a mortar pestle with a pinch of PVP (Poly Vinyl Pyrrolidone) and crushed into powdered form in liquid nitrogen. DNA extraction buffer with the composition of Tris HCl (100 mM, pH 8.2), EDTA (20 mM) NaCl (0.5 M, w/v), CTAB (2%, w/v) and beta-mercaptoethanol (2%, v/v) was added in equal volume to it and incubated at 60 °C for an hour in a water bath. The mixture was cooled down to room temperature, then centrifuged after emulsifying it with 24:1 chloroform: isoamyl alcohol solution. The precipitation of genomic DNA was made by adding chilled isopropanol into supernatant following dehydration in 70% ethanol and dissolved in T10E1(Tris EDTA) buffer (pH 8.2).Genomic DNA quality and quantity of all the varieties were verified in 0.8% agarose gel. The presence of traces of RNA impurity was purified with the treatment of 150 µl of RNAase-A treatment (10 mg ml−1 stock) at 37 °C for 2 h. Subsequently, precipitation of DNA was done in chilled ethanol mixing with 3 mM sodium acetate following 24:1 chloroform: isoamyl alcohol separation. Further purification was carried out using a silica-based purification kit (Thermo Fisher Scientific) to obtain purified DNA suitable for use in PCR (Polymerase Chain Reaction). The final dilution (25 ng µl−1) of purified DNA was prepared using T10E1 buffer for PCR reactions.

Polymerase chain reaction

PCR was performed as per the protocol of Willams et al. [56] to generated SSR profiles using the set of 22 Musa-specific primers used earlier [8]. Each PCR reaction mixture (25 µl) was prepared as per our earlier protocol [52]. Briefly, dNTPs, Taq DNA polymerase, 10 × PCR assay buffer, 25 ng of primer, and 25 ng of template DNA were mixed and the PCR reaction was performed in a thermal cycler (Bio-Rad T100, USA) following our earlier protocol [52]. The amplificons were separated in 1.5% agarose gel at a constant 55 V for 1½ h and were stained with ethidium bromide (0.5 μg ml−1) dissolved in TAE buffer (40 mM Tris–acetate, pH 8.0; 2 mM EDTA). DNA bands were visualized using a gel documentation system (Geldoc XR system, Bio-Rad, USA) under UV light and were photographed. Gene ruler (100 bp DNA ladder) was used as the size standard in the gel. The PCR reactions were performed three times for each sample and only constantly obtained amplification products were scored.

Data analysis

The size of the amplified bands was scored for each allele in each variety for each set of primers. The Excel score sheet obtained from the SSR profile was used to prepare a similarity matrix [29]. A dendrogram was created by cluster analysis following the UPGMA (unweighted pair group method with arithmetic mean) using the NTSysPC version 2.20 [43].

Results and discussions

Cytological characterization

Musa exhibits a range of ploidy levels which was developed in the natural hybridization process of M. acuminata (AA genome) and M. balbisiana (BB genome)—two wild diploid species. The chromosome number in studied varieties is variable in different sections of the family Musaceae with Eumusa having x = 11 chromosomes [49]. In our study, it was found that all varieties of banana and plantain had 2n = 3 × = 33 (Fig. 1, Table 2, Supplementary Table 1). The traditional means of squash preparation and counting of chromosome numbers are still desirable before confirming the ploidy determination by non-counting methods e.g. flow cytometry or by meiotic chromosome study [1]. The chromosome size varied from small to medium in the 12 studied Musa varieties of the present study (Table 2, Supplementary Table 1). The disparity between small and medium chromosomes was quite considerable. Chromosome numbers were found to be the same in all the varieties (Fig. 1, Table 2). The length of the total chromosome varied from 39.75 µm in var. Shankara to 81.50 µm in var. Kathia. Relatively median and nearly median chromosomes were predominant in all twelve varieties. The TF% varied from 35.65% in var. Amritapani to 45.31% in var. Desi Dakhini Sagara. The number of secondary constrictions was 3 in var. Grand Naine, var. Robusta, var. Gaja Bantala to 12 in var. Amritapani, var. Desi Dakhini Sagara. The length of the total chromosome varied from 10.78 µm3 in var. Robusta to 32.24 µm3 in var. Desi Dakhini Sagara. Distinct karyotype variations were observed in different varieties of studied bananas as evident by karyograms and idiograms (Fig. 2). The Interphase Nuclear volume varied from 625.13 µm3 in var. Gaja Bantala to 2048.37 µm3 in var. Patakpura (Table 2). The variation in the number of Type C, Type D suggest the origin of Type-A chromosomes because of structural alteration of chromosome arms through translocation during the micro-evolution [13]. A greater variation and presence of Type C chromosomes rather than Type D chromosomes in most of the species suggested the greater rate of translocation of Type D chromosome to Type C chromosome during the early evolution. The variation in the karyotype formula that indicates structural changes might have led to the diversification of bananas and plantains [14, 23]. The alteration in the TF% might be due to structural alteration of chromosomal arms at an early stage of the natural evolutionary process in Musa as evident by fluorescent in situ hybridization in somatic and meiotic chromosomes [1, 7, 9, 18, 23].

Fig. 1
figure 1

Metaphase plate of the twelve Musa acuminata varieties, a var. Amritapani, b var. Champa, c var. Grand Naine, d var. Patakpura, e var. Dwarf Cavendish, f var. Kathia, g var. Chini Champa, h var. Robusta, i var. Gaja Bantala, j var. Paunsia Bantala, k var. Shankara, l var. Desi Dakhini Sagara. Bar = 10 µm

Full size image
Table 2 Detail karyotype analysis of the twelve Musa varieties with different chromosomal parameters
Full size table
Fig. 2
figure 2

Comparative karyograms of the twelve studied genotypes of Musa of the corresponding metaphase plates

Full size image

The karyotype study of all varieties consisted of chromosomes of the nearly median, median, nearly submedian, and sub-median types (Table 2, Supplementary Table 1, Fig. 3). However, the origin of the cultivated varieties was suggested from wild varieties with nearly median and nearly sub-median type chromosomes [18, 23] following the process of reciprocal translocation as evident in M. acuminata ssp. malaccensis [1]. According to Stebbins’s [51] karyotype system, the 2A, 1B, 2B 2C typification of karyotypes were found among the studied varieties (Fig. 3, Table 3).The recorded variation of intra-chromosomes was found mostly the same in all the varieties except var. Paunsia Bantala and var. Gaja Bantala (1B). The inter-chromosomal asymmetry varied among the varieties with the highest value in var. Dwarf Cavendish (2C). The karyotype asymmetry parameters like AI (heterogeneity of chromosome length and/or centromeric index), A1 (intra-chromosomal asymmetry index), A2 (inter-chromosomal asymmetry index), CVCL (relative variation in chromosome length), CVCI (relative variation in centromeric index), DI (Dispersion Index), MAR (Mean Arm Ratio), MCA (Mean Centromeric Asymmetry), Syi (index of karyotype symmetry) varied significantly among the varieties (Table 3). It was observed that most of the intra-chromosomal asymmetry parameters like MAR, AsK% (karyotype asymmetry index), A1, AI, MCA had peak values in var. Dwarf Cavendish and most of the inter-chromosomal asymmetry parameters like A2, DI, CVCL had peak values in var. Desi Dakhini Sagara. Cyto-morphological data create an immense impact on plant evolutionary and diversification studies [35, 50, 51]. With the chromosome number remaining constant, the increasing karyotype asymmetry can be best explained through pericentric inversion and unequal translocation that occurred in the primitive karyotypes leading to advanced ones [51, 54].

Fig. 3
figure 3

Scatter plots of four pairs of karyotypic parameters for twelve Musa varieties: a A1 index versus A2 index, b CVCL index versus CVCI index, c DI index versus AI index, d Rec index versus Syi index, CVCL versus MCA

Full size image
Table 3 Detail karyotype parameters of the twelve Musa varieties under Eumusa
Full size table

Scatter plots between a pair of intra- and inter-chromosomal indexes differentiate the varieties according to their asymmetry value. Karyotype with higher intra-chromosomal asymmetry positioned higher on Y-axis and with higher inter-chromosomal asymmetry placed more to the right side of the X-axis. The asymmetry difference found less among the tabletop varieties as compared to cooking varieties (Fig. 3). The karyomorphological representation of the karyotypes by karyogram (Fig. 2) and scatter plots found highly correlated, except the Rec versus Syi plot (Fig. 3d). None of the scatter plot diagrams efficiently described the intra-chromosomal asymmetry evident from the contradictory positions of scattered value according to the intra-chromosomal asymmetry index of varieties. The dendrogram constructed using the five asymmetry parameters (A, A1, AsK%, Syi, and TF%) showed affinity among the dessert banana varieties that placed all the studied varieties under a major cluster due to similar asymmetry values where Cavendish group like var. Dwarf Cavendish, var. Grand Naine and var. Robusta formed a sub-group except for var. Chini Champa and var. Kathia. These two varieties formed a sub-cluster in Cluster-II having AAB genomic constituent, which suggests their ancestry with cooking type. The cooking varieties also got separated from the dessert varieties forming separate clusters with more genetic affinity (Fig. 4).

Fig. 4
figure 4

Dendrogram showing genetic relationships among different varieties of banana of the basis of karyotype parameters

Full size image

Molecular characterization

The 22 sets of SSR primers have been tested for polymorphism in 12Musa varieties and produced 89 amplified fragments of ~ 140–230 bp in size (Fig. 5a, b, Table 4). An average polymorphic percentage was recorded at 41.89% among the primers for all studied varieties. A number of 37 alleles found polymorphic out of the total amplified 89 bands that formed 41.89% polymorphism. The dendrogram obtained from SSR markers showed Cluster-I with all the dessert banana varieties and Cluster-II with cooking varieties. Cluster analysis confirmed more genetic homogeneity among banana varieties as compared to the plantain varieties where var. Kathia showed distinct differences in their genetic makeup. Cluster-I differentiates var. Champa, var. Chini Champa, var. Grand Naine, var. Robusta var. Amritpani and var. Patakpura forming single Sub-cluster-I. The Cluster-II also formed Sub-cluster-I and –II with two plantain varieties each (Fig. 6). The genetic relationships of var. Kathia showed high genetic resemblance with Cluster-I (with mostly banana) forming an outgroup (Sub-cluster-II).The involvement of a non-reduced 2n gamete and an n gamete in the natural hybridization process might have resulted in the modern triploid banana [40]. The tabletop banana diversity might be due to their origin from a common ancestor clone and their subsequent vegetative propagation following somatic mutations or involvement of genetically related parents in some subgroups through the sexual reproduction process. The diversity of plantain might be due to somatic mutations of few introduced varieties as reported from time to time [12, 17, 49, 53]. Meiotic restitution in bananas due to unbalanced segregation of chromosome complements through monads, dyads, triads, and tetrad stage of diploid banana might be a cause of genetic variation in the sexual reproduction process [1]. However, the low genetic differences among various accessions of a subgroup cannot elucidate the huge phenotypic diversity observed inside the subgroups [27]. Only a few primers among the tested primers produced amplified fragments in our analyzed varieties. Earlier studies also reported a low fraction of primer amplification using SSR markers [39]. In addition, the primer-anchoring sequences flanking the microsatellite loci might vary adequately to avoid product amplification [11]. So, new SSR primers are required for the genetic analysis of Musa as variations in band intensities were reported in Musa [11] which is in accordance with our finding (Fig. 1). The variations in band intensity could be due to mutations in annealing sequences [28] leading to a null allele or competitive amplification [5]. The alleles of the same length (bp) might have arisen from altered alleles with convergent evolution, and could be the possible elucidation in the evolution of microsatellite markers [20]. In this study, the SSRs used for the phylogenetic analysis of the studied varieties proved an efficient tool in describing the diversity among the varieties that appear consistent with the phylogenetic tree obtained from the karyological data and asymmetry indices. However, capillary-based genotyping could resolve better genotypic information that can contribute more discrete information regarding the genetic variability and systematic position of the varieties [24].

Fig. 5
figure 5

SSR band polymorphism among the twelve varieties of M. acuminata amplified with primer a MuS2 and b MuS15. Lane M = Molecular weight marker, 1 = var. Amritapani, 2 = var. Champa, 3 = var. Paunsia Bantala, 4 = var. Desi Dakhini Sagara, 5 = var. Dwarf Cavendish, 6 = var. Grand Naine, 7 = var. Kathia, 8 = var. Chini Champa, 9 = var. Patakpura, 10 = var. Shankara, 11 = var. Robusta, l2 = var. Gaja Bantala

Full size image
Table 4 Amplification pattern of SSR bands of 12 varieties of Musa genotypes in different primers (*Christelova´ et al. [8], F = Forward primer, R = Reverse primer)
Full size table
Fig. 6
figure 6

Dendrogram showing genetic relationships among twelve varieties of banana on the basis of SSR profile

Full size image

Conclusion

The genetic diversity in cultivated bananas (Musa spp.) has remained a major concern among germplasm conservationists and banana breeders. The detailed karyotype analysis of 12 tabletop bananas and plantains belonging to Eumusa provided structural details of somatic chromosomes of Musa. Our study reconfirms the interrelationship between the karyotype asymmetry and the SSR-based characterization for the phylogenetic analysis of bananas mostly cultivated in Odisha, India. The genetic diversity assessment is able to differentiate the banana and the plantain into separate groups and also demonstrate more genetic variability among plantains rather than the tabletop banana. The narrow genetic variability of tabletop bananas could be increased by the strategic banana breeding programs with a more heterogeneous group of bananas to tackle the disease susceptibility of tabletop bananas.