The most significant result of this investigation is that the 18 RAPD primers produced monomorphic banding patterns amongst all the 27 EAHB landraces (Figs. 1, 2 and 3).
The sample represents about a third of the approximately 80 EAHB landraces identified in Uganda (Karamura 1998). The complete absence of genomic variation within the EAHB is comparable to the results obtained by Nyine and Pillay (2011) with a combined RAPD and AFLP study by Kitavi et al. (2016) using 100 SSR markers and 92 EAHB landraces, 48 of which were from Uganda. The latter study concluded that this banana subgroup is genetically homogeneous and derived from a single ancestral clone that underwent population expansion by vegetative propagation. DNA sequence similarity of the EAHB was also reported by Christelova et al. (2017) using 19 SSR markers and 22 samples from the Mutika-Lujugira group. Likely, the SSR and RAPD markers assessed only a small number of loci, and the genome coverage in these studies was low, as Noyer et al. (2005) suggested.
However, insignificant or no genetic diversity was also reported for studies that assessed the molecular genetic diversity of homogenomic subgroups of bananas. Among them are the RAPD analysis of the bananas (AAA) from Rwanda (Pillay et al. 2001), AFLP analysis of the EAHB with two primer combinations (Tugume et al. 2002), AFLP assessment of plantains (AAB) (Ude et al. 2003) and AFLP and SSR of plantains (Noyer et al. 2005). On the contrary, as expected, greater genetic diversity has been described for studies that used a mixed group of species or subgroups of bananas (Brown et al. 2009; Opara et al. 2010; de Jesus et al. 013; Nzawele et al. 2018; Hapsari et al. 2018).
Microsatellites have been propositioned as highly informative, codominant, multi-allelic, and highly reproducible genetic markers (Gebhardt 2007; Ben-Ari and Lavi 2012). They have been used widely for estimating gene flow, diversity, crossing-over rates, and evolution to uncover intraspecific genetic relatedness (Vieira et al. 2016). Despite the several advantages of SSR markers, they appear unsuitable for revealing genomic variability in the EAHB, or the genome coverage with these markers is too low. While there may be an apparent bias in using SSR as the ideal marker for diversity analysis in the literature, this study showed that the results obtained with the RAPD technique parallel those of previous SSR studies, especially in the EAHB.
The absence of DNA sequence variation in the EAHB contrasts with the broad phenotypic variability reflected in this subgroup of bananas (Karamura 1998; Karamura et al. 2011). In addition to phenotypic variability, differences in the fertility and seed set (Ssebuliba et al. 2006), Vitamin A content (Fungo and Pillay 2011), iron and zinc levels (Pillay and Fungo 2016) and cytogenetic variation in the form of translocations and aneuploidy (Nemeckova et al. 2018) have been reported for the EAHB. In addition, minor differences in genome size were reported in a sample of 38 in vitro-maintained EAHB (Nemeckova et al. 2018).
The DNA sequence uniformity provided by Kitavi et al. (2016), Christelova et al. (2017), and this study is a conundrum since the EAHB shows clear morphological differentiation. The sequence similarity has led researchers to believe that the EAHB has a monophyletic origin. This discrepancy may have arisen because the phenotype of an organism is dependent not only on its genetic constitution but also results from the combination of its genotype, the environment, and genotype-by-environment interactions (Flint-Garcia 2013). Perrier (1993) supported this view, indicating that the banana phenotype is not a direct gene expression of the genome but is influenced by environmental and other effects. Currently, most of the variability in the EAHB has been attributed to somatic mutations and chimerism (Karamura 1998). Recent evidence suggests that transposons and epigenetics, individually or in combination, are other major mechanisms that can influence the phenotype of organisms. Furthermore, polyploidy, which was once considered an evolutionary dead end by Stebbins (1950), is now widely proposed as another mechanism that influences plant phenotypic variation (Comai 2005; Madlung 2013). Several lines of evidence from different research fields suggest that the organellar genomes also contributed to plant adaptation (Budar and Roux 2011).
Since vegetative propagation bypasses the meiotic step, the offspring are expected to be identical to the progenitor except when mutations occur (Forneck 2005). However, some clonally propagated crops, including sweet potatoes (Villardon and LaBonte 1996), Cynodon (Caetano-Anollés et al. 1997; Caetano-Anollés 2002), Agave fourcroydes (Infante et al. 2003), Dioscorea spp. (Velasco-Ramirez et al. 2014), and Vitis (Riaz et al. 2018) show high genetic diversity. Despite its clonal reproduction, bananas also exhibit a high level of morphological diversification in pseudostem color, fruit size, shape and color of petiole bases, plant height and growth habit (Karamura et al. 1998; Bakry et al. 2009; Li et al. 2013).
Many mechanisms can intensify the genetic diversity in clonally propagated plants to enhance variation and provide an open system for adaptation and selection (Forneck 2005; Flint-Garcia 2013). Mutation types that usually affect a single gene include point mutations that result in single nucleotide polymorphisms (SNPs) and insertions or deletions (InDels) of base pairs. Entire genes can be deleted or duplicated (gene loss or gain). Larger-scale mutations occur at the chromosomal level, including deletions and duplications of chromosomal regions, whole chromosomes (aneuploidy), or entire genomes (polyploidy). In addition, recombination between homologous chromosomes results in the shuffling of genes during meiosis, resulting in new combinations of genes in the progeny. Inversions, homologous recombination and reciprocal translocations have been reported in bananas (Jeridi et al. 2011; Baurens et al. 2019; Martin et al. 2020). One of the factors that reduced the genetic diversity of the EAHB is domestication.
Reduced diversity of bananas via domestication
Domestication introduces major germplasm changes and reduces crop species diversity (Meyer and Purugganan 2013; Shi and Lai 2015), although Allaby et al. (2017) offer a different viewpoint. Reduced diversity during the domestication bottleneck has been proposed for bananas (Li et al. 2013; Perrier et al. 2011). Unlike crops such as maize or wheat, domestication of bananas did not completely change the plant's phenotype since cultivated bananas maintain the genetic signature of the best-known wild diploid species, Musa acuminata and M. balbisiana. This is plausible because domestication involved a limited number of major genes under monogenic recessive control (Gepts 2004). Conscious and unconscious selection of a few “domestication traits”, including parthenocarpy, seedlessness, increased fruit pulp content, increased number of useful nutrients, and perhaps increased palatability from wild diploid species about 7000 years ago, is considered to have reduced genetic diversity in edible bananas (Denham et al. 2003; 2020). Initially, the admixture of still fertile domesticates allowed mating with other fertile diploids among the subspecies to produce the vast diversity of extant bananas (Perrier et al. 2019).
On the other hand, fertile diploid bananas produced triploid bananas through diplo-gametes and meiotic restitution (Simmonds 1962; Raboin et al. 2005). Vegetative propagation made these selections less fertile or lost their ability for sexual reproduction but conferred several advantages as a means of reproduction (Denham et al. 2020). This led to a gradual reduction in the initial diversity of the selections.
Reduced genetic diversity due to domestication is ascribed to the genetic bottleneck effect, in which the population size of a crop is diminished (Kantar et al. 2017). Additional bottlenecks may have occurred when human migrations took these bananas to different regions of the world far away from the point of origin (Kantar et al. 2017).
It is known that the domestication of several crops, including bananas, was not a single-step event but a multi-step process over some time (Simmonds and Shepherd 1955; Careel et al. 2002; McKey et al. 2012). The reticulate nature of banana domestication is also supported by other research (De Langhe and de Maret 1999; Careel et al. 2002). It is assumed that each of these domestication steps introduced further changes to the germplasm, and the more domesticated germplasm pools had a narrower range of genotypic diversity (Smykal et al. 2018). Although reduced diversity during domestication is true for most crops, it is important to mention that it is not the general rule. For example, broader phenotypic diversity of fruit shape, size, and the color is observed in domesticated tomatoes (Rodriguez et al. 2011). In maize, cultivars have narrower phenotypic diversity relative to the less domesticated landraces (Troyer 1999).
Probable mechanisms for increased phenotypic diversity in the EAHB
Several mechanisms could have increased phenotypic variability in bananas. They include somatic mutations, somaclonal variation, the activity of transposable elements, new genome combinations, polyploidy, genome duplications, mitotic recombination and recombination of novel alleles (Nyine and Pillay 2011). For example, somatic mutations, somaclonal variation, retrotransposition, chimerism and epigenetic changes are the main sources of genetic variation in grapevine clones (Forneck 2005). Recently, the role of genome plasticity and epigenetic mechanisms are also implicated in inducing changes in the phenotype of organisms (Duncan et al. 2014). The widely accepted view for phenotypic changes in the EAHB is that somatic mutations provided the plants with much morphological and ecological diversity.
Possible Mechanisms Affecting Phenotypic Variability
The relatively low genetic diversity of the EAHB, in contrast to its morphological variation, is probably due to the accumulation of somatic mutations (Šimoníková et al. 2020).
Somatic mutations
Somatic mutations are an important source of variation in clonally propagated crops (Mckey et al. 2010). It has been stated that bananas are good examples of the power of somatic mutations to provide genetic variation that contributes to adaptive evolution and increased phenotypic variability (Whitham & Slobodchikoff 1981; Karamura et al. 2010). There is consensus that the phenotypic variability in the EAHB, and bananas in general, is a result of the accumulation of somatic mutations (De Langhe 1961; Ude et al. 2003; Šimoníková et al. 2020) and human selection (Christelova et al. 2016).
Karamura (1998) suggested that the phenotypic diversity of the EAHB results from somatic mutations in the meristems of the lateral rhizomes, which produce suckers that mature into plants. While some somatic mutations may be deleterious, others may offer a direct selective advantage or create a novel phenotype (McKey et al. 2010). Favourable mutations are easily identified and propagated to produce new cultivars with desirable traits (McKey et al. 2010).
Plants with obligate vegetative reproduction have higher rates of somatic mutations than those that reproduce sexually (Caetano-Anolles 2002). For instance, in Bermuda grass, a plant with obligate vegetative reproduction, the level of somatic mutations was high (10 per triploid genome). Many morphology characters in bananas, including plant statute, pseudostem color, the shape of the petiolar canal section, plant height, growth habit, bunch and fruit shape, astringency, and color of the fruit pulp, are prone to mutations (Karamura 2011; Bakry et al. 2013). However, no direct link has been associated with any character or specific mutation. Segregation tests, cloning and sequencing of relevant genes are necessary to prove any character is due to mutations or epigenetic variation (Yu et al. 2021).
Plants grown under in vitro conditions are subject to stress (Desjardins et al. 2009). Oxidative stress is one of the main reasons for both spontaneous and induced mutations and a driving force for improving crops (Sen 2012). The EAHB could have undergone similar stresses when they were transported from their origin to different regions of the world. Although the climate in the East African highlands and South East Asia is described as tropical, the East African highlands perhaps have a different microclimate that influences the phenotypes of the plants, leading to variation. Plastic adaptation to different growth environments is reported to initiate phenotypic variation, especially in vegetatively propagated crops (Denham et al. 2020).
Transposable elements as agents of diversity.
Transposable elements (TEs) are considered a major driving force in genome evolution and gene duplication resulting in rapid plant phenotypic changes (Niu et al. 2019; Quesneville 2020). The mutating potential, genomic and phenotypic changes due to transposable elements are discussed adequately in several publications (Lisch 2013; Vitte et al. 2014; Razali et al. 2019). Similarly, the key roles of TEs in fine-tuning the regulation of gene expression leading to phenotypic plasticity are reviewed by several researchers (Lisch 2013; Wei and Cao 2016; Razali et al. 2019). Researchers agree that TEs are agents of genetic diversity on which selection can act (Vitte et al. 2014; Razali et al. 2014). It is reported that the contribution of TEs to genetic diversity may be underestimated since TEs can be more active when organisms are under stress, such as in their natural environment (Horváth et al. 2017; Lanciano and Mirouze 2018). Plants in any environment are constantly under environmental changes and are affected by abiotic (light, water, and temperature) and biotic factors such as pathogens and pests. They develop various genetic mechanisms to cope with habitat heterogeneity supplemented by phenotypic plasticity (Wang et al. 2020).
TEs have influenced phenotypic traits such as fruit shape, leaf variegation, inflorescence structure, skin colour, seedless fruit development, plant height, response to disease and pest resistance, apical dominance, variation in flowering time, fruit variation, vitamin E accumulation, leaf angle, parthenocarpy etc. of many crop plants (Vitte et al. 2014; Wei and Cao 2106). Although transposable elements and retrotransposons are present in the banana genome (Balint-Kurti et al. 2000; Nouroz et al. 2017; Pratama et al. 2021), no association has been identified between any transposable element and trait. Several of the traits mentioned above are known to be highly variable in the EAHB (Karamura et al. 2012). It is known that TEs can induce spikes in mutation rates, as shown for wild populations of three sunflower species, barley and some rice cultivars (Lisch 2013).
Further evidence is required to establish whether the somatic mutations observed in bananas are due to TEs. New technologies such as NGS and improved statistical tools may make it possible to confirm whether TE-mediated polymorphisms can be linked with phenotypes and or/environmental variation.
Epigenetic variations contribute to plant evolution.
There is strong support that epigenetic mechanisms, including DNA methylation and histone modifications, generate genome rearrangements in response to biotic and abiotic environmental stresses (Duncan et al. 2014). Epigenetic mechanisms can bring about both genomic and phenotypic plasticity and produce different phenotypes when exposed to environmental variation (Duncan et al. 2014). It is likely to assume that the movement of bananas from their origin to different ecological zones, including the East African Highlands, was accompanied by abiotic and biotic stresses that influenced phenotypic variation.
Noyer et al. (2005) were the first to report on epigenetic marks in the form of a high degree of methylation polymorphism in plantains (AAB). However, no correlation was observed between the phenotypic classification and methylation diversity. Epigenetic variation is due to genomic changes that do not affect an organism's DNA sequence but can bring about several changes in gene expression (Kapazoglou et al. 2018). This may be why, despite the phenotypic variation in the EAHB, no genome sequence changes were detectible with molecular marker technology such as SSR and RAPDs.
Epigenetic variation in crop plants has affected several phenotypic characteristics such as modification of plant stature, fruit development and ripening, fertility, leaf shape, seed size, flowering time, floral symmetry, and anthocyanin pigmentation (Wang et al. 2020; Gupta and Salgotra 2022). Considerable variation in the pseudostem height, color, and disposition occurs in bananas and is used to distinguish cultivars (Bakry et al. 2009; Karamura et al. 2011). The role of epigenetics in determining these traits in bananas is unknown. It is believed that the EAHB harbors significant epigenetic diversity with heritable epialleles that can contribute to morphological diversity (Kitavi et al. 2020). Bakry et al. (2009) proposed that DNA methylation plays an important role in bananas' pathogenic response to Foc TR4. DNA methylation changes in response to salt stress have been reported in Musa acuminata (Ranganath 2018), suggesting that DNA methylation could be used to fine-tune gene expression. High throughput sequencing techniques and biochemical techniques to manipulate epigenetic marks will allow us to see the influence of epigenetics on phenotype, plasticity and evolution (Duncan et al. 2017).
Polyploidy
Polyploidy or genome doubling is a prominent characteristic of plant genomes, including bananas, and results in high levels of gene duplication (Flagel and Wendel 2009; Marfil et al. 2018). The duplicated genes can acquire new or slightly varied functions and provide the basis for gene sub-/neo-functionalization, further promoting plant species' adaptation and genome plasticity (Madlung 2013; Cheng et al. 2018). New genome combinations created via polyploidy can influence the morphological, ecological, physiological, biochemical and cytological characteristics associated with diversifying traits and adapting plants to new environments (Weiss-Schneeweiss 2013; Wang et al. 2017).
Compared to their diploid ancestors, polyploid plants outperform their diploid relatives in many aspects and exhibit superior traits such as larger organs, increased vigor etc. (Sattler et al. 2017; Wang et al. 2017). This may be true in the case of the allopolyploid EAHB with three A genomes. Two studies have shown that polyploidy can affect morphological traits in bananas. Simmonds (1952) compared petiole strength in diploids, triploid and polyploid bananas and observed that the tetraploids had the weakest petiole-breaking strength compared to the triploids and the diploids. Vandenhout et al. (1995) reported that ploidy affected fruit traits and plant height in plantain (ABB) hybrids. Induced polyploidy in bananas has shown that many features, including anthocyanin concentration and leaf pigmentation, female sterility, mitotic chromosomal irregularities, the root system, number of suckers, plant height, leaf morphology, number of living leaves at flowering and harvest, pseudostem diameter, length and diameter of fruits, greater fruit and bunch weights, and disease resistance differed compared with the original diploids (Vakili 1967; Kanchanapoom & Koarapatchaikul 2012; do Amaral et al. 2015). Recently, ten induced tetraploids generated from six diploid banana AA genotypes showed that the tetraploids generally displayed inferior vegetative characteristics than the original diploids and had about 20% lower bunch weights (Amah et al. 2019). The same study also reported a 50% decrease in fruit provitamin A carotenoids but increased lutein in the induced tetraploids. Further, pollen viability tests indicated over 70% viability for the induced tetraploids than the diploid controls.
Polyploid induction in other plants also reported cytological, morphological, anatomical, phytochemical characteristics, yield, and qualitative characteristic differences between the polyploids and their respective diploids (Viehmannová et al. 2012; Tavan et al. 2015). Combining multiple sets of chromosomes from different subspecies may increase the probability of accumulating alleles and novel epistatic interactions. This may be true in the case of the EAHB.
Chimerism
Chimerism also appears to have affected trait variability in the EAHB (Karamura 2011). Chimerism is common in plants propagated vegetatively for a long period (Tilney-Bassett 1986). The presence of chimerism has been alluded to and described in bananas, although there is no in-depth investigation. A common feature in the EAHB is that during vegetative propagation, not all the suckers produced from a single mat are identical as expected in a vegetative crop (Karamura et al. 2011). It is believed that the meristem of the mother plant may give rise to highly variable suckers due to chimerism. Chimerism is common in in vitro propagated banana plants due to the primary explants' chimeric heterogeneity (Israeli et al. 1995). Several reports have shown that chimerism has influenced the grapevine's genetic variability (Franks et al. 2002; Hocquigny et al. 2003).
Genomic and phenotypic plasticity
Since the genotype determines the phenotype, genotypic and phenotypic plasticity must be considered together in any discussion. Phenotypic plasticity has been documented in bananas (Blomme et al. 2005; Karamura et al. 2012; Taulya et al. 2014). Polyploidy introduces a high degree of plant genomic plasticity (Zhang et al. 2019). We can assume that polyploidy in the EAHB provides indirect evidence for genome plasticity in bananas. Like other plants, the EAHB has demonstrated tolerance to changing chromosome numbers (aneuploidy and polyploidy), genome size, transposable elements, insertions, deletions, and epigenome restructuring factors that induce genome plasticity (Leitch and Leitch 2008). These large-scale genomic changes restructure the transcriptome, metabolome, and proteome, resulting in altered phenotypes (Leitch and Leitch 2008).