Introduction

Bananas (Musa) are monocotyledonous, perennial plants which are grown in many tropical and subtropical countries. They are one of the most important food crops, particularly in the developing world [1]. While the sweet fruits of dessert bananas are popular in Europe and North America, plantains or cooking bananas are commonly eaten as a staple food in Africa and Latin America where they provide food security, as well as employment opportunities [2]. Furthermore, banana fruits are rich in several health-promoting minerals and beneficial phytochemicals such as vitamins and flavonoids [3].

Flavonoids are a major group of plant specialised metabolites that share a basic structure of two aromatic C6-rings connected by a heterocyclic ring [4]. Reorganisation and modification of the carbon skeleton, such as oxidation, glycosylation, acylation, and methylation create a versatile group comprising more than 9,000 different flavonoid derivatives [5]. Consequently, flavonoids do not only contribute to the nutritional value of fruits, but also play important roles in manifold processes. While the group of coloured anthocyanin pigments attracts animals for pollination and dispersal of seeds by colouring flowers and fruits, other flavonoids protect plants against UV-B irradiation or increase plant fertility [6,7,8,9,10]. Flavonoids from many species have been reported to have anti-pathogenic properties, this includes flavonoids from carnation (Dianthus caryophyllus) which have antifungal activity against the plant’s major pest Fusarium oxysporum f.sp. dianthi [11, 12]. The tropical race 4 (TR4) of the banana Fusarium wilt (commonly known as ‘Panama disease’) is caused by another Fusarium subspecies called Fusarium oxysporum f. sp. cubense (Foc) and threatens the global banana production [13]. Transcriptome analyses of susceptible and resistant banana cultivars infected by Foc TR4 revealed an increased transcription of flavonoid biosynthesis related genes in the resistant cultivar, suggesting an involvement of flavonoids in the defence against Foc TR4 [14].

Flavonoid biosynthesis is one of the best characterised pathways of the specialised metabolism and has been extensively studied in many plant species [15]. In banana, several flavonoid biosynthesis related enzymes have been identified and characterised [16, 17]. Regulation of structural genes on a transcriptional level allows a specific response to environmental influences as well as development and organ specific expression [18,19,20]. MYB transcription factors are common transcriptional regulators of flavonoid biosynthesis. While some MYBs act independently, others interact with basic helix-loop-helix (bHLH) and WD40 proteins to form a protein complex called MBW complex [21]. MYB proteins are present in all eukaryotes and characterised by highly conserved DNA-binding domains [22]. These MYB domains consist of up to three imperfect amino acid repeat sequences, based on which they are classified. R2R3-MYBs are the most abundant class of plant MYBs and reveal versatile functions in plant-specific processes [23]. Besides core- and specialised metabolism they are also involved in cell fate and -identity definition, developmental processes and the response to biotic and abiotic stresses [23]. Well-known R2R3-MYBs which act as activators of anthocyanin biosynthesis include C1 (COLOURED ALEURONE1) from maize (Zea mays), as well as PAP1 (PRODUCTION OF ANTHOCYANIN PIGMENT1/MYB75) and PAP2 (MYB90) from Arabidopsis (Arabidopsis thaliana) [24, 25]. They act as part of an MBW complex and control the promoters of the anthocyanin biosynthesis related structural genes as for example ANTHOCYANIDIN SYNTHASE (ANS) and DIHYDROFLAVONOL 4-REDUCTASE (DFR) [26,27,28,29].

In banana, 285 R2R3-MYB proteins have been identified in a genome-wide study, including several putative regulators of flavonoid biosynthesis [30]. In addition, MYB31, MYB4 and MYBPR1 – MYBPR4 have been identified as negative regulators of flavonoid biosynthesis in banana [31, 32]. Despite the recent identification of two proanthocyanidin biosynthesis activating R2R3-MYBs [33], little functional data is available on positive regulators (activators) of flavonoid and in particular anthocyanin biosynthesis in M. acuminata.

Here, we describe the regulatory properties of three MaMYBs, named MaMYBA1, MaMYBA2 and MaMYBPA2, with a possible role in the regulation of anthocyanin biosynthesis. As one of these MaMYBs was very recently published under the name MaMYBPA2 [33], we used this name to avoid confusion due to multiple protein naming. Regulatory activity was assessed by in planta complementation experiments of the anthocyanin deficient A. thaliana regulatory mutant pap1/pap2 and co-transfection experiments in A. thaliana protoplasts (see Supplementary File 1 for detailed methods). Our results show that MaMYBA1, MaMYBA2 and MaMYBPA2 are able to activate the promoters of AtANS and AtDFR as part of an MBW complex. Furthermore, we show that the activation potential of MaMYBA1, MaMYBA2 and MaMYBPA2 is increased when combined with the monocotyledonous bHLH ZmR instead of the dicot bHLH protein ENHANCER OF GLABRA3 (AtEGL3).

Main text

We aimed to analyse the regulatory properties of three MaMYBs which have been previously assigned to a possible role in positive regulation of anthocyanin biosynthesis (Ma06_g05960 or MaMYBA1, Ma09_g27990 or MaMYBA2, Ma10_g17650 or MaMYBPA2). Since all three MaMYB genes were detected in the haploid M. acuminata reference genome sequence DH (doubled-haploid) Pahang v2 [34, 35], these MYBs appear to be present in the same sub-genome, suggesting that they are different genes and not haplo-copies. We attempted to amplify the corresponding coding sequences (CDSs) on a template collection containing cDNA from different banana samples. The CDSs of all three MaMYBs were successfully amplified on cDNA derived from peel tissue of M. acuminata (AAA group) cultivar ‘Grand Naine’ grown in the field in Lucknow, India.

In a first approach, we performed a complementation assay using the regulatory A. thaliana pap1/pap2 double mutant (pap1: transposon tag allele RIKEN_PST16228 in Nö-0 background; pap2: T-DNA insertion allele SALK_093731 in Col-0 background [26]), which cannot produce anthocyanins in the seedling (Fig. 1). Seedlings were grown on anthocyanin synthesis-inducing media to analyse the ability of MaMYBs under the enhanced cauliflower mosaic virus 35 S promotor (2 × 35 S) to complement the pap1/pap2 anthocyanin deficiency. While wild-type seedlings (Col-0: Nottingham Arabidopsis Stock Centre (NASC) ID N1092; Nö-0: NASC ID N3081) accumulated high levels of red anthocyanin pigments, pap1/pap2 seedlings did not. Although MaMYBA1, MaMYBA2 or MaMYBPA2 were successfully expressed in the transgenic seedlings (Supplementary Figure S1), the anthocyanin level in pap1/pap2 plants expressing MaMYBA1, MaMYBA2 or MaMYBPA2 did not differ from that of the double mutant. Accordingly, MaMYBA1, MaMYBA2 and MaMYBPA2 do not appear to be able to complement the mutant phenotype and thus to regulate anthocyanin biosynthesis in A. thaliana in combination with the bHLH and WD40 genes expressed in A. thaliana seedlings.

Fig. 1
figure 1

MaMYBA1, MaMYBA2 and MaMYBPA2 cannot complement the anthocyanin deficient phenotype of A. thaliana pap1/pap2 mutant seedlings. (A) Representative pictures of anthocyanin accumulation in 6-day-old MaMYB-expressing pap1/pap2 seedlings. One representative plant per construct is shown. (B) Photometric measurement of the sucrose induced anthocyanin content in pap1/pap2 seedlings expressing 2 × 35 S-driven MaMYBs. Col-0, Nö-0 (wildtypes) and pap1/pap2 were used as controls. Error bars indicate the standard deviation of three biological replicates. The different numbers (1-4) represent individual, independent, transgenic lines. The numbers in the table below the graph indicate the relative anthocyanin content and the corresponding standard deviation of individual transgenic lines.

Full size image

Lloyd et al. [28] analysed the Z. mays anthocyanin regulator C1 by generating transgenic A. thaliana overexpression lines. In their study, Lloyd et al. generated three independent ZmC1-expressing A. thaliana lines that did not show an increased anthocyanin content compared to wildtype. Similar results were obtained in transgenic tobacco. However, further experiments suggested that ZmC1 must interact with the maize bHLH ZmR to activate anthocyanin biosynthesis in the heterologous A. thaliana system [28]. Phylogenetic analysis [36] showed that the anthocyanin biosynthesis regulating R2R3-MYBs from several monocots such as Z. mays and rice (Oryza sativa), are part of a different phylogenetic clade than the anthocyanin regulators from several dicots, such as A. thaliana or grapevine (Vitis vinifera) and further monocots, including onion (Allium cepa) and lily (Lilium hybrida). Experiments in snapdragon (Antirrhinum majus) have also shown that the monocot AcMYB1 can activate anthocyanin production in dicots [36]. The phylogenetic differences between anthocyanin biosynthesis activating R2R3-MYBs and the dependence of ZmC1 on ZmR in dicots suggest an explanation for our observations. For example, the MaMYBs may depend on their endogenous or at least monocot bHLH for effective activation of structural anthocyanin biosynthesis genes. To investigate the phylogenetic differences between anthocyanin biosynthesis regulating R2R3-MYBs from different plant species, an approximate maximum-likelihood tree was constructed (Fig. 2A). The resulting tree, which also included R2R3-MYBs that activate other branches of flavonoid biosynthesis, revealed two distinct clades of anthocyanin-related R2R3-MYBs (highlighted in red). MaMYBA1, MaMYBA2 and MaMYBPA2 form a clade with MYB10 from Triticum aestivum and exhibit a close evolutionary relationship with ZmC1, OsC1 and anthocyanin regulating MYBs from other monocots. In contrast, AtPAP1 and AtPAP2 fall into a second clade of anthocyanin-related MYBs. Furthermore, differences between anthocyanin biosynthesis regulating bHLHs were analysed in a second approximate maximum-likelihood tree (Fig. 2B). The tree showed that monocot bHLH proteins involved in the regulation of anthocyanin biosynthesis form a separate clade (highlighted in red). Both, ZmR and several putative anthocyanin biosynthesis regulating bHLHs from banana, fall into this clade. These phylogenetic analyses show that anthocyanin biosynthesis regulating R2R3-MYB and bHLH proteins from several monocot species appear to be distinct from other anthocyanin biosynthesis regulating R2R3-MYB and bHLH proteins, for example from A. thaliana. These differences may imply that MaMYBA1, MaMYBA2 and MaMYBPA2 are dependent on a banana or other monocot bHLH.

Fig. 2
figure 2

Rooted approximately maximum-likelihood trees of MYB (A) and bHLH (B) transcription factors. Circle sizes represent bootstrap values. Gene identifiers of banana proteins are highlighted in blue.

Full size image

To follow up this idea, we further investigated the regulatory properties of the three MaMYBs by performing co-transfection assays (Fig. 3) in hypocotyl-derived, dark-cultured A. thaliana At7 protoplasts [37] with different bHLH proteins from A. thaliana (AtEGL3) or Z. mays (ZmR) and the A. thaliana WD40 protein TRANSPARENT TESTA GLABRA1 (AtTTG1). Their potential to activate the promoters of AtDFR and AtANS, which are important structural genes of anthocyanin biosynthesis [38], was analysed. While none of the MaMYBs was able to independently activate proAtANS or proAtDFR, which both contain conserved cis-regulatory elements, a slight activation of proAtDFR was detected when MaMYBA1 or MaMYBPA2 was combined with AtEGL3 and AtTTG1. In combination with AtEGL3 and AtTTG1, MaMYBPA2 showed the strongest activation of proAtDFR and was also able to activate proAtANS. In combination with ZmR and AtTTG1, the three MYBs MaMYBA1, MaMYBA2, and MaMYBPA2 showed a significant activation potential on proAtDFR and proAtANS. All three MaMYBs showed a higher activation potential in an MBW complex with ZmR and AtTTG1 than in combination with AtEGL3 and AtTTG1.

These results show that MaMYBA1, MaMYBA2 and MaMYBPA2 are able to activate proAtANS and proAtDFR as part of an MBW complex and that MaMYBPA2 shows the strongest activation potential. Furthermore, MaMYBA1, MaMYBA2 and MaMYBPA2 show a higher activation potential when combined with the monocot bHLH ZmR instead of the dicot bHLH AtEGL3.

The phylogenetic differences mentioned above could explain the observed higher activation potentials of the analysed MaMYBs in combination with ZmR and AtTTG1 instead of AtEGL3 and AtTTG1. It is conceivable that these differences impede the interaction between the tested MaMYBs and the dicot bHLH AtEGL3. Since MaMYBPA2 shows the highest activation potential and is also able to activate proAtANS in combination with AtEGL3 or ZmR, differences between the three MaMYBs affecting the interaction with the AtEGL3 are likely and should be further investigated. Interestingly, MaMYBPA2 did not complement the anthocyanin deficient phenotype of A. thaliana pap1/pap2 mutant seedlings, but was able to activate proAtANS and proAtDFR when combined with AtEGL3 in co-transfection experiments in A. thaliana protoplasts. Although previous studies have shown AtEGL3 expression in A. thaliana seedlings [21], it is possible that the level was too low to activate anthocyanin biosynthesis in combination with MaMYBPA2. Furthermore, the At7 cell line used for co-transfection experiments was established more than 25 years ago. This long period of propagation in suspension cell culture has caused a number of genomic and transcriptomic changes [39]. These changes may also explain the differences between seedling and cell culture analyses.

Fig. 3
figure 3

MaMYBA1, MaMYBA2 and MaMYBPA2 can activate proAtDFR and proAtANS as part of an MBW complex. The ability of MaMYBs to activate proAtDFR-GUS and proAtANS-GUS reporter constructs in combination with different bHLH proteins (AtEGL3, ZmR) and a WDR (AtTTG1) was analysed by co-transfection in A. thaliana At7 protoplasts. The relative promoter activity refers to the measured GUS reporter enzyme activity. Promoter activity is given relative to the values obtained for the A. thaliana MBW complex (AtPAP1, AtEGL3, AtTTG1). Error bars indicate the standard deviation of five independent biological replicates. Statistical significance is indicated by asterisks which mark p-values < 0.05.

Full size image

The expression profiles of MaMYBA1, MaMYBA2 and MaMYBPA2 have previously been analysed in embryogenic cell suspension, seedling, root and different developmental stages of leaf, pulp and peel [30]. The expression of MaMYBA1 was strongest in pulp (developmental stage S2-S3), MaMYBA2 expression was relatively low in all samples and MaMYBPA2 showed highest expression in seedlings and early developmental stages of pulp (S1). This gene expression pattern may indicate organ specific MaMYBA1, MaMYBA2 and MaMYBPA2 activity in banana. Since flavonoid biosynthesis is largely regulated at the transcriptional level, it would be particularly interesting to analyse the expression levels of MaMYBA1, MaMYBA2, and MaMYBPA2 in anthocyanin rich organs such as bract or pseudostem to correlate anthocyanin content with MaMYB transcript abundance.

Deng et al. [32] performed an expression analysis of genes related to flavonoid biosynthesis using leaves of banana plants overexpressing the anthocyanin repressor MaMYB4. They found that the expression of MaMYBA1 and MaMYBPA2, together with MaDFR and MaANS, is decreased in plants overexpressing MaMYB4, compared to wildtype. These data support a proposed functionality of MaMYBA1 and MaMYBPA2 as transcriptional activators of anthocyanin biosynthesis in banana, as MaMYB4 could cause a feedback regulation of the positive regulators of anthocyanin biosynthesis, including MaMYBA1 and MaMYBPA2.

Recently, Rajput et al. [33] showed that MaMYBPA2 can activate the banana ANS, ANR and LAR promoters. They also showed that MaMYBPA2 can partially rescue the proanthocyanin deficiency of the A. thaliana tt2-1 mutant. The reported ability of MaMYBPA2 to activate proMaANS supports the proposed role of MaMYBPA2 in the regulation of anthocyanin biosynthesis. However, the partial complementation of the proanthocyanin deficient phenotype of the A. thaliana tt2-1 R2R3-MYB mutant, as well as the ability to activate the banana ANR and LAR promoters, indicates a role in the regulation of proanthocyanidin biosynthesis. Such dual role in the regulation of flavonoid biosynthesis has previously been suggested for R2R3-MYB transcription factors from grapevine, blueberry (Vaccinium species) and apple (Malus domestica) [40,41,42,43]. In several species, including A. thaliana, Z. mays and Petunia hybrida, anthocyanin biosynthesis is regulated by an MBW complex [21, 28, 44,45,46]. Our co-transfection assays showed that MaMYBA1, MaMYBA2 and MaMYBPA2 are able to activate proAtANS and proAtDFR as part of an MBW complex with ZmR and AtTTG1 in planta. As shown in a previous study by Pucker et al. [30], it is known that MaMYBA1, MaMYBA2, and MaMYBPA2 are all known to contain a bHLH-binding consensus motif [27]. Thus, the regulation of anthocyanin biosynthesis by MaMYBA1, MaMYBA2, and MaMYBPA2 is likely to depend on bHLH and WD40 proteins. Based on our tree (Fig. 2B), the bHLH encoding genes Ma05_g23010, Ma11_g19640, Ma11_g16060, Ma08_g11190 and Ma11_g03740 could be suitable candidates for further studies on the transcriptional activation of anthocyanin biosynthesis by MBW complexes. In addition, future analyses should additionally include MaTTG1 to identify a complete functional MBW complex in banana and to determine wether the use of endogenous WD40 protein further enhances the activation potential of the MBW complex.

The results presented, in particular the approximate maximum-likelihood trees and the co-transfection assays, show that MaMYBA1, MaMYBA2, and MaMYBPA2 have the ability to transcriptionally activate expression of structural anthocyanin biosynthesis genes in an MBW complex with a suitable bHLH partner. The activation potential of the tested MaMYBs is increased when the MaMYBs are combined with the monocot bHLH ZmR instead of the dicot bHLH AtEGL3.

This is a step towards deciphering the MBW complex-mediated transcriptional activation of flavonoid biosynthesis in banana. It also provides a basis for further research to increase anthocyanin production in banana, which could improve fruit quality and disease resistance.

Limitations

The co-transfection assays revealed that MaMYBA1, MaMYBA2, and MaMYBPA2 can activate two structural genes of anthocyanin biosynthesis from A. thaliana as part of an MBW complex. To further investigate the promotor activation potential of the MaMYBs, the co-transfection analysis could be expanded to other structural genes of anthocyanin biosynthesis, as well as promoters and bHLH and WDR candidates from banana. In addition, yeast two-hybrid or other protein-protein interaction experiments could be performed to investigate the affinity between MYB and bHLH proteins. To elucidate the regulatory role in banana and to confirm possible target genes, future studies should include an overexpression of the three MaMYBs in banana.