Introduction

Banana (Musa spp.) is one of the most important fruit and food crops in the world, which is cultivated in more than 130 countries worldwide, with annual production of more than 100 Million Metric Tons . China is the second largest producer with annual production of more than 10 Million Metric Tons in recent years (FAOSTAT 2023). In Africa, banana is a staple food for millions of people (Dotto et al. 2018). Banana is rich in nutrients (such as starch, vitamins and minerals/K) and plays a very important role in food security (Dotto et al. 2018), food industry (Kumar et al. 2019; Martins et al. 2019; Lau et al. 2020) and some other aspects, such medicine (Vilhena et al. 2018) and industrial wastewater treatment (Mouiya et al. 2019).

However, Fusarium wilt, a soil-borne disease caused by Fusarium oxysporum f. sp. cubense (Foc), is threatening banana industry and the plantations of the producing regions worldwide (Ploetz 2015; Koberls et al. 2017; Mostert et al. 2017; Staver et al. 2020). Foc is classified into several physiological races. Among them, race 1 (termed as Foc1) affects ‘Gros Michel’ (Musa spp. AAA), ‘Silk’ (Musa spp. AAB) and ‘Pisang Awak’ (Musa spp. ABB), race 2 (Foc2) attacks ‘Bluggoe’ (Musa spp. ABB), while race 4 (Foc4) is virulent not only on Cavendish (Musa spp. AAA) but also on Foc1- and Foc2-susceptible varieties (Ploetz 2015). This pathogen infects banana roots through the wound and causes wilt symptom (yellowing of old leaves, withering and even death of plants) through production of gums and gels in the xylem (Ploetz 2015) and causes heavy yield losses among susceptible bananas worldwide (Kumar et al. 2010; Scheerer et al. 2018; Staver et al. 2020; Heck et al. 2021).

Unfortunately, there are no known effective chemical fungicides and biological control agents for this disease. The efficiency of cultural methods is also limited. The best way to control this disease is growing resistant cultivars (Ploetz 2006). However, conventional breeding method has been restricted by several factors, such as the parthenocarpy nature of most cultivars and long growth cycle resulting into long breeding cycle of estimated 10–17 years (Nyine et al. 2018). Selection of disease-resistant germplasm, whatever they are from field screening, or mutation breeding, requires large-scale field evaluations and is labor- and time-consuming (Javed et al. 2004) due to its gigantic size, and long-life cycle.

Molecular markers are very efficient tools in crop improvement due to their functions in linkage mapping, genetic analysis, germplasm characterization and fingerprinting, as well as molecular marker-assisting breeding (Kalia et al. 2017; Kengo 2019; Jian et al. 2021; Maleita et al. 2021; Park et al. 2022; Chen et al. 2023; Habe et al. 2023). These molecular markers include inter-simple sequence repeat (ISSR), random amplified polymorphic DNA (RAPD), sequence characterized amplified regions (SCARs), sequence-related amplified polymorphism (SRAP), single nucleotide polymorphism (SNP), simple sequence repeat (SSR) and so on. Development of early, reliable, and reproducible selection strategies are considered as the efficient approach which could speed up the selection procedure and, eventually, be benefit for banana breeding and the control of banana Fusarium wilt (Javed and Othman 2005). RAPD markers were the first molecular markers successfully used to identify banana resistant germplasm to Foc from susceptible one (Javed et al. 2004; Zambrano et al. 2007; Prasad et al. 2018). But its application is limited due to their drawbacks, such as their dominance nature and low reproducibility (Arinaitwe et al. 2019). SSR and ISSR markers have also been used to distinguish the resistant banana germplasm from the susceptible ones, however they are not successful (Li et al. 2012; Rebouças et al. 2018). No SNP markers involved in the resistance of banana to Foc have been developed successfully yet (Arinaitwe et al. 2019).

SCAR markers are defined as DNA fragments amplified by the PCR using specific 15–30 bp primers, based on the nucleotide sequences generated from cloned RAPD and ISSR fragments related to a trait of interest (Nadeem et al. 2018). When compared to RAPD and ISSR, SCAR markers owe some advantages, such as less sensitive to amplification conditions due to its primer size and hence are more specific and reproducible (Joshi and Chavan 2012). In banana, only three SCAR markers associated with the susceptibility of banana to Foc4 have been developed previously (Wang et al. 2012, 2018). However, there are no SCAR markers related Foc4 have been applied successfully in Foc4-resistant banana germplasm screening.

In the present study, we have developed two SCAR markers with the same set of resistant and susceptible cultivars (7 resistant cultivars/lines, 23 susceptible ones), and the discriminatory power was 96.7% for SC4-F/SC4-R and 100% for SC14-F/SC14-R, suggesting their high reliability. To further confirm their reliability and reproducibility, potential resistant banana germplasm were identified with combined utilization of these two SCAR markers from 53 accessions from cultivars either resistant or susceptible to Foc4, and the results show that the consistency between these two SCAR markers was 83.0%.

Materials & methods

Plant materials used for the development of molecular markers

A total of 30 banana cultivars/mutants belonging to Musa spp. AAA Cavendish were selected as plant materials to generate SCAR markers related banana resistance/susceptibility to Foc4. These 30 cultivars/mutants and their degrees of resistance or susceptibility to Foc4 were shown in Table S1.

DNA preparation

The cetyltrimethylammonium ammonium bromide (CTAB) method (Murray et al. 1980) was employed to extract the DNA from the 30 banana cultivars/mutants. After determination the quality and concentration of DNA with a nucleic acid protein analysis (Eppendorf BioPhotometer Plus, Germany), the concentration was adjusted to 50 ng/μL with ddH2O before use. The size and the integrity of the genomic DNA fragment were tested by gel electrophoresis on 1.0% (w/v) agarose (2 µL DNA + 1 µL 6 × loading buffer + 4 µL ddH2O, 120 V, 15–20 min) followed by visualized on a UV transilluminator and photographed.

Cloning and sequencing of SRAP products

To get SRAP products for further development of SCAR markers associated with banana resistance/susceptibility to Foc4, 100 pairs of random primers (each containing the original 10 bases) were synthesized, and five resistant (‘Nongke 1’, ‘Nantianhuang’, ‘Dongjiao 1’, bxm51, and kangku1) and susceptible cultivars/mutants (g30, ‘Dongguan Zhongba’, ‘Qiwei’, ‘Baxijiao’ and ‘Beida Ai’) were selected as the plant materials. These primers are designed according the rules described by Li et al. (2001) and obtained from Sangon Biotech (Shanghai) Co., LTD (Table S2). PCR amplifications ran in a Bio-Rad PCR (Bio-Rad T100, Bio-Rad, Hercules, CA, USA) according to the manufacturers’ instructions. The total reaction mixture was referred to Table S3 and the conditions for reaction were mentioned in Table S4.

To get the SRAP products, the PCR products (8–10 μL) were fractionated by gel electrophoresis on 1.8% (w/v) agarose gels in 0.5 × Tris–borate-EDTA (TBE) buffer (pH 8.0) after loaded with 4 μL of 6 × loading buffer. Gel was ran at 120 V for approximately 40 min and then visualized using a UV transilluminator and photographed with gel imaging system (Guangzhou Ewell Bio-Technology Co., Ltd., Tanon 4120, CHN) after staining with Goldview. After excised from agarose gels, the PCR products were purified by Tiangel Midi purification kit (Tiangen, CHN) according to the manufacturers’ instructions. To obtain the fragment information of the linked SRAP (the PCR products), they were sent to Sangon Biotech (Shanghai) Co., LTD for sequencing.

Generation of SCAR markers

The SRAP sequences obtained were analyzed by BioXM2.7 software, followed by blasting in NCBI and http://banana-genome-hub.southgreen.fr/blast. Primer3.0 software was employed to design primers for the development SCAR markers from SRAP sequences (Table S5 and Table S6). DNA samples and PCR reaction mixture were the same as that for random markers mentioned above. The reactions were programmed according to Table S7 and Table S8 for SC4-F/SC4-R and SC14-F/SC14-R, respectively. The method for gel electrophoresis of the amplified products was the same as that for SRAP products described above.

Validation of SCAR markers

The SCAR markers obtained were validated firstly with four resistant and susceptible cultivars/mutants, respectively. The additional 22 cultivars/mutants showing different degrees of resistance to Foc4 (Table 1) was employed to validate the reliability of the SCAR markers using the pair of primers with single band after amplification. The PCR reaction system is shown in Table S3 while the conditions for PCR reactions were described above as for the validation of SCAR markers in Table S7 and Table S8.

Table 1 Thirty banana cultivars/mutants (Musa AAA Cavendish) used in the present study and their resistance to Foc4 predicted by two SCAR markers
Full size table

Molecular marker assistant selection of banana germplasm resistant to Foc4

These two SCAR markers developed in the present study were employed to screen potential resistant banana germplasm from the same group of new accessions (in total 53, of which 21 are natural mutants or mutated from resistant cultivars, 29 from susceptible ones, while the origin of the other three ones was unknown). The leaf from young seedling was used to extract DNA.

Results

DNA quality evaluation

The quality of the DNA extracted from banana leaves was checked with BioDrop μLite (Holliston, MA, USA). The result showed that the concentration of DNA samples was more than 400 ng/μL, the OD260/OD280 ratio for each sample was in the range of 1.7–2.0, indicating they could be used for further experiment. Fig. S1 showed DNA extracted from some representative samples.

SRAP molecular markers related to Foc4 resistance/susceptibility

After amplification with a primer set Me1-Em2, a special band of approximately 800 bp appeared in the five susceptible cultivars/mutants (g30, ‘Dongguan Zhongba’, ‘Qiwei’, ‘Baxijiao’ and ‘Beida Ai’) (Fig. 1a). But this was not the case for the five resistant ones (‘Nongke1’, ‘Nantianhuang’, ‘Dongjiao 1’, bxm51 and kangku1). Similarly, a special band of approximately 400 bp was only present in the five susceptible cultivars/mutants after amplified with primer set Me9-Em1 (Fig. 1b). These two pairs of primers were potentially related to the susceptibility of banana to Foc4.

Fig. 1
figure 1

The PCR products amplified with random primer sets in 10 banana (Musa spp. AAA) cultivars/mutants either resistant or susceptible to Foc4. a Primer set Me1-Em2; b Primer set Me9-Em1. M: 2000 bp marker; lanes 1–5: resistant cultivars/mutants; lanes 6–10: susceptible cultivars/mutants; Foc4: Fusarium oxysporum f. sp. cubense race 4; Arrows indicate the specific bands related to disease susceptibility

Full size image

The PCR products from both Me1-Em2 and Me9-Em1 primer sets mentioned above were cut and purified followed by sequencing. The sequences from ‘Baxijiao’ were selected as the representative (the sequences from different cultivars/mutants were nearly the same) and shown in Fig. 2a (from primer set Me1-Em2) and Fig. 2b (from primer set Me9-Em1), respectively.

Fig. 2
figure 2

The sequences amplified with primer sets Me1-Em2 (a) and Me9-Em1 (b) from banana (Musa spp. AAA)

Full size image

Development of SCAR markers from SRAP ones

For the development of a SCAR marker, four pairs of primers (Table S5) were designedfor the PCR product from primer set Me1-Em2 using DNA from eight banana cultivars showing resistance or susceptibility to Foc4 as the template. Another four pairs of primers (Table S6) were designed for the PCR product from primer set Me9-Em1.

A single, clear band of 251 bp or 295 bp was present only in the four susceptible cultivars/mutants (‘Dongguan Zhongba’, ‘Qiwei’, ‘Baxijiao’ and ‘Beida Ai’) but not in the other four resistant ones (‘Nongke 1’, ‘Nantianhuang’, ‘Dongjiao 1’, and bxm51) after amplified with primer set SC4-F/SC4-R (Fig. 3a) or SC14-F/SC14-R (Fig. 3b), respectively. Unfortunately, there were no any assignments with these two fragments when blasted in NCBI, but this was not the case when blasted in Banana Genome Hub database (http://banana-genome-hub.southgreen.fr/blast). With all available Musa genome sequences, we employ the BLAST program to find potential locations of these two markers. Interestingly, both markers seem to locate in the mitochondrion genome. The 251 bp fragment from SC4-F/SC4-R is located in the position from 72 to 322 bp of the original SRAP band and showed a 99.60% similarity to a fragment of 250 bp located in chrUn_random (from 6,569,491 to 6,569,741) of the M. acuminate Pahang (v2) and in the mitochondrion genome (from 2,709,912 to 2,710,162) of the M. acuminate Pahang (v4). Differently, the 295 bp fragment from SC14-F/SC14-R is imperfectly matched to multiple regions of the mitochondrion genome. It was located in the position 15 to 309 bp of the original SRAP band and showed only a 71.53% similarity to a fragment of 211 bp located in Contig78419 of the M. acuminate calcutta4 (v1).

Fig. 3
figure 3

The PCR products amplified from banana cultivars/mutants (Musa spp. AAA) either resistant or susceptible to Foc4 with SCAR primer sets. a SC4-F/SC4-R, 251 bp; b SC14-F/SC14-R, 295 bp; M: 2000 bp marker; lanes 1–4: resistant cultivars/mutants; lanes 5–8: susceptible cultivars/mutants; Foc4: Fusarium oxysporum f. sp. cubense race 4; and SCAR: sequence characterized amplified region

Full size image

To test the reliability of those two SCAR markers mentioned above, they were used to distinguish the additional 22 cultivars/mutants with known resistance or susceptibility to Foc4. As shown in Fig. 4 and Table 1, both SCAR markers were able to discriminate the resistance from susceptibility of banana to Foc4. An unexpected, weak band only appeared in one (kangku5) out of the 30 cultivars/mutants after amplified with SC4-F/SC4-R. The discriminatory power was 96.67% for SC4-F/SC4-R while it was 100% for SC14-F/SC14-R. These results suggest that the two SCAR markers developed in the present study can be applied in molecular marker-assisted selection for banana germplasm resistant to Fusarium.

Fig. 4
figure 4

Validation of SCAR markers in 23 banana cultivars (Musa spp. AAA) either resistant or susceptible to Foc4. a SC4-F/SC4-R, 251 bp; b SC14-F/SC14-R, 295 bp; M: 2000 bp marker; lanes 1–4 and 23: resistant cultivars/mutants; lanes 5–22: susceptible cultivars/mutants; Foc4: Fusarium oxysporum f. sp. cubense race 4; SCAR: sequence characterized amplified region. * ‘Nongke 1’ is the cultivar also among the eight ones in Fig. 3

Full size image

Identification of potential resistant germplasm with two SCAR markers

In the present study, two SCAR markers were employed to screen potential Foc resistant banana germplasm from 53 accessions (50 from resistant or susceptible cultivars, 3 from resistance-unknown accessions). SC4-F/SC4-R revealed that 17 accessions were identified as resistant germplasm while the other 36 as susceptible ones, and it was 16 resistant accessions and 37 susceptible ones for SC14-F/SC14-R (Fig. 5, Table 2). Forty-four out of the 53 tested accessions revealed the same resistance to Foc between these two SCAR markers and the consistence was 83.0% (it was 88.0% when the 30 cultivars/mutants used for the development of SCAR markers were included in). These results further confirmed the reliability and reproducibility of the two SCAR markers developed in the present study.

Fig. 5
figure 5

Screening of potential resistant gerplasm from 53 banana accessions (Musa spp. AAA Cavendish) from cultivars either resistant or susceptible to Foc4. Resistant-origin: 5 (nk4), 6 (rk5), 7 (reke1), 14–17, 21 (B40-Yueyoukang 1), 22 (B20-Yueke), 23 (B40-Yueke), 24–25 (B20-Reyan1), 26 (B30-Reyan 1), 27–28 (B20-Nantianhuang), 29 (B30-Nantianhuang), 30 (k1j1), 38 [l1n1(1)], 40 (l3-3), 44 (l3), 49 (r1-2–1); Susceptible-orgin: 1 (hongyan 3 ny3), 2 (‘Hongyan 2’), 3 (‘Hongyan 3’), 4(‘Hongyan 5’), 8 (‘Reke2’), 9–11 (2–21), 12–13 (1–10), 18–20 (B40-BX), 31 (bx1-1 nai), 32 (b23-1), 33 [g20k71(1)], 34,36 (g20k72), 37 (g20-11–71), 39 (g20k71), 41 (g20-1 k); 42 (rk2), 43 (rk3), 45 [g20k71(2)], 46 (g20-1–1), 50 (wghd3), 51 (wghd 51), 52 (wghd 52), 53 (wghd); Unknown origin: 35 (wgh), 47 (hb1), 48 (hb3)

Full size image
Table 2 Banana accessions (Musa spp. AAA) screened by two SCAR markers developed in the present study
Full size table

In the present study, most tested accessions kept the resistance or susceptibility from their mother plants, but four susceptible-origin accessions (bx1-1 nai, hongyan 3 hy3, ‘Hongyan 5’, and ‘Reke 2’) were identified as resistant germplasm with both SCAR markers. On the contrary, nine resistant-origin accessions [l3, l3-3, r1-2-1, nk4, B40-Yueyoukang 1 (1–5)] were identified as susceptible ones. In total, two newly obtained cultivars, four mutants and 6 offsprings generated directly from mutated cultivars were identified as resistant germplasm by both SCAR markers developed in the present study (Fig. 5, Table 2).

Discussion

It is estimated that by 2040, only Foc tropical race 4 (Foc TR4) might spread to 17% of the world’s banana production regions and the yield is estimated as 36 million tons which is worth over US$10 billion (Staver et al. 2020). A banana breeding program aiming to control this devastating disease is extremely important and necessary. Nevertheless, traditional hybrid breeding strategy is not suitable for Musa AAA cultivars due to parthenocarpy. Some other strategies, such as somaclonal variants (Hwang and Ko 2004), field tested in more tropical regions (Molina et al. 2016), and mutation breeding (Kishor et al. 2018) have been employed commercially. However, they are time-consuming and costly due to the necessity of verifying the results by field evaluation over an extended period (Javed et al. 2004). Furthermore, there is a risk in spread of the pathogen when field disease resistance test is carried out. Thus, the development of rapid and reliable methods to screen resistant germplasm at earlier developmental stages would be of great interest for the banana breeding against Foc (Collard and Mackill 2008). Molecular markers associated with the resistance or susceptibility fit this requirement very well due to their several advantages, such as accuracy, stability and effectiveness of operation (Park et al. 2022). Most importantly, molecular-assisted breeding only needs a sample of less than 0.5 g from a new germplasm, instead of a large quantity of offsprings from the new germplasm as used in conventional disease resistance test. This will greatly shorten the breeding cycle.

Till today, several types of molecular markers, e.g. RAPD, SSR and ISSR, have been employed in studing the resistance/susceptibility of bananas to Fusarium wilt. For example, the genetic diversity of wild Musa acuminata ssp. Malaccensis in Malaysia showing varying degrees of resistance to Foc TR4, was assessed with RAPD markers by Javed et al. (2004). The authors found that a band specific to either resistant or susceptible seedlings could be obtained from three primers (namely primer OPA-03, 24 and 21), respectively. A RAPD marker named OPK-03 was identified as capable of distinguishing the susceptible cultivar from the resistant ones, because it could amplify a band of 485 bp only in the susceptible banana genotype (Zambrano et al. 2007).Recently, Kishor et al. (2018) have also identified two RAPD primers, namely OPN-06 and OPR-07, to screen potential mutants resistant to Foc1. However, the application of RAPD markers in molecular-assisted breeding has been restricted due to their disadvantages. Firstly, the PCR products from RAPD primers often have more than one band after amplification of multiple gene loci. Secondly, they have relatively lower reproducibility when compared to some other molecular markers. The amplification reactions for RAPD markers require highly standardized experimental protocols because they are very sensitive to the reaction conditions (Jarocki et al. 2016). Moreover, RAPD markers generally require DNA in relatively higher quality and at higher molecular weight (Athe et al. 2018). SSR markers owe some advantages over other molecular markers, such as RAPD and SCAR. These include the multi-allelic and co-dominant natures of the genome providing wide coverage as well as high reproducibility these features enable them to distinguish homozygotes from heterozygotes (Athe et al. 2018). However, till today, the SSR markers have been able to detect genetic diversity but could not differentiate between banana genotypes susceptible to Foc4 from those that are resistant (Rebouças et al. 2018). Similar results were observed with ISSR markers in relation to Fusarium wilt of banana (Li et al. 2012), though SSR and ISSR markers have been utilized successfully in studying Fusarium wilt resistance in some other crops (Anjani et al. 2018; Mahmoud and Abd El-Fatah 2020).

SCAR markers, in contrast to RAPDs, are dominant and could detect a single locus. In banana, Cunha et al. (2015) have successfully generated a SCAR marker associated with the susceptibility of banana to Foc1 from a RAPD primer OPP-12. The PCR products from this marker were present only in the 20 tested susceptible banana cultivars but not in the resistant ones. Wang et al. (2012) have successfully converted two RAPD primers into SCAR markers (ScaU1001 and ScaS0901), which could distinguish bananas resistant and susceptible to Foc4. However, the cultivar number for the development of these SCAR markers was only five, which might result in low reliability and reproducibility. Moreover, one of these two markers was identified as not efficient (Silva et al. 2016). More cultivars showing different resistance to the pathogen used in developing molecular markers will be benefit for their reliability and reproducibility. Wang et al. (2018) have also successfully developed one SCAR marker linked to Foc4 susceptibility but without application of this marker in resistant banana germplasm screening. In the present study, two SCAR markers associated with banana susceptibility to Foc4 were successfully generated from SRAP markers using 30 cultivars/mutants that were either resistant or susceptible to Foc4. We used both SCAR markers to screen 53 banana germplasm accessions from resistant and susceptible cultivars. The results were highly consistent, confirming the marker’s reliability and reproducibility. The combined use of these two SCAR markers will accelerate banana breeding against Foc4.

Two SCAR markers obtained in the present study were predicted to be located in mitochondrion genome. Similar result was observed by Wang et al. (2018). Interestingly, we have analyzed available sequences amplified from banana Foc1 or Foc4 resistant linked markers in previous studies (Wang et al. 2012; Cunha et al. 2015) and found that two out of three were also present in mitochondrion genome. These findings indicated that mitochondrion DNA plays a crucial role in in banana resistance to Foc. Molecular markers linked to several aspects, including disease resistance, have been successfully generated from mitochondrial DNA (mtDNA) of some other plants (Ding et al. 2021; Murali et al. 2021) and many evidences have already demonstrated that mitochondrial proteins contributed to plant resistance/tolerance to biotic and abiotic stresses (Robles et al. 2018; Kim et al. 2021; Qiu et al. 2021; Meng et al. 2022).

In the present study, the two SCAR markers developed do not exhibit any annotated functional genes in adjacent regions (upstream 100 kb and downstream 100 kb). This is, at least partially, due to the fact that no complete banana mitochondrial chromosome data is currently available. More transcriptome data might be helpful in verifying the direct association of this marker with coding genes. Another alternative explanation is that the locus of marker SC4-F/SC4-R functions in cis. For better understanding the functions of mitochondrial proteins, further work should be focused on assembly and analysis of the complete mitochondrial genome sequences of banana cultivars either resistant or susceptible to Foc4. We believe that Pacbio hifi-sequencing will play a crucial role in resolving this problem in the near future.

Conclusions

In the present study, two reliable SCAR markers (SC4-F/SC4-R and SC14-F/SC14-R) were developed from SRAP markers successfully using 30 cultivars/mutants showing varying resistance to Foc. In addition, we have combined the use of them to screen potential resistant banana germplasm and have got highly consistent results. This method could screen large amount of mutants or any other newly obtained germplasm rapidly and at low cost. We are sure the combination application of these two SCAR markers could further improve the reliability and reproducibility of identification, and finally greatly speed up banana breeding against Foc4.