Transcriptome analysis reveals key defense genes associated with resistance to banana blood disease

Abstract

Background

Bananas are a globally significant crop, but their production is threatened by banana blood disease (BBD), caused by Ralstonia syzygii subsp. celebesensis (Rsc). While resistant cultivars like ‘Khai Pra Ta Bong’ (AAA genome) have been identified, the genetic basis of resistance remains unclear. This study uses RNA sequencing (RNA-seq) to investigate gene expression changes in ‘Khai Pra Ta Bong’ over seven days post-infection with Rsc isolate MY4101. Candidate genes associated with resistance were also evaluated in other cultivars, including resistant, moderately resistant, and highly susceptible cultivars.

Results

RNA-seq analysis identified several genes associated with BBD resistance mechanisms, showing significant upregulation as early as 12 h post-inoculation. Key molecular processes, including xyloglucan endotransglucosylase hydrolases, receptor-like kinases, and glycine-rich proteins, were enriched at 24 h post-inoculation, highlighting the activation of effector-triggered immunity (ETI). Quantitative real-time RT-PCR (qRT-PCR) validation further confirmed the differential expression of several key defense-related genes in other banana cultivars.

Conclusions

The study highlights the potential of these differentially expressed genes as genetic markers for BBD resistance, which could be valuable for breeding programs aimed at enhancing resistance in banana cultivars. These findings provide important insights into the molecular mechanisms of disease resistance and contribute to developing more efficient strategies for sustainable banana production and global food security.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12864-025-12056-0.

Background

Bananas (Musa spp.) are one of the world’s most economically vital crops. Beyond serving as a staple food, they are crucial in ensuring food security and providing income for smallholder farmers. However, banana production faces severe challenges from banana blood disease (BBD), caused by Ralstonia syzygii subsp. celebesensis (Rsc), a bacterial pathogen that leads to devastating losses through wilting symptoms, internal bleeding-like discoloration, and eventual plant death [1]. First reported in Indonesia, BBD has since spread to other regions, including Malaysia and Thailand, threatening major banana-growing areas. Effective management strategies and disease-resistant varieties are urgently needed to sustain banana production and ensure food security. It has been reported that BBD causes a yield loss of up to 100% in severely infected fields, particularly in susceptible cultivars such as Musa acuminata and Musa balbisiana hybrids, including popular dessert and cooking bananas. In Southeast Asia, where the disease is endemic, yield loss typically ranges from 30 to 80%, depending on cultivar susceptibility, environmental conditions, and management practices [2, 3].

Significant advances in plant immunity emerged in the 1990 s with the identification of the first plant resistance (R) genes, which were categorized into five functional classes: detoxifying enzymes, intracellular kinases, intracellular and cell-surface receptors, and receptor kinases. Notable examples include HM1 in maize, which encodes a toxin-detoxifying enzyme [4], and Pto in tomato [5] encoding a serine/threonine kinase. The most extensively studied group comprises NBS-LRR proteins (e.g., RPS2, RPM1, N, L6), which are key in pathogen recognition and signaling [6–8]. Remarkably, the N gene shares a conserved TIR domain with innate immune receptors in animals, highlighting evolutionary parallels across kingdoms [9]. These discoveries laid the foundation for understanding plant immune mechanisms, particularly the broad-spectrum resistance conferred by NBS-LRR genes [10, 11]. However, despite such progress in model plants, the molecular basis of resistance remains poorly characterized in banana, especially in response to Ralstonia syzygii subsp. celebesensis (Rsc), the causative agent of banana blood disease. Recent studies have identified cultivars with varying resistance levels to Rsc. For instance, Kawicha et al. [12]. revealed a broad spectrum of resistance in 85 Thai banana germplasm accessions, with only two accessions, ‘Khai Pra Ta Bong’ and ‘Pisang Ambon’ (both from the AAA genome), exhibiting high resistance. Similarly, Nitayaros et al. [13]. demonstrated the resistance of ‘Khai Kasetsart 2’ (AA genome) and moderate resistance of ‘Raksa’ (AA genome), correlating resistance levels with the expression of pathogenesis-related protein genes, such as PR3 and PR10, using qPCR assays. These findings highlight the genetic diversity available for breeding programs but also underscore the limited understanding of the molecular mechanisms underlying resistance, which hampers the development of effective molecular tools for disease management.

Advancements in molecular techniques, particularly RNA sequencing (RNA-seq), offer powerful tools to address this knowledge gap. RNA-seq enables high-resolution analysis of gene expression dynamics, facilitating the identification of differentially expressed genes (DEGs) involved in plant-pathogen interactions [14–16]. This approach provides insights into resistance pathways and supports comparative analyses across resistant and susceptible cultivars under varying conditions. Molecular markers identified through such studies have revolutionized breeding programs by allowing marker-assisted selection and accelerating the development of disease-resistant varieties [17–19].

This study focuses on the highly resistant cultivar ‘Khai Pra Ta Bong’ (AAA genome), identified as a key genetic resource for breeding programs [12]. Using RNA-seq, the study examines gene expression changes in response to infection with Rsc isolate MY4101 over a seven-day period. Differentially expressed genes associated with resistance mechanisms, including those involved in receptor-like kinase signaling, were identified as potentially indicative of effector-triggered immunity (ETI) responses. Candidate genes were further evaluated in other cultivars with varying resistance levels to validate their role in defense mechanisms. By elucidating these molecular pathways, the findings provide a foundation for developing molecular markers and resistance breeding strategies, ensuring the sustainability of banana production in the face of this devastating disease.

Materials and methods

Plant materials and sample preparation

‘Khai Pra Ta Bong’ (AAA genome, highly resistant; accession no. HB162) and ‘Hin’ (ABB genome, highly susceptible; accession no. HB241) cultivars [12] were multiplied using tissue culture techniques as described by [13]. Subsequently, the plantlets were transplanted into pots and grown in a controlled environment greenhouse until reaching 20 days of age. Then, the plantlets were transferred to 4-inch pots, and seven days later, the pathogen was inoculated onto the banana plants, following the method outlined [13]. The Rsc strain MY4101 was cultured in a CPG medium at 28 °C for three days. Subsequently, the cultures were used to prepare a suspended inoculum with a concentration of 10⁸ colony-forming units per milliliter (CFU/mL). The inoculum prepared previously was applied by making wounds near the roots. The procedure involved using a cutter with a blade width of 18 mm and length of 100 mm. The blade was pressed vertically into the soil at 2 cm from the base of the test plant, penetrating to a depth of 5 cm, and then pressed on the opposite side. The test plants were watered one day prior to both Rsc and mock inoculations. The prepared inoculum was applied by pouring 10 mL per plant around the wounded root area. The mock inoculation, defined as the control, involved the application of sterile water. Root tissue samples were collected for 12 h, 1 day, and 7 days post-inoculation for subsequent RNA extraction. The susceptible cultivar ‘Hin’ (ABB genome) was inoculated with Rsc in parallel to confirm the success of the Rsc inoculation conditions. Disease symptoms, disease severity scores, and the disease severity index of both cultivars were observed over a 14-day period. The disease severity assay was conducted in triplicates to ensure the reliability of the phenotypic evaluation.

RNA extraction

RNA was extracted from root samples of both the Rsc-inoculated and control groups at three time points after bacterial inoculation using the RNeasy Plant Kit (QIAGEN, Germany). Each time point included three replicates for both groups, resulting in a total of 18 samples. The extraction process involved grinding frozen roots with liquid nitrogen (0.1 g) using a mortar and pestle until a fine green powder was obtained. The powdered material was transferred to a 1.5-mL tube, and 450 µL of RLT buffer was added. The mixture was immediately vortexed to ensure thorough mixing and then transferred to a QIA shredder spin column. Centrifugation was performed at 8000 g at room temperature for 2 min. The Liquid obtained after centrifugation was transferred to a new tube, and half the volume of 96% ethanol was added, mixed, and transferred to a QIA shredder spin column. Centrifugation was then performed at 8000 g at room temperature for 15 s. The Liquid was discarded, and 700 µL of RW1 buffer was added to the column, followed by centrifugation at 8000 g at room temperature for 15 s. The Liquid was discarded, 500 µL of RPE buffer was added to the column, followed by centrifugation at 8000 g at room temperature for 15 s. The Liquid was discarded, and 50 µL of RNase-free water was added to the column. The column was centrifuged at 8000 g at room temperature for 15 s. The eluted RNA was then used for purity and concentration assessment using a NanoDrop™ Lite spectrophotometer (Thermo Fisher Scientific, USA), and the results were further confirmed by gel electrophoresis on a 1% agarose gel. Subsequently, the obtained RNA was used for the next steps in the study.

RNA sequencing and transcriptome analysis

The obtained total RNA from 18 samples (three replicates per group at each time point for both the Rsc-inoculated and control groups) was used to generate RNA Libraries. The nucleotide sequences were determined using the paired-end method with the NovaSeq 6000 system (Illumina, USA) through Getz Healthcare (Thailand). The data output was approximately 6 GB and the quality base Q30 was > 80% or equivalent. This dataset will be further analyzed to investigate the expression patterns of genes. MultiQC [20] was used to create a single report visualizing the output of the qualities assessment from Fastqc [21] across every sample.

The reference transcriptome M. acuminata DH Pahang (version 4.3), obtained from the banana genome hub [22], was indexed and used as a reference transcript for quantifying transcripts by Salmon (version 1.9.0), the alignment-free algorithm [23]. The resulting quantification data were loaded into R (version 4.2.1), and differential expression analysis was conducted using DESeq2 (version 1.42.0) [24]. Results were visualized by generating MA plots and volcano plots. Statistical significance was determined through the Wald test. Differentially expressed genes (DEGs) were identified using a threshold of log2 fold change > 1 and a Benjamini–Hochberg adjusted p-value ≤ 0.05.

Gene ontology enrichment and combined pathway analysis

The single copy predicted genes in each differentially expressed gene set were annotated using BLASTP hits to the NCBI RefSeq plant protein database. The GO enrichment analysis was analyzed using the GO enrichment function in Banana Genome Hub service with criteria as sensitive merging similar GO-term, removing uninformative parents’ node with cutoff value at 0.05, value cut off at 0.1 and maximum GO to display at a plot at ten categories. The pathway analysis was performed using the Combined pathway analysis in OmicsBox version 3.2.2 [25] with a default parameter, which integrates two major public pathway databases: Reactome [26] and KEGG [27]. Reactome provides curated pathways and reactions, including orthologous events across 15 non-human species, while KEGG represents molecular interaction and relational networks across various biological systems and processes.

Measurement of gene expression by quantitative real-time RT-PCR (qRT-PCR)

Candidate genes identified as significantly differentially expressed in the RNA-Seq analysis were evaluated for their expression using qRT-PCR in other resistant banana cultivars reported by [12], including ‘Khai Kasetsart 2’ (AA genome; accession no. HB032) and ‘Pa Pattalung’ (AA genome; accession no. HB222), as well as the moderately resistant cultivar ‘Pisang Papan’ (AAA genome; accession no. HB037) and the highly susceptible cultivar ‘Hin’ (ABB genome; accession no. HB241). Banana plants were multiplied and inoculated with Rsc or water (mock inoculation; control) as described earlier. Total RNA was extracted from root samples at 12, 24, and 168 h post-inoculation with Rsc or mock treatments. The relative expression level and fold change (FC) of each target gene in Rsc-inoculated samples compared to mock-inoculated samples were calculated using the 2^−ΔΔCT method, incorporating PCR efficiency. The housekeeping gene ribosomal protein S (RPS2) was used for normalization. A two-sample t-test was performed to assess the statistical significance of the results between the samples.

Results

Disease response of ‘Hin’ and ‘Khai Pra Ta Bong’ banana cultivars to Ralstonia syzygii subsp. celebesensis (Rsc)

The differential response of the banana cultivars ‘Hin’ (highly susceptible) and ‘Khai Pra Ta Bong’ (highly resistant) to Rsc strain MY4101 was evaluated over 14 days. As shown in Fig. 1a, non-inoculated ‘Hin’ plants remained symptom-free, while Rsc-inoculated plants exhibited wilting and leaf drooping by 7 days after inoculation (DAI), progressing to severe wilting and chlorosis by 14 DAI indicative of high susceptibility. In contrast, ‘Khai Pra Ta Bong’ (Fig. 1b) maintained a healthy phenotype throughout, with both control and inoculated plants showing no visible symptoms, underscoring strong resistance. Quantitative disease assessments in Table 1 support these observations. ‘Hin’ displayed a disease severity score of 1 (20% index) at 7 DAI, escalating to 3.5 (70% index) by 14 DAI. In contrast, ‘Khai Pra Ta Bong’ consistently exhibited a score and index of 0 at all-time points, indicating complete resistance. These findings highlight a clear contrast in disease progression between the two cultivars. ‘Khai Pra Ta Bong’ maintained an average disease severity index (DSI) below 5%, whereas ‘Hin’ surpassed 90% DSI by 14 DAI. Together, the visual and numerical data in Fig. 1; Table 1 emphasize the robust resistance of ‘Khai Pra Ta Bong’, suggesting the presence of strong genetic defense mechanisms warranting further investigation.

Fig. 1.

Disease symptoms of ‘Hin’, highly susceptible (HS) (a), and ‘Khai Pra Ta Bong,’ highly resistant (HR) (b), after inoculation with the Rsc isolate at 1, 7, and 14 days after inoculation (DAI). Cross sections of the stem and root were observed at 14 DAI

Table 1.

Disease severity scores and disease severity index of Rsc- and mock-inoculated ‘Khai Pra Ta bong’ and ‘Hin’ cultivars at 1, 7, and 14 days after inoculation

Day after inoculation (DAI)	‘Khai Pra Ta Bong’ (HR; accession no. HB162)		‘Hin’ (HS; accession no. HB241)
	Disease severity score (0–5)	Disease Severity Index (%)	Disease severity score (0–5)	Disease Severity Index (%)
1	0	0	0	0
7	0	0	1	20
14	0	0	3.5	70

HR highly resistant, HS highly susceptible

Differentially expressed genes (DEGs) in ‘Khai Pra Ta Bong’ banana cultivar in response to Rsc Infection

This study investigates the molecular responses of the resistant banana cultivar ‘Khai Pra Ta Bong’ to infection by Rsc. RNA seq revealed the differential gene expression at each time point, highlighting upregulated and downregulated genes. Genes with |Log2FoldChange| >1 and padj < 0.05 are considered significantly differentially expressed (Fig. 2). The molecular responses associated with resistance to Rsc strain MY4101 were investigated by collecting root tissues from both Rsc- and mock-inoculated ‘Khai Pra Ta Bong’ at 12 h, 1 day (24 h), and 7 days (168 h) post-inoculation for RNA extraction and RNA-seq analysis. Transcriptomic analysis revealed significant differences in gene expression in response to Rsc inoculation. At 12 h, there were 28 upregulated and 13 downregulated DEGs; at 24 h, this increased to 221 upregulated and 32 downregulated genes, demonstrating an active transcriptional response. By 7 days, the DEGs reduced to 6 upregulated and 28 down-regulated, indicating potential resolution or adaptation phases in host-pathogen interaction (Fig. 3; Table 2).

Fig. 2.

Volcano plots showing differentially expressed genes (DEGs) in ‘Khai Pra Ta Bong’ at 12 h, 1 day (24 h), and 7 days (168 h) post-inoculation. The blue and red dots represent significant p-values at 0.05 and 0.01, respectively

Fig. 3.

Gene expression patterns of (a) downregulated DEGs and (b) upregulated DEGs in ‘Khai Pra Ta Bong’ 24 h after Rsc and mock inoculation. The color scale represents relative expression levels, with red indicating downregulated and green denoting upregulated genes

Table 2.

Number of up- and downregulated genes in ‘khai Pra Ta bong’ in response to Rsc inoculation (|Log2FoldChange| >1 and padj < 0.05) at different time points post-inoculation

Time	Upregulated	Down-regulated
12 h.	28	13
24 h.	221	32
168 h.	6	28

Gene ontology (GO) enrichment analysis

These GO enrichment results, Fig. 4, provide an overview of the complex regulatory mechanisms underlying the plant’s response to Rsc infection, including the up-regulation of defense-related pathways and down-regulation of non-essential processes. The size of the dots in each plot represents the gene count (number of DEGs involved in each GO term), while the color gradient reflects the adjusted p-value (padj), indicating the statistical significance of the enrichment. The GO enrichment analysis suggests that in response to inoculation, ‘Khai Pra Ta Bong’ activates genes involved in stress response, metabolism, and structural reinforcement (up-regulated), while down-regulating genes related to nutrient transport and general metabolism. This dynamic gene regulation likely supports defense and adaptation to pathogen stress. The upregulated DEGs highlight (a) Biological Process (BP) that are activated in response to inoculation. Notable enriched terms include nitrogen compound metabolic process, response to salt stress, and response to oxidative stress, suggesting that upregulated genes are involved in metabolic adjustments and stress responses. (b) Molecular Function (MF): the key enriched molecular functions include hydrolase activity, structural constituent of cell wall, and oxidoreductase activity. This implies that enzymes involved in metabolic breakdown and oxidative processes are significantly up-regulated. (c) Cellular Component (CC) the enriched cellular components include ‘cell wall,’ ‘cytoplasmic membrane-bounded vesicle’, and ‘extracellular region’, suggesting that upregulated genes are associated with structural changes and extracellular defense mechanisms. While the downregulated transcripts highligh (d) Biological Process (BP) shows suppressed biological processes such as sulfate transport, RNA metabolic process, and response to stimulus, indicating reduced activity in nutrient transport and general metabolic regulation and (e) Molecular Function (MF) include sulfate transmembrane transporter activity and nucleotide binding, pointing to decreased transport and binding activities, possibly reflecting a metabolic shift under stress conditions.

Fig. 4.

GO enrichment of upregulated DEGs in ‘Khai Pra Ta Bong’ at 24 h post-inoculation: (a) biological process, (b) molecular function, and (c) cellular component. GO enrichment of downregulated DEGs at the same time point: (d) biological process and (e) molecular function

Key immune response processes by DEGs and integration of pathway analysis

These DEGs were categorized into three key immune response processes: pathogen detection, signal transduction, and pathogen elimination, which are listed in Table 3. The dynamic reallocation of plant resources highlighted mechanisms that optimize immune responses through receptor activation, hormonal signaling, and structural defense.

Table 3.

Differentially expressed transcript plays a key response process

Response process category		log2FC	p-adj
Pathogen Detection
Macma4_09_g14360	Leucine-rich repeat extensin-like protein 3	4.09	< 0.0001
Macma4_06_g28280	putative Leucine-rich repeat extensin-like protein 3	1.24	0.0033
Macma4_03_g05590	putative Leucine-rich repeat extensin-like protein 2	2.23	< 0.0001
Macma4_05_g33330	Leucine-rich repeat receptor-like protein kinase PXC2	1.94	0.0256
Macma4_11_g22690	putative Glycine-rich protein DOT1	2.92	0.0288
Signal Transduction Pathways
Macma4_09_g05760	AP2/ERF domain-containing protein	2.38	0.0436
Macma4_04_g38390	Knot1 domain-containing protein	2.59	0.0014
Macma4_10_g19290	Receptor-like serine/threonine-protein kinase At1g78530	−4.13	0.0024
Pathogen Elimination
Macma4_03_g28930	Xyloglucan endotransglucosylase/hydrolase	3.02	0.0075
Macma4_08_g28050	Xyloglucan endotransglucosylase/hydrolase	2.67	< 0.0001
Macma4_04_g03090	Mannan endo-1,4-beta-mannosidase	1.57	0.0025
Macma4_05_g17100	Glucan endo-1,3-beta-glucosidase 11	1.99	0.0433

log2FC; log2 Fold Change

p-adj; adjust p-values, controlling the false discovery rate (FDR)

Pathogen detection

In this study, receptor-like kinases (RLKs) and glycine-rich proteins were significantly upregulated, enhancing the recognition of pathogen-associated molecular patterns (PAMPs) and initiating pattern-triggered immunity (PTI). Effector-triggered immunity (ETI) was further reinforced through the activation of receptor-like proteins (RLPs), including leucine-rich repeat (LRR) proteins and glycine-rich protein DOT1. Notably, both LRR proteins and DOT1 were strongly upregulated in Rsc-inoculated plants compared to mock-treated controls. The resistant cultivar ‘Khai Pra Ta Bong’ exhibited a rapid induction of these immune-related genes within 12–24 h post-inoculation. The identification of glycine-rich proteins and RLKs underscores the cultivar’s robust ability to perceive pathogenic signals, contributing to early immune activation and reinforcing cell wall integrity to prevent pathogen invasion.

Signal transduction pathways

Transcriptomic analysis revealed that effective plant defense involves tightly coordinated signal transduction and transcriptional regulation, with transcription factors playing central roles. In this study, differentially expressed AP2/ERF transcription factors were observed in Rsc-inoculated banana plants, with several genes upregulated as early as 24 h post-inoculation, suggesting their involvement in early immune responses. These factors are known to regulate pattern-triggered immunity (PTI), effector-triggered immunity (ETI), and hormone-mediated pathways, including jasmonic acid (JA) and ethylene signaling. Conversely, certain AP2/ERF genes linked to disease resistance were downregulated, suggesting a dynamic regulatory balance. Additionally, transcriptomic data indicated a strategic reallocation of cellular resources toward immune responses, with concurrent suppression of mitochondrial and general stress-related pathways. Hormone signaling pathways were prominently activated, particularly those related to JA, auxin, and brassinosteroids, contributing to immune regulation and stress mitigation. Upregulation of chorismate biosynthesis genes precursors for salicylic acid (SA) synthesis was also detected within 24 h post-inoculation. Knot1 domain-containing proteins, including plant lectins and antimicrobial peptides, were persistently upregulated and are associated with broad-spectrum antimicrobial activity. Furthermore, genes involved in secondary metabolite biosynthesis, such as phenylpropanoid and flavonoid pathways, were significantly induced, enhancing oxidative stress tolerance and cell wall reinforcement. Gene Ontology (GO) enrichment analysis highlighted elevated hydrolase activity, cell wall modification, and apoplast-associated processes, supporting pathogen degradation and extracellular defense remodeling. Altogether, these findings suggest that banana immune responses to Rsc infection involve early transcriptional reprogramming, hormone crosstalk, antimicrobial compound production, and structural fortification of host tissues.

Pathogen elimination

Structural reinforcement emerged as a primary pathogen elimination strategy. Genes encoding xyloglucan endotransglucosylase hydrolase and mannan endo-1,4-beta-mannosidase were significantly upregulated, highlighting their roles in strengthening physical barriers. Furthermore, candidate genes such as PR8 (class III acidic chitinase), which degrades fungal cell walls, and xyloglucan endotransglucosylase hydrolase, which modifies cell walls, were also upregulated. These genes likely play critical roles in fortifying structural defenses and activating immune signaling. Enhanced expression of glucan endo-1,3-beta-D-glucosidase underscored its function in degrading pathogen cell walls, effectively targeting invaders. The extracellular matrix played a pivotal role, as modifications were observed in apoplast components and secretory vesicles, essential for containing and neutralizing pathogens. Together, these mechanisms reflect a coordinated pathogen-targeting system tailored to suppress infection and Limit pathogen spread. Furthermore, orthologous genes encoding peroxidases were upregulated within 24 h post-inoculation. Peroxidases play a crucial role in plant defense against pathogens by participating in cell signaling following infection. They contribute to cell wall fortification by catalyzing the polymerization of macromolecules, which are subsequently deposited on the extracellular surface, thereby impeding pathogen invasion.

Integration of pathway analysis

Pathway integration highlighted robust immune reprogramming. The upregulation of jasmonic acid biosynthesis, flavonoid biosynthesis, and transcriptional machinery reinforced findings from DEG analysis. Notably, suppression of tryptophan metabolism and gluconeogenesis reflected a deliberate reallocation of resources, favoring immune activation over general metabolic activity. Chromatin remodeling and transcription regulation further underscored the plant’s adaptability to sustained pathogen resistance. Therefore, the immune response in ‘Khai Pra Ta Bong’ demonstrates a highly dynamic and organized defense system. Early pathogen detection through receptor-like kinases and glycine-rich proteins, strategic reprogramming of signaling pathways, and robust pathogen elimination mechanisms underscore its resilience. These insights offer valuable targets for breeding disease-resistant banana cultivars and deepen our understanding of plant-pathogen interactions.

Gene expression analysis by qRT-PCR

Candidate genes significantly upregulated in Rsc-inoculated ‘Khai Pra Ta Bong’, including Macma4_11_g22690 encoding putative glycine-rich protein DOT1, Macma4_01_g21320 (PR8) encoding classIII acidic chitinase, Macma4_08_g28050 encoding xyloglucan endotransglucosylase hydrolase, and Macma4_04_g03090 encoding mannan endo-1,4-beta-mannosidase, were selected to measure their expression in other resistant banana cultivars. The results of the qRT-PCR analysis for selected genes associated with the response to Rsc strain MY4101 in four banana cultivars ‘Khai Kasetsart 2’ (resistant, R), ‘Pisang Papan’ (moderately resistant, MR), ‘Pa Pattalung’ (resistant, R), and ‘Hin’ (highly susceptible, HS) are shown in Fig. 5. The expression levels of five genes were quantified at three time points: 12 h, 24 h, and 168 h (7 days) post-inoculation (HPI). The genes analyzed include putative glycine-rich protein DOT1 (Fig. 5a-d, graph 1), PR8 Class III acidic chitinase (Fig. 5a-d, graph 2), xyloglucan endotransglucosylase/hydrolase (Fig. 5a-d, graph 3), and mannan endo-1,4-beta-mannosidase (Fig. 5a-d, graph 4).

Fig. 5.

Gene expression of candidate genes in blood disease-resistant banana cultivars: (a) ‘Khai Kasetsart 2,’ (b) ‘Pisang Papan,’ (c) ‘Pa Pattalung,’ and (d) the susceptible cultivar ‘Hin’ under mock (green bars) and Rsc inoculation (blue bars) conditions. R: resistant, MR: moderately resistant, HS: highly susceptible. Candidate genes include: (1) Macma4_11_g22690 encoding putative glycine-rich protein DOT1, (2) Macma4_01_g21320 (PR8) encoding Class III acidic chitinase, (3) Macma4_08_g28050 encoding xyloglucan endotransglucosylase/hydrolase, and (4) Macma4_04_g03090 encoding mannan endo-1,4-beta-mannosidase

The putative glycine-rich protein DOT1 gene showed moderate upregulation in the resistant cultivars, with peak expression observed at 24 HPI in ‘Pa Pattalung’ (Fig. 5c, graph 1) and at 168 HPI in ‘Khai Kasetsart 2’ (Fig. 5a, graph 1). The susceptible cultivar ‘Hin’ exhibited upregulation at 12 HPI (Fig. 5d, graph 1), which does not appear to correlate with an effective defense response.

The PR8 Class III acidic chitinase gene was upregulated in the resistant cultivars ‘Khai Kasetsart 2’ (Fig. 5a, graph 2) at 12 and 168 HPI and ‘Pa Pattalung’ (Fig. 5c, graph 2) at 12 HPI. In the moderately resistant cultivar ‘Pisang Papan,’ upregulation was observed at 24 HPI (Fig. 5b, graph 2). However, the susceptible cultivar ‘Hin’ also showed some induction (Fig. 5d, graph 2), aligning with its susceptibility to Rsc.

Xyloglucan endotransglucosylase/hydrolase was upregulated in both resistant and susceptible cultivars (Fig. 5a-d, graph 3), suggesting its role in cell wall modification. However, its function may not be directly linked to defense mechanisms. Similarly, mannan endo-1,4-beta-mannosidase was induced in both resistant and susceptible cultivars (Fig. 5a-d, graph 4), indicating a possible general role in plant responses rather than a resistance-specific function.

Overall, the qRT-PCR results corroborate the RNA-seq data, confirming the differential expression of key defense-related genes in response to Rsc inoculation. However, the distinct expression patterns between resistant and susceptible cultivars were not consistently observed, emphasizing the complexity of these genes’ roles in mediating resistance to bacterial wilt in banana plants.

Discussion

Plants have a well-defined defense system to counter environmental threats and pathogens [28], with two main branches. The first branch uses pattern recognition receptors (PRRs) to detect pathogen-associated molecular patterns (PAMPs), triggering PAMP-triggered immunity (PTI) to limit pathogen growth. The second branch involves resistance (R) genes that recognize pathogen effector molecules through NB-LRR proteins, leading to effector-triggered immunity (ETI) and further defense responses [29]. If pathogens overcome PTI, it leads to effector-triggered susceptibility, but ETI can restore resistance.

In this study, the typical defense response of banana against the BBD pathogen, involving recognition processes, was observed. Receptor-like proteins (RLPs), such as leucine-rich repeat proteins, and glycine-rich protein DOT1, were found to be upregulated in inoculated plants (Rsc-inoculated) compared to mock-treated controls. Notably, the resistant cultivar ‘Khai Pra Ta Bong’ demonstrated a rapid upregulation of these genes within 12–24 h post-inoculation. This suggests that resistant cultivars possess efficient signal transduction mechanisms to activate resistance responses against invading pathogens through multiple layers of defense [30]. Plant receptor leucine-rich repeat proteins and glycine-rich proteins are thought to play a crucial role in cellular stress responses, including both abiotic and biotic stress, as well as in signaling pathways [31, 32]. Additionally, the putative glycine-rich protein (GRP) DOT1 was significantly upregulated in inoculated plants compared to mock-treated controls within 24 h of inoculation. This may be attributed to the role of GRPs in plant defense, particularly their involvement in innate immune activation to protect plants from pathogens. Lin et al. [33]. reported that a class II GRP, such as LsGRP1, might be a potential entry point for pathogens while simultaneously inducing immune mechanisms in monocot species. Moreover, the leucine-rich repeat receptor kinase (LRR-RK) family plays a key role in triggering the production of reactive oxygen species (ROS) via the activation of the NADPH oxidase RBOHD [34]. In this study, peroxidase genes were also upregulated in inoculated plants compared to mock-treated controls within 24 h of inoculation. The increased expression of peroxidase genes may facilitate rapid, localized cell death at the pathogen penetration site, a response strongly associated with disease resistance.

Upon pathogen recognition via RLPs, plants activate downstream signaling networks. These downstream signaling networks are well-known to be regulated by calcium-dependent protein kinase (CDPK) and mitogen-activated protein kinase (MAPK) cascades [35, 36]. For MAPK activations to trigger secondary and later defense responses, they may need to surpass specific thresholds in both duration and intensity and potentially interact differentially or synergistically with calcium signaling. MAPK and CDPK pathways regulate the activity and synthesis of various transcription factors, enzymes, hormones, peptides, and antimicrobial compounds in ways that are both specific and overlapping, thus contributing to a coordinated defense response against multiple pathogens [35, 37–39]. Moreover, the effective defense requires coordinated signal transduction and gene regulation, with transcription factors such as WRKY, bHLH, AP2/ERF, NAC, and MYB playing pivotal roles in this process [28, 29, 40–43]. In this study, the different expression levels of orthologous transcription factors from the AP2/ERF family were observed to be both upregulated and downregulated in inoculated banana plants compared to mock-treated controls. Certain loci of the AP2/ERF genes were significantly up-regulated, with increased expression detected as early as 24 h post-inoculation (Supplementary Data, Table S2). This up-regulation likely contributes to the activation of innate plant immunity, restricting the pathogen at the site of infection. AP2/ERF transcription factors play a critical role in plant immunity by regulating genes associated with pattern-triggered immunity (PTI), effector-triggered immunity (ETI), and hormone-mediated pathways, such as jasmonic acid and ethylene and phytoalexin synthesis [31]. Conversely, the down-regulation of some AP2/ERF genes involved in disease resistance pathways was also observed. This suggests that the resistant banana cultivar exhibited well-coordinated signal transduction and finely tuned gene expression regulation, enabling an effective immune response to the pathogen. This correlates well with the ERF subfamily, specifically Pti4 and Pti5, in tomato plants when challenged by the virulent P. syringae. The Pst-induced phosphorylation increases Pti4 and Pti5 binding to their target sequences in defense-related genes [44]. Similarly, the upregulation of tomato ERFs Pti4, Pti5, and Pti6 was detected in Arabidopsis, inducing defense responses and contributing to resistance against P. syringae [45]. These findings confirm the role of AP2/ERF TFs in regulating disease resistance pathways [46].

In this study, orthologous genes encoding the AP2/ERF transcription factor family were not only upregulated at 24 h post-inoculation, but the gene for a Knot1 domain-containing protein also exhibited upregulation, which persisted through 24 h. Knot1 encompasses a diverse group of proteins, including plant lectins and antimicrobial peptides. These proteins possess a multitude of functions including inhibitory, cytotoxic, antiviral or hormone-like activity. Again, the xyloglucan endotransglucosylase/hydrolase (Macma4_03_g28930 and Macma4_08_g28050) and Mannan endo-1,4-beta-mannosidase (Macma4_04_g03090) were upregulated at 12 h till 24 h. The Xyloglucan endotransglucosylase/hydrolase and mannan endo-1,4-beta-mannosidase families were classified as being enriched in hydrolase activity, specifically hydrolyzing O-glycosyl compounds and exhibiting xyloglucosyl transferase activity. These enzymes are typically associated with the plant’s response to microbial pathogens, including fungi and oomycetes, which secrete various glycoside hydrolases (GHs) onto their cell surfaces and the surrounding extracellular environment [47]. The broad specificity of plant XETs (xyloglucan endotransglucosylase) underscores their crucial roles in the continuous restructuring and remodeling of the cell wall [48]. Therefore, the elevated expression of the genes such as xyloglucan endotransglucosylase/hydrolase (Macma4_03_g28930 and Macma4_08_g28050) and mannan endo-1,4-beta-mannosidase (Macma4_04_g03090) in inoculated banana plants upon exposure to the BBD pathogen suggests their potential roles in regulating critical defense processes and signaling pathways. This is consistent with the findings of [16], who reported that the down-regulation of MAPK (Ma08_g21950) and upregulation of CDPK (Ma06_g31170) gene expression in banana following banana bunchy top virus (BBTV) infection may play a role in modulating critical defense processes through kinase signaling pathways.

Importantly, among the differentially expressed genes (DEGs) identified in this study, several stand out as strong candidates for use in banana resistance breeding. These include the glycine-rich protein DOT1 (Macma4_11_g22690), class III acidic chitinase PR8 (Macma4_01_g21320), xyloglucan endotransglucosylase/hydrolase (Macma4_08_g28050), and mannan endo-1,4-beta-mannosidase (Macma4_04_g03090). Their consistent upregulation in resistant cultivars at early time points suggests that these genes are associated with rapid and effective defense responses. The development of functional markers such as dCAPS for these loci can facilitate marker-assisted selection (MAS) in breeding programs aimed at developing banana cultivars resistant to Rsc. Thus, these genes provide promising molecular targets for future resistance breeding efforts.

Conclusion

This study provides a comprehensive analysis of gene expression dynamics associated with blood disease resistance in banana cultivars, focusing on the resistant ‘Khai Pra Ta Bong’ cultivar. Using RNA-seq analysis, we tracked gene expression changes over a critical seven-day period following inoculation with the Rsc strain MY4101. The findings reveal a distinct temporal pattern of gene activation, with significant upregulation of genes involved in disease resistance mechanisms as early as 12 h post-inoculation. Notably, key molecular processes such as hydrolase activity and xyloglucan endotransglucosylase hydrolase activity were significantly enriched, providing insights into the defense responses employed by the resistant variety. qRT-PCR validation confirmed the differential expression of crucial genes, including PR8 Class III acidic chitinase, putative glycine-rich protein DOT1, xyloglucan endotransglucosylase/hydrolase, and mannan endo-1,4-beta-mannosidase, underscoring their potential as genetic markers for resistance to blood disease in other resistant and moderately resistant cultivars. The observed expression changes in these genes may contribute to resistance; however, further research involving a larger number of resistant and susceptible cultivars is necessary to confirm their functions. Additionally, other potential genes involved in resistance should be explored to provide a more comprehensive understanding of the molecular mechanisms at play. These molecular insights not only enhance our understanding of the defense mechanisms in bananas but also offer practical applications in breeding programs. The identified genes and pathways provide valuable targets for marker-assisted selection, enabling the development of banana varieties with enhanced resistance to wilt disease. This advancement is crucial for sustainable banana production, ensuring crop resilience against this devastating pathogen and contributing to global food security.

Authors’ contributions

TT, PK, RB, and PT initialed a research concept and design; AAT provided clean tissue culture plants used in the experiment; LR performed the experiment and collected data; PT performed the data analysis; TT, AS, PK, and PT interpreted the data; TT and PT wrote the article; AS and PK performed critical revision; ALL authors have read and approved the final manuscript.

Funding

Kasetsart University, FF(KU)3.66.

Data availability

The raw reads generated via Illumina sequencing were deposited in the NCBI SRA database (BioProject ID: PRJNA1163445).

Declarations

Ethics approval and consent to participate

No specific permission is required to collect all the samples described in this study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.