Arabica coffee SBP transcription factor family members respond to brown leaf spot stress by regulating their expression

Abstract

Coffee brown spot is a prevalent foliar fungal disease that seriously threatens the survival and yield of Coffea arabica L. To explore the molecular basis of disease resistance, we conducted a genome-wide identification and analysis of the SBP gene family in C. arabica L. A total of 22 CaSBP genes were identified and classified into three phylogenetic subfamilies, distributed across 12 chromosomes and including nine segmentally duplicated gene pairs. Cis-element analysis revealed a high proportion (55.37%) of hormone-related regulatory elements in CaSBP promoters. Integrated transcriptomic and metabolomic analyses showed that brown spot infection significantly altered the expression of seven CaSBP genes, which was further confirmed by qRT-PCR, accompanied by marked changes in jasmonic acid (JA) and salicylic acid (SA) levels. Functional enrichment analysis indicated that five differentially expressed CaSBP genes were associated with the brassinosteroid (BR) signaling pathway. Together, these results suggest that CaSBP genes may participate in hormone-mediated defense responses during brown spot infection, providing molecular insights for breeding disease-resistant coffee cultivars.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12870-026-08184-8.

Introduction

Coffea arabica L. var. nana, a member of the genus Coffea in the Rubiaceae family, is the most economically important cultivated species within this genus [1]. Owing to its superior cup quality and distinctive flavor attributes, C. arabica L. dominates global coffee production and trade, accounting for approximately 60–65% of worldwide consumption [2]. Its cultivation is typically restricted to specific agroecological conditions, such as high-altitude regions with relatively cool temperatures, which contribute to the development of its complex sensory profile [3]. In recent years, the global coffee market has increasingly emphasized high-quality, single-origin, and specialty coffee products. In line with this trend, China’s Yunnan Province has adopted a “boutique coffee” development strategy aimed at improving bean quality and aligning cultivation practices with international standards for premium Coffea arabica production [4].

As one of the most valuable cash crops worldwide, the yield stability and quality of C. arabica L. are closely associated with the livelihoods of millions of smallholder farmers and the sustainability of the global coffee industry [5]. However, coffee production is severely constrained by biotic stresses, particularly foliar diseases caused by fungal pathogens [6]. Coffee brown leaf spot (CBLS), primarily caused by Cercospora coffeicola, is one of the most widespread foliar diseases affecting C. arabica L. plantations across major coffee-growing regions, including Latin America, Africa, and Southeast Asia [7]. Typical symptoms include circular to irregular necrotic lesions surrounded by chlorotic halos, which may coalesce under severe infection, leading to premature defoliation. This substantially reduces photosynthetic efficiency and ultimately compromises both yield and bean quality [8]. Recent studies have demonstrated that climate change-associated increases in temperature and humidity have further intensified the incidence and severity of CBLS, posing an increasing threat to sustainable coffee production. Although chemical fungicides remain the primary means of disease control, their long-term and intensive use has raised concerns regarding pathogen resistance, environmental contamination, and rising production costs [9].

Increasing evidence from diverse plant species suggests that SQUAMOSA promoter-binding protein (SBP/SPL) transcription factors play important roles in plant growth, development, and stress responses [10]. In several crop species, SBP genes have been implicated in the regulation of defense-related signaling pathways, including responses to fungal and bacterial pathogens [11]. Notably, SBP transcription factors are closely linked to phytohormone signaling networks, particularly jasmonic acid (JA) and salicylic acid (SA), which are central regulators of plant immune responses [12]. Consistent with this, our preliminary metabolomic analyses revealed pronounced alterations in JA and SA levels following brown leaf spot infection in C. arabica L. These observations led us to hypothesize that CaSBP genes may participate in hormone-mediated defense regulation during pathogen challenge, thereby contributing to the brown spot resistance response in coffee.

Transcription factors (TFs) are proteins that bind to specific DNA sequences in regulatory regions to control gene transcription, thereby activating or repressing gene expression. As central regulators of transcriptional networks, TFs play indispensable roles in plant growth and development, metabolism, responses to environmental stresses, and disease resistance [13]. With the rapid development of high-throughput sequencing technologies, whole-genome sequencing has been completed for numerous model plants and major crops, including Arabidopsis thaliana, rice (Oryza sativa), and poplar (Populus spp.), enabling large-scale identification and functional characterization of transcription factor families involved in stress responses [14]. Early studies highlighted the roles of WRKY transcription factors in plant defense regulation [15], followed by investigations into bZIP family members involved in stress and hormone signaling pathways [16]. More recently, MYB transcription factors have been shown to participate in complex regulatory networks governing both biotic and abiotic stress responses [17]. Collectively, these studies have provided important insights into how transcriptional regulatory networks coordinate plant adaptation to adverse environmental conditions and have identified potential genetic targets for improving crop stress tolerance. Among plant-specific transcription factor families, SQUAMOSA promoter-binding protein (SBP) transcription factors are characterized by a highly conserved SBP DNA-binding domain and have been implicated in diverse biological processes. Accumulating evidence indicates that SBP proteins function not only in plant growth and developmental regulation but also in responses to both biotic stresses and abiotic challenges, such as drought, cold, and salinity [18]. Given their involvement in multiple stress-related regulatory pathways, SBP transcription factors represent promising candidates for investigating the molecular mechanisms underlying pathogen defense in coffee and other economically important crops [19].

SQUAMOSA promoter-binding protein (SBP) transcription factors constitute a plant-specific class of transcriptional regulators. The SBP family was first identified in Antirrhinum majus by Klein et al. (1996), who demonstrated that two proteins, AmSBP1 and AmSBP2, specifically bind to the promoter of the SQUAMOSA gene, a key regulator of floral meristem development. This seminal study established both the nomenclature and the conceptual framework for SBP transcription factors. Subsequent studies reported the identification and characterization of SBP genes in a broad range of plant species. Early investigations expanded SBP research to horticultural and crop plants, including grape (Vitis vinifera) [20], birch (Betula alba) [21], and petunia (Petunia hybrida) [22]. This was followed by studies in chili pepper (Capsicum annuum) [23, 24] and tea plant (Camellia sinensis) [23, 24]. More recent studies further extended SBP functional and genomic analyses to additional crop species, including melon (Cucumis melo) [25] and cabbage (Brassica oleracea) [26]. Genome-wide identification of SBP gene families has also been reported in apple (Malus domestica) [27]. Most recently, research in tomato (Solanum lycopersicum) demonstrated that the SBP-box protein SlCNR negatively regulates genes involved in fruit cutin deposition [28]. With the continuous accumulation of high-quality genome assemblies, comparative genomics approaches have increasingly been employed to investigate the phylogenetic relationships, structural conservation, and evolutionary diversification of SBP genes across plant lineages, providing insights into the evolutionary history and functional divergence of this transcription factor family.

A defining feature of SQUAMOSA promoter-binding protein (SBP) transcription factors is the presence of a highly conserved SBP DNA-binding domain of approximately 76 amino acids (aa). This domain is responsible for recognizing and binding specific cis-regulatory elements in the promoter regions of target genes [29]. Structural and functional studies have established that SBP transcription factors are characterized by a highly conserved SBP domain containing two zinc-binding motifs, C3H and C2HC, which are essential for DNA-binding activity and protein stability [30]. A conserved nuclear localization signal (NLS), partially overlapping with the C2HC motif, is required for nuclear import of SBP proteins, as demonstrated by subcellular localization analyses showing that NLS deletion leads to cytoplasmic retention and loss of transcriptional regulatory function [20]. Phylogenetic analyses further indicate that the SBP domain represents a core and evolutionarily conserved signature across plant species. In addition to structural conservation, SBP gene expression is tightly regulated at the post-transcriptional level. In Arabidopsis thaliana, several SBP genes, including SPL9 and SPL15, contain conserved miR156/miR157 target sites in their 3′ untranslated regions, enabling precise temporal and spatial control of gene expression [31]. Together, these multilayered regulatory mechanisms provide a molecular basis for the diverse roles of SBP transcription factors in plant development and stress-responsive processes.

SBP transcription factors regulate plant development, metabolism, and stress responses by binding to GTAC core motifs in the promoters of target genes through their conserved SBP domains. Functional studies in model plants, including Arabidopsis thaliana and Oryza sativa, have demonstrated that specific SBP family members are involved in phytohormone-mediated defense signaling pathways. For example, SPL7 and SPL9 have been reported to respond to jasmonic acid (JA) and brassinosteroid (BR) signaling, thereby contributing to pathogen defense through the regulation of defense-related genes, such as pathogenesis-related proteins, callose synthases, and cell wall biosynthesis-associated genes [32]. Mechanistically, SPL12 has been shown to influence cellulose synthase complex (CSC) assembly via the BR signaling pathway, which may affect cell wall integrity during pathogen challenge. Similarly, SPL7 interacts with JA-dependent signaling to promote callose deposition, reinforcing cell wall-based physical barriers against pathogen invasion [33]. Collectively, these studies suggest that SBP transcription factors function as integrators of hormone signaling and structural defense responses, highlighting the SBP gene family as an important regulatory component in phytohormone-mediated disease resistance.

In this study, we performed a systematic genome-wide analysis of the SBP transcription factor family in Coffea arabica to characterize its potential involvement in responses to brown spot disease. The objectives of this study were to: (i) identify SBP family members in the C. arabica L. genome; (ii) analyze their physicochemical properties, phylogenetic relationships, gene structures, conserved domains, and motif compositions; and (iii) investigate their chromosomal distribution, duplication patterns, and collinearity relationships. In addition, cis-acting regulatory elements in SBP gene promoters were analyzed to provide insights into potential transcriptional regulatory features. Functional annotation was further explored using Gene Ontology (GO) and KEGG pathway enrichment analyses. To assess the possible association between SBP genes and brown spot disease responses, expression patterns of SBP family members were examined under brown spot stress conditions. Collectively, these analyses provide a comprehensive descriptive framework for the SBP transcription factor family in C. arabica L. and establish a basis for future functional studies related to disease response and stress adaptation.

Materials and methods

Data and material acquisition

Coffea arabica genome data: Genome and annotation files for coffea arabica were obtained from the Coffee Genome Hub (https://coffee-genome-hub.southgreen.fr/) [34]. The conserved domain for the SBP transcription factor family (SBP, ID: PF03110) was retrieved from the Pfam database (http://pfam.xfam.org/). Gene IDs for the A. thaliana SBP family were first acquired from the online platform (https://niblrrs.ucdavis.edu/), and corresponding gene data were subsequently downloaded using these IDs from the TAIR database (https://www.arabidopsis.org/). All data related to A. thaliana SBP transcription factors were directly searched for and downloaded from TAIR. A. thaliana genome data: Genome and annotation files for A. thaliana were downloaded from the TAIR website (https://www.arabidopsis.org/). Apple genome data: Genome (Malus domestica v1.0) and annotation files for apple were obtained from the Rosaceae Genome Database (https://www.rosaceae.org/. Initial analysis was hindered by significant inconsistencies in sequence identifiers across the downloaded files. The original dataset comprised 67,523 sequences (FASTA, 67,567 protein entries, and 523,729 annotation features (GFF. To resolve identifier conflicts, we systematically curated all files to ensure cross-referencing integrity. The curation workflow included: 1 Extracting all chromosome and scaffold names from the genome FASTA file to establish a primary reference list; 2 Creating a mapping file using command-line tools (Awk 4.2.0, sed 4.8, and Python 2.7 to reconcile divergent identifiers in the GFF and protein FASTA files with the primary reference. The core criterion was to adopt the simplest and most consistent nomenclature present in the original files (e.g., "Chr1".."Chr22", discarding redundant prefixes or suffixes. 3 Regenerating all downstream files using the curated GFF as the authoritative source. Following curation, the standardized dataset contained 22 chromosome sequences, 67,281 non-redundant protein entries, and a correspondingly synchronized GFF file. This rigorous manual synchronization was a critical prerequisite for achieving robust chromosome-level biological feature visualization.

Identification and physicochemical characterization of the SBP family of transcription factors

In this study, members of the SBP gene family in C. arabica L. were identified using an integrated approach combining HMMER and BLAST. First, the Hidden Markov Model (HMM) profile of the SBP domain (Pfam ID: PF03110) was obtained from the Pfam database and used to scan the C. arabica L. genome using HMMER (v3.3.2) with default parameters (hmmsearch), including the use of the gathering threshold disabled, sequence E-value cutoff set to 1 × e⁻⁵, and domain E-value cutoff set to 1 × e⁻⁵ [35]. Only sequences with significant hits covering the conserved SBP domain were retained. In parallel, BLASTP analysis was performed using 16 known A. thaliana SBP protein sequences as queries to retrieve homologous sequences from the C. arabica L. protein dataset, with an E-value threshold of 1 × 10^–5 and default scoring parameters [36]. Candidate genes obtained from both methods were merged, and redundant sequences were removed. The resulting non-redundant sequences were further validated for the presence of complete SBP structural domains using National Center for Biotechnology Information (NCBI—CD-Search) and the SMART database. Sequences were excluded if they contained incomplete or atypical SBP domains, defined as domains lacking one or both conserved zinc finger motifs or exhibiting less than 80% coverage of the canonical SBP domain length [37]. The final set of genes was named according to their chromosomal locations as CaSBP followed by a serial number. The physicochemical properties of CaSBP proteins—including amino acid length, molecular weight (kDa), theoretical isoelectric point (pI), instability index, aliphatic index, and grand average of hydropathicity (GRAVY)—were predicted using the ExPASy ProtParam online tool. The instability index was used to evaluate protein stability, with values > 40 indicating potentially unstable proteins and values ≤ 40 indicating relative stability. GRAVY values ranged from negative to slightly positive among CaSBP members, where negative values indicate hydrophilic proteins and positive values indicate hydrophobic characteristics, suggesting diversity in protein solubility and interaction potential. The aliphatic index was used as an indicator of thermostability, with higher values reflecting greater stability across temperature variations. Subcellular localization was predicted using the Plant-mPLoc tool. This multistep screening and characterization strategy ensured the accuracy and reliability of SBP family identification in C. arabica L. [38].

In this study, multiple sequence alignment of SBP protein sequences from Coffea arabica L. was performed using the Clustal Omega online tool (https://www.ebi.ac.uk/jdispatcher/msa/clustalo) with default parameters. Pairwise sequence similarity was calculated based on the percentage of amino acid identity derived from the Clustal Omega alignment output. The resulting percentage identity matrix, which reflects sequence similarity rather than evolutionary distance, was exported and manually organized into a matrix format using Microsoft Excel. Based on the processed sequence similarity matrix, a phylogenetic heatmap was generated using TBtools software. In TBtools, the similarity matrix was directly used as input without further transformation, and hierarchical clustering was performed using the default clustering algorithm and distance calculation settings to visualize similarity patterns among SBP family members. This heatmap representation allowed intuitive comparison of homology relationships among SBP proteins. To further elucidate the evolutionary relationship between C. arabica L. and the SBP transcription factor family of the model plant A. thaliana, SBP protein sequences from both species were identified and integrated for comparative analysis based on genomic datasets. Multiple sequence alignment of SBP protein sequences was performed using MEGA11.0 with default parameters to identify conserved structural domains and sequence variations. The initial alignment was subsequently optimized using TBtools, in which low-confidence and poorly aligned regions were automatically trimmed using the default alignment trimming settings, ensuring improved alignment quality and reliability [36]. Based on the trimmed alignment, phylogenetic analysis was conducted using the Maximum Likelihood (ML) method implemented in both MEGA11.0 and IQ-TREE [38]. In IQ-TREE, the optimal amino acid substitution model was determined using ModelFinder with the -m MF option, and model selection was performed based on the Bayesian Information Criterion (BIC). The best-fit model was identified as LG + G4. Phylogenetic tree construction was then carried out using this model with 1,000 ultrafast bootstrap replicates to assess node support, and an appropriate outgroup was specified to root the tree [39]. In MEGA11.0, ML phylogenetic analysis was performed using default parameters, including automatic model selection and standard bootstrap settings, to confirm the robustness of the inferred topology. The resulting phylogenetic trees clearly illustrated evolutionary divergence and subfamily clustering patterns among SBP proteins. All phylogenetic trees were visualized and annotated using the Evolview online platform (http://www.evolgenius.info/evolview/#/treeview) to enhance clarity and presentation quality. Finally, according to previously established classification criteria for A. thaliana SBP proteins, the 22 C. arabica L. SBP proteins were categorized into corresponding subfamilies [40].

Chromosome localization, duplicate gene, and covariance analysis

To investigate the evolutionary associations of SBP transcription factor families in A. thaliana, apple, and coffee, this study was conducted following a systematic procedure. First, genome annotation files (GFF) for apples were collected from the A. thaliana TAIR database and the Rosaceae GDR database, while simultaneously integrating chromosomal localization information of the genomic protein sequences of coffee. Utilizing the integrated annotation data, the 'Gene Location Visualize from GFF' module of TBtools software [36] was employed to accurately extract the chromosomal locations of the target genes and to generate chromosomal localization maps through its visualization capabilities. In the covariance analysis phase, the MCScanX algorithm was utilized to detect homologous gene duplication events among species. Combined with the multigenome comparison function of TBtools, the genomic covariance patterns among members of the SBP family in A. thaliana, apple, and C. arabica L. were systematically analyzed [41], thereby revealing the conserved features and potential functional associations of this TF family throughout species evolution. Finally, calculate Ka, Ks, and Ka/Ks values. The divergence time formula for duplicate gene pairs is: T = Ks/2λ, (λ = 6.5 × 10^–9) [42].

Analysis of gene structure, structural domains, and conserved motifs

To further analyze the structural features and functional element distribution of the SBP gene family in C. arabica L., TBtools was used to examine gene structural characteristics, including exon–intron organization, based on the C. arabica L. genome annotation file (GFF3). In addition, conserved motifs of SBP proteins were identified using the Multiple Em for Motif Elicitation (MEME) online platform (https://meme-suite.org/meme/) [43]. The MEME analysis was performed with the following parameters: the maximum number of motifs was set to 10, the motif width was allowed to range from 6 to 50 amino acids, and the E-value cutoff for motif significance was set to 1 × 10^–5, while other parameters were kept at their default settings. The identified gene structural features and conserved motif distributions were subsequently integrated and visualized using the integrated visualization module of TBtools, generating a comprehensive schematic representation of exon–intron structures and motif compositions. This visualization facilitated the comparative analysis of conserved functional elements and evolutionary patterns among SBP family members.

Cis-acting element analysis

A cis-acting element is a noncoding DNA regulatory sequence located in the flanking region of a gene that regulates transcriptional activity by specifically binding to trans-acting factors, such as transcription factors. Identification of cis-acting elements was performed by extracting the promoter region 2,000 base pairs (bp) upstream of the translation start site (ATG) of each gene, based on genome annotation information using TBtools. Because experimentally validated transcription start site (TSS) information was unavailable for most genes, the annotated ATG was used as a reference point for promoter extraction. The extracted promoter sequences were subsequently submitted to the PlantCARE online analysis platform (https://bioinformatics.psb.ugent.be/webtools/plantcare/html/) to predict cis-regulatory elements and their putative biological functions using established bioinformatics approaches [44].

To address the limitations of database-dependent annotation, manual curation was subsequently performed to specifically search for several well-characterized stress-responsive cis-regulatory elements, including ABRE, DRE/CRT, MYB, MYC,and LTR motifs, based on their reported consensus sequences in previous studies. This combined strategy, integrating PlantCARE-based prediction with manual inspection, ensured a more comprehensive and reliable identification of stress-related cis-acting elements in the promoter regions of the analyzed genes.

GO function and KEGG pathway enrichment analysis of the SBP gene

Gene Ontology (GO) annotation and KEGG pathway enrichment analysis of SBP genes provide an important framework for elucidating their biological functions, including molecular functions, cellular components, biological processes, and associated metabolic networks. In this study, a multidimensional analysis strategy was employed. First, the 22 C. arabica L. CaSBP protein sequences were functionally annotated using the EggNOG-Mapper online platform (http://eggnog-mapper.embl.de/) based on the eggNOG v5.0 database, with orthology inference performed using the fine-grained orthology mode to improve annotation accuracy [45]. Conserved functional domains were identified through direct comparison with homologous orthologous group databases. Subsequently, GO functional classification and KEGG pathway enrichment analyses were conducted using TBtools software. For KEGG enrichment analysis, the entire set of annotated C. arabica L. genes was used as the background gene set. Statistical significance was evaluated using the hypergeometric test, and multiple testing correction was performed using the false discovery rate (FDR) method (Benjamini–Hochberg correction). Pathways with an adjusted FDR value < 0.05 were considered significantly enriched. Finally, bar charts and bubble plots representing the enrichment results were generated using the ChiPlot interactive visualization platform (https://www.chiplot.online/bar_plot_width_category) to facilitate intuitive visualization and interpretation of the functional enrichment patterns [46].

Isolation, Identification, and Inoculation of Cercospora coffeicola

Diseased coffee leaves exhibiting typical brown spot symptoms were collected from multiple locations within a small-leaf coffee (Coffea arabica) plantation in Baoshan City, Yunnan Province, China (Supplementary Fig. S1). Fungal isolation was performed from infected leaf tissues, and a total of eight dominant fungal isolates were obtained and designated as strains A-H. Based on morphological characterization on potato dextrose agar (PDA) medium, only strain H exhibited typical characteristics of the genus Cercospora. Strain H formed circular colonies on PDA plates, initially grayish-white and gradually turning dark gray to black. The colonies showed rapid growth and covered the entire plate surface within 5–7 days at 25 °C. The mycelia were well developed with a floccose appearance, and the reverse side of the colonies was black(Supplementary Fig. S2). For molecular identification, genomic DNA of strain H was extracted and the 5.8S rRNA-ITS region was amplified using universal primers ITS1 and ITS4. A 524-bp PCR product was obtained (Supplementary Fig. S3) and submitted to the NCBI database for BLAST analysis. The ITS sequence showed more than 99% identity with several Cercospora species, including Cercospora coffeicola, C. musigena, C. cf. sigesbeckiae, C. cf. malloti, and C. cf. zinniae. Phylogenetic analysis was conducted using the Maximum Likelihood (ML) method based on ITS sequences of 11 closely related Cercospora species (Supplementary Fig. S4), which demonstrated that strain H clustered most closely with C. coffeicola. Based on combined morphological characteristics and ITS-based phylogenetic analysis, strain H was identified as Cercospora coffeicola and used for subsequent experiments. For inoculum preparation, C. coffeicola strain H was cultured on PDA plates at 25 °C for 5 days. The colony surface was gently washed with 5 mL sterile physiological saline, and the mycelium was scraped and transferred into a 50-mL conical flask containing 30 mL yeast extract beef (YEB) medium. The culture was incubated on a rotary shaker at 25 °C for 48 h. The resulting fungal suspension was diluted with sterile saline, and the spore concentration was determined using a hemocytometer. The final concentration was adjusted to 1 × 10⁶ CFU/mL for inoculation. Spore viability was assessed prior to inoculation using a germination assay. Briefly, 100 μL of the spore suspension was spread onto PDA plates and incubated at 25 °C for 12 h. Spores were considered viable if germ tubes were visible under a light microscope. Only spore suspensions with a germination rate exceeding 90% were used for pathogenicity assays. For pathogenicity testing, 20 μL of the spore suspension (1 × 10⁶ CFU/mL) was evenly applied to the abaxial surface of each fully expanded leaf using a micropipette. Control leaves were treated with an equal volume of sterile physiological saline.

Biostress treatment of brown spot disease in C. arabica L. and sequencing of leaf transcriptome and determination of hormone expression

In this study, young leaves of Yunnan Coffea arabica cuttings cultivated in the greenhouse experimental base of Southwest Forestry University were used as experimental materials. Eighteen uniformly growing, healthy 8-month-old seedlings were selected and maintained under controlled greenhouse conditions, including a light intensity of 300 μmol·m⁻²·s⁻¹, a 14 h/10 h (light/dark) photoperiod, a temperature of 22 ± 2 °C, and relative humidity of 65 ± 5%. The plants were randomly divided into a control group (CK) and a treatment group (T), each consisting of nine seedlings. Each group contained three biological replicates (CK1-CK3 and T1-T3, respectively), with three plants per replicate. Both groups were placed in the same open area and spaced 2 m apart to minimize cross-contamination while ensuring identical environmental conditions. Inoculation was performed at 09:00 a.m. on July 25, 2024. For each plant in the treatment group, three fully expanded leaves were selected and wounded with a sterile needle at approximately 1–2 cm from the left or right leaf margin, creating a single scratch of approximately 2 mm in diameter. A fungal spore suspension adjusted to a concentration of 1 × 10⁶ spores/mL using a hemocytometer was applied to each wound site. After inoculation, transparent humidity covers were used to maintain relative humidity above 90% and to prevent external pathogen intrusion. Disease progression was monitored at 0, 3, 6, and 9 days post-inoculation (dpi). At each time point, one leaf was randomly selected from each seedling, and disease severity was quantitatively evaluated by measuring the diameter of visible lesions (mm) using a digital caliper (Table S1). Leaf samples for transcriptomic and metabolomic analyses were collected only once, at 9 dpi (09:00 a.m.), immediately frozen in liquid nitrogen, and sent to Myvi Metabolic Biology (Wuhan, China) for downstream analyses.

qRT-PCR analysis

To analyze the expression patterns of CaSBP14, CaSBP15, CaSBP16, CaSBP17, CaSBP19, CaSBP20, and CaSBP22 under brown spot stress, fully expanded, healthy leaves of C. arabica L. plants were used as control samples, while leaves from plants inoculated with a Cercospora coffeicola spore suspension (1 × 10⁶ spores mL⁻¹) were used as treated samples. For each treatment, three independent biological replicates were prepared. Control samples were designated as CK1, CK2, and CK3, representing leaves collected from three independently grown, non-inoculated plants, whereas treated samples were designated as T1, T2, and T3, representing leaves collected from three independently inoculated plants subjected to brown spot stress.Total RNA was extracted from coffee leaves using a Plant RNA Extraction Kit (DP432), following the manufacturer’s instructions. First-strand cDNA was synthesized using the FastKing RT Kit (KR116). The resulting cDNA was diluted tenfold and used as the template for quantitative real-time PCR (qRT-PCR). Gene-specific primers were designed using Primer 6.0 and are listed in Table S2. qRT-PCR was performed using a QIAGEN SYBR Green PCR kit on a real-time PCR system under the following thermal cycling conditions: an initial denaturation at 95 °C for 5 min, followed by 40 cycles of denaturation at 95 °C for 10 s, annealing at 58–60 °C for 20 s, and extension at 72 °C for 30 s. A melt-curve analysis was conducted at the end of the amplification program to confirm primer specificity. Each reaction was performed with three technical replicates. The Actin gene was used as the internal reference for normalization, as it has been widely reported to exhibit stable expression across different tissues and stress conditions in coffee and other plant species [47]. Relative gene expression levels were calculated using the 2^−ΔΔCt method. Statistical analyses and graphical visualization were performed using GraphPad Prism 10 software.

Results

Characterization and analysis of the physicochemical properties of the SBP gene family in C. arabica L.

The members of the SBP transcription factor family in C. arabica L. were systematically identified using various bioinformatics methods. Initially, potential SBP gene family members were identified through Blast and hmmsearch searches in the C. arabica L. genome database, based on the conserved motifs of known SBP structural domains. Candidate genes were subsequently screened, and false-positive results, which included incomplete SBP structural domains or redundant sequences, were eliminated through structural domain validation using the Conserved Domain Search from the CDD database and Pfam protein family annotation. This process resulted in the identification of 22 high-confidence SBP genes in C. arabica L. (Fig. 1), which were sequentially named CaSBP1-CaSBP22 based on their chromosomal positions. The physicochemical characterization of the proteins encoded by the candidate genes revealed significant variability in sequence lengths among the family members, ranging from CaSBP7 (137 aa) to CaSBP1 (1099 aa). Notably, CaSBP7 represented the smallest functional unit at 15.45 kDa, while CaSBP10 exhibited the largest molecular weight at 114.58 kDa. The distribution of isoelectric points (pI 5.94–9.34) suggests potential differential interaction properties in various pH microenvironments. Protein stability assessments indicated that the instability index ranged from 47.62 (CaSBP8) to 87.10 (CaSBP17), reflecting the possible requirement for posttranslational modifications in some members. Furthermore, subcellular localization analysis performed using the Plant-mPLoc 2.0 algorithm predicted that all CaSBP family members are likely to localize to the nucleus, which is consistent with their putative roles as transcriptional regulators (Table S3).

Fig. 1.

Physicochemical property analysis of CaSBP genes. The figure is divided into three colors by clustering, Group I, Group II, and Group III. The genes in each grouping represent the same subfamily

Sequence similarity and phylogenetic analysis of the SBPs gene family in C. arabica L.

In this study, the phylogenetic relationships of SBP gene family members in C. arabica L. were analyzed using the ClustalW multiple sequence alignment tool to assess the homology of candidate protein sequences. The phylogenetic clustering revealed that members within the same subfamily exhibited high sequence conservation. Notably, proteins CaSBP2, CaSBP3, CaSBP4, and CaSBP5 clustered within Group III and shared a pairwise sequence similarity of up to 98.89%, accompanied by strong bootstrap support for this clade in the phylogenetic tree (Fig. 2), indicating a high degree of evolutionary conservation among these members. To elucidate the evolutionary relationships of SBP genes across species, a phylogenetic tree was constructed based on the amino acid sequences of 22 SBP proteins from Coffea arabica and 16 SBP proteins from A. thaliana. As shown in Fig. 3A, these proteins were clustered into three major subfamilies (Group I, Group II, and Group III), each supported by bootstrap values greater than 70%, which were used as the primary criterion for defining subfamily boundaries. Comparative analysis revealed that all three subfamilies contained SBP genes from both C. arabica L. and A. thaliana. Specifically, Group III comprised the largest number of members (19 genes), including eleven CaSBP genes (50.0% of the CaSBP family) and eight AtSBP genes (50.0% of the A. thaliana SBP family). Group I and Group II together accounted for the remaining SBP genes, with Group I containing eight members and Group II containing eleven members across the two species (Fig. 3B). Within each subfamily, SBP proteins from C. arabica L. and A. thaliana frequently clustered together into mixed-species subclades with pairwise amino acid sequence similarities exceeding 60% and strong bootstrap support (> 70%), indicating conserved evolutionary relationships between orthologous SBP genes in the two species.

Fig. 2.

Heatmap of sequence similarity analysis of CaSBPs Plotted based on TBtools bioinformatics platform. The figure adopts a continuous color scale to characterize the gradient change of sequence similarity (right scale: 15.06–100%), in which the high similarity region is marked by a warm color system (red), and the low similarity region corresponds to a cold system, which intuitively reflects the highly conserved characteristics of subclade members of Group III and other subclades

Fig. 3.

Phylogenetic tree of A. thaliana and C. arabica L. SBP gene families. A phylogenetic tree plot, constructed by ML method using a total of 1,000 this replicated bootstrap values, with different subgroups denoted by different colors; AtSBP genes denoted by purple pentagrams in the plot, and green triangles denote CaSBP genes, where the genes marked in red are genes with known biological functions. Gray circles indicate self-expansion values between 70 and 90%, and black circles indicate self-expansion values > 90%. B Plot of the number of A. thaliana and C. arabica L. SBP genes in each subfamily, with kiwi indicated in blue and Arabidopsis indicated in pink-purple

Chromosomal localization, duplicate gene analysis, and covariance analysis of the SBP gene in C. arabica L.

The chromosomal localization of the 22 CaSBP genes was analyzed using TBtools based on the C. arabica L. genome annotation (Fig. 4). In C. arabica L., the genome comprises 22 chromosomes derived from an allotetraploid origin, which are designated with alphabetical suffixes “c” and “e” to distinguish homologous chromosomes originating from the two ancestral subgenomes. Accordingly, the chromosomes analyzed in this study included 1c-11c and 1e-11e. The results showed that the CaSBP genes were unevenly distributed across the genome. Among the 22 chromosomes, 10 chromosomes did not contain any CaSBP genes, whereas the remaining 12 chromosomes harbored at least one CaSBP gene. Notably, chromosomes 11c and 6e contained the highest number of CaSBP genes, with three genes each. In addition, chromosomes 2e, 3c, 3e, 5c, 5e, and 6c each contained two CaSBP genes, while the other chromosomes harbored only a single CaSBP gene. Overall, the distribution pattern suggests a non-random chromosomal localization of CaSBP genes in the C. arabica L. genome. Gene duplication refers to the increase in the number of copies of a specific gene or gene segment within a genome and may arise through multiple mechanisms, including whole-gene duplication, segmental duplication, tandem duplication, or whole-genome duplication events [48]. To investigate the expansion patterns of the CaSBP gene family in Coffea arabica, we systematically analyzed the 22 identified CaSBP genes for duplication events based on their chromosomal distribution and syntenic relationships, with particular attention to the presence and characteristics of segmentally duplicated gene pairs. The results revealed nine duplicated gene pairs resulting from segmental duplication within the family, distributed across seven chromosomes, with six pairs (66.7% of the total) concentrated on chr5c and chr5e. This distribution suggests that segmental duplication events are the primary factor driving the expansion of the CaSBP gene family in C. arabica L.

Fig. 4.

Localization and covariance maps of C. arabica L. SBP genes. Gray lines indicate all duplicated genes in the C. arabica L. genome, and colored lines indicate C. arabica L. SBP fragment duplicated genes; heatmaps and line plots are gene densities, with line densities increasing from blue to pink to red, and orange rectangles are chromosomes, with chromosome names shown between each chromosome and gene density

To further investigate the relationship between gene duplication, divergence, and selection pressure within the SBP gene family of Coffea arabica, the nonsynonymous substitution rate (Ka), synonymous substitution rate (Ks), and their ratio (Ka/Ks) were calculated for duplicated gene pairs to evaluate the evolutionary forces acting on this gene family. Previous studies have shown that a Ka/Ks ratio greater than 1 generally indicates positive selection, a ratio equal to 1 suggests neutral selection, and a ratio less than 1 reflects purifying selection. The results showed that all identified duplicated SBP gene pairs exhibited Ka/Ks values lower than 1, indicating that these genes have predominantly been subjected to strong purifying selection during evolution, resulting in functional conservation. In addition, the Ks values of the duplicated gene pairs ranged from 0.011 to 3.812, from which the divergence times were estimated to be approximately 0.85—293 million years ago (Mya) [42]. These findings suggest that the duplication events of the C. arabica L. SBP gene family span both ancient and relatively recent evolutionary periods.

In this study, we analyzed the syntenic relationships of SBP genes among Coffea arabica, A. thaliana, and apple (Malus domestica) (Fig. 5). A total of 18 syntenic gene pairs were identified between C. arabica L. and A. thaliana, whereas 40 syntenic gene pairs were detected between C. arabica L. and apple. This quantitative difference indicates a higher degree of syntenic conservation between C. arabica L. and apple than between C. arabica L. and A. thaliana. Furthermore, MCScanX analysis identified 16 conserved syntenic blocks between the C. arabica L. and A. thaliana genomes, and three CaSBP genes showed significant synteny and sequence homology with functionally characterized SBP genes in A. thaliana.

Fig. 5.

Distribution of covariates in Arabidopsis, C. arabica L. and its close relative, apple. The blue line connects the covariates of C. arabica L., A. thaliana and apple. Gray lines connect the covariates of other genes. At-, Ca-, and Md- denote the chromosomes of A. thaliana, C. arabica L., and apple, respectively

Analysis of gene structure, structural domains, and conserved motifs of the SBP transcription factor family in C. arabica L.

Conserved structural domains play crucial roles in protein interactions and are essential for life processes by regulating cellular physiological activities [49]. In this study, motif characterization of the CaSBP protein family was performed using the MEME online platform (Fig. 6). After filtering based on motif significance, a total of 10 conserved motifs (designated Motif1-Motif10) were retained for subsequent analysis. Among these, motif1, motif2, and motif4 exhibited a high degree of conservation throughout the evolutionary process and were widely distributed among the majority of family members. Combined with phylogenetic analysis (Fig. 3A), we found that, despite belonging to the same Group I subfamily, CaSBP3-5 and CaSBP14-17 demonstrated significant divergence in motif composition, suggesting that differences in phylogenetic branching may influence motif evolution. Notably, the spatial arrangement and combination of conserved motifs varied among different CaSBP proteins. Members of the Group II subfamily, CaSBP10 and CaSBP12, exhibited nearly complete motif compositions, lacking only motif8, indicating a relatively high level of motif conservation within this subfamily. In addition, a duplication of motif9 was observed in the CaSBP6 protein sequence. Overall, differences in motif number and distribution were detected among CaSBP family members, reflecting structural diversity in motif organization.

Fig. 6.

Phylogenetic relationships, gene structures, structural domains and conserved motif distribution of CaSBP transcription factor family genes. Phylogenetic tree analysis of 22 CaSBP transcription factor family genes constructed by MERGA; distribution of conserved motifs of CaSBP, motifs are represented by colored boxes and black lines indicate the relative length of proteins; conserved structural domains of CaSBP genes; and exon–intron structure map of CaSBP genes, with green denoting UTRs, yellow denoting CDSs, and the middle line segment denoting the introns

Differences in the distribution of conserved motifs within the CaSBP gene family may be influenced by variations in gene structure and associated functional domains. For example, CaSBP10 and CaSBP12 contain all identified motifs except motif 8, and these two genes are the only members harboring the Ank_2 structural domain. Furthermore, domain annotation indicated that all 22 CaSBP proteins contain conserved domains belonging to the SBP and ANKYR superfamilies, which are known to play important roles in signal transduction processes [50].

Analysis of gene structure provides insights into sequence variation, exon–intron organization patterns, and the retention and evolution of structural features among different CaSBP transcription factors during their evolutionary history. Based on the genome annotation files, gene structure analysis was performed for 22 CaSBP transcription factors. Untranslated regions (5′ and 3′ UTRs) were included in the gene structure visualization, and were clearly distinguished from coding sequences (CDSs) in the structural diagrams. The results showed a small discrepancy between the number of introns and exons among CaSBP genes in C. arabica L., with intron numbers ranging from 1 to 10 and exon numbers ranging from 1 to 11. The inclusion of UTRs further revealed variation in non-coding regions among family members, which may contribute to regulatory diversity at the transcriptional and post-transcriptional levels.

Sequence Comparison and 3D Structure Prediction of SBP Protein in C. arabica L.

To examine the structural characteristics of Group I SBP proteins, five CaSBP proteins were selected for comparative analysis with three A. thaliana SBP proteins belonging to the same subfamily, focusing on amino acid sequence features and predicted three-dimensional (3D) structures. Sequence alignment showed that all analyzed SBP proteins contained the characteristic SBP domain, including two conserved zinc-binding motifs (Zn1 and Zn2) and a predicted nuclear localization signal (NLS) (Fig. 7A), consistent with previously reported SBP domain architectures [51]. Three-dimensional structures of the selected CaSBP and AtSBP proteins were predicted using the SWISS-MODEL server (Fig. 7B). The resulting models exhibited comparable overall folding patterns among Group I members. The Global Model Quality Estimation (GMQE) scores ranged from 0.64 (CaSBP15) to 0.76 (AtSBP3), indicating moderate to high confidence in the predicted structures. Structural superposition further revealed similar spatial arrangements of the conserved SBP domain among these proteins. In addition, the predicted NLS sequence of AtSBP3 (RRKR), located at the C-terminal region of the SBP domain, was mapped onto the corresponding region of its three-dimensional model (Supplementary Fig. S5). This region showed a conserved structural position among the analyzed Group I SBP proteins.

Fig. 7.

Multiple sequence comparison of the structural domains of proteins SBP in Group I, subfamily and prediction of 3D structures. A Multiple sequence comparison of the structural domains of SBP proteins in subfamily Group I. Yellow, green and blue colors indicate the whole conserved residues, 75% conserved residues and 60% conserved residues, respectively; B 3D structures of proteins in Group I subfamily

Analysis of vital cis-acting elements in the SBP promoter of C. arabica L.

The analysis of cis-acting elements provides insights into tissue-specific expression patterns and potential mechanisms underlying multiple stress responses. Utilizing the PlantCARE database to predict cis-elements within the promoter region of the CaSBP gene, we identified 16 cis-acting elements associated with responses to hormones, light, adversity, and developmental regulation (Fig. 8). The proportions of these elements were 55.37%, 15.96%, 23.13%, and 5.54%, respectively. Key components included auxin-responsive elements, growth hormone-responsive elements, drought-inducible MYB-binding sites, cis-acting elements involved in defense and stress responses, low-temperature-responsive cis-acting elements, cis-acting regulatory elements necessary for anaerobic induction, light-responsive elements, cis-acting regulatory elements for circadian control, as well as elements responsive to gibberellins (GA), abscisic acid, SA, and methyl jasmonate (MeJA)-. Five cis-acting elements related to phytohormone responses were identified: the abscisic acid (ABA)-responsive element, growth hormone (auxin)-responsive element, GA responsive element, MeJA responsive element, and SA responsive element. Notably, MeJA response elements are prevalent among members of the CaSBP gene family, and (JA plays a critical role in regulating plant growth, development, and resistance to pests and diseases [52, 53]. Furthermore, SA serves as a vital signaling molecule for enhancing resistance against a broad spectrum of pathogens and plant diseases; in poplar, it has been demonstrated that SA can coordinate lignin synthesis to bolster plant defense and disease resistance [54]. Cis-acting element analysis of the CaSBP gene promoters revealed the presence of multiple categories of regulatory elements related to hormone responsiveness and stress responses. Among these, hormone-responsive elements accounted for 55.37% of the total identified cis-elements, including predicted binding sites associated with five hormone types: ABA, auxin, GA, MeJA, and SA. In addition to hormone-related elements, the promoter regions also contained cis-elements associated with drought stress, low-temperature stress, hypoxia response, and light responsiveness. The number and distribution of these elements varied among different CaSBP genes. Overall, the cis-element profiles indicate that CaSBP promoters harbor a diverse set of regulatory elements related to hormonal signaling and environmental responsiveness.

Fig. 8.

Cis-acting elements of the CaSBP gene family in C. arabica L.. Different colors represent different promoter cis-acting elements, including the location, type, and number of acting elements

GO function enrichment

In this study, we investigated the potential molecular regulatory roles of SBP genes in C. arabica L. under drought stress using GO functional enrichment analysis. Functional annotation of the 22 CaSBP proteins was performed based on the EggNOG-mapper database and further integrated with transcriptomic data, revealing significant enrichment across the three major GO categories (Fig. 9; Table S4). At the molecular function level, DEGs were significantly enriched in terms related to metal ion binding (GO:0046872), DNA binding (GO:0003677), and nuclear localization (GO:0042025), with ten genes showing upregulated expression and twelve showing downregulated expression. In the biological process category, enriched terms were mainly associated with the regulation of phytohormone metabolism and ion transmembrane transport. The cellular component category was predominantly enriched in the nuclear region and membrane-associated structures. Notably, five CaSBP family members were enriched in the brassinosteroid-related sterol signaling pathway. Among these, three target genes showed functional similarity to A. thaliana AtSBP3, a gene previously reported to possess copper ion-binding activity, as indicated by GO annotation. Consistently, all three target genes were annotated with metal ion-binding functions. These results suggest that SBP genes may contribute to plant adaptation to adverse environmental conditions through coordinated regulation of hormone signaling pathways and cellular ion homeostasis.

Fig. 9.

GO annotations of the SBP gene family in C. arabica L.. The different colors represent the three gene ontology categories, Biological Process (BP), Cellular Component (CC) and Molecular Function (MF). From the inside out, the 1 st circle bar indicates the ratio of the number of differentially significant genes enriched to the same GO term to the total number of genes. The 2nd circle indicates the number of genes enriched to GO term upregulated and downregulated, with green color indicating downregulation and red color indicating upregulation. The 3rd circle heatmap represents the total number of genes enriched on the corresponding GO term. The 4th circle represents the GO number

Analysis of the expression pattern of C. arabica L. under brown spot biotic stress

When plants are exposed to environmental stresses, cellular gene expression programs are dynamically adjusted to support adaptive responses. These processes are coordinated through multilayered gene regulatory networks composed of multiple TF families, which integrate diverse signaling pathways to modulate stress-responsive gene expression [55]. In this study, transcriptome sequencing was employed to investigate the expression patterns of the CaSBP gene family in C. arabica L. under brown spot disease stress. As shown in Figs. 10 and 11 (Table S5), all 22 CaSBP genes were detected as differentially expressed in response to pathogen infection, among which 12 genes were significantly upregulated and 10 genes were downregulated. Based on stringent screening criteria (|log₂FC|≥ 1 and p < 0.05), seven CaSBP genes—CaSBP14, CaSBP15, CaSBP16, CaSBP17, CaSBP19, CaSBP20, and CaSBP22—were identified as significantly differentially expressed. Among these genes, CaSBP14, CaSBP16, and CaSBP20 showed decreased expression levels under brown spot stress, whereas CaSBP15, CaSBP17, CaSBP19, and CaSBP22 exhibited increased expression. This contrasting expression pattern suggests functional diversification within the CaSBP gene family during pathogen infection. Collectively, these results indicate that different CaSBP members may participate in brown spot disease responses through distinct regulatory roles, potentially contributing to both defense activation and the maintenance of cellular homeostasis in C. arabica L. leaves.

Fig. 10.

Comparison between before and after biostressing of brown spot disease in C. arabica L

Fig. 11.

Heatmap of differential gene expression of SBP in C. arabica L. brown spot biostress. The magnitude of the relative content in the left panel is shown by the difference in color, the redder the color the higher the expression and the bluer the lower the expression. Ck is the control group and t is the treatment group; the right panel is the volcano plot of the transcriptome, the x-axis is the fold difference in gene expression (log₂ Fold Change, log₂FC); the y-axis is the statistical significance (-log₁₀(p-value) or -log₁₀(adjusted p-value); the higher the value, the higher the statistical significance. log₂FC); Y-axis is the statistical significance of differential expression (-log₁₀(p-value) or -log₁₀(adjusted p-value) The higher the value, the stronger the significance (the smaller the p-value.). Red circles indicate differentially significant upregulated genes; green squares indicate differentially significant downregulated genes; light blue diamonds indicate differentially expressed genes

We validated seven differentially significant genes (CaSBP14, CaSBP15, CaSBP16, CaSBP17, CaSBP19, CaSBP20, and CaSBP22) using quantitative reverse transcription polymerase chain reaction (qRT-PCR). The results indicated (Fig. 12, Table S6) that the expression of CaSBP15, CaSBP17, CaSBP19, and CaSBP22 was upregulated, while CaSBP14, CaSBP16, and CaSBP20 were downregulated under brown spot stress (9 d) compared to the control (CK). The expression trends of the genes measured by qRT-PCR were generally consistent with those obtained from RNA sequencing analysis (R² = 0.9616). These findings suggest that the RNA-seq data are reliable, providing further confirmation of the accuracy of the transcriptome data through qRT-PCR, with a high correlation between the results and the transcriptome data. The reliability of the differential genes identified in the transcriptome data was thus validated, demonstrating that these genes are key contributors to resistance against brown spot disease stress in C. arabica L..

Fig. 12.

qRT-PCR validation of seven CaSBP differentially significant genes in C. arabica L. all data were averaged using three biological replicates, and the error line indicates the standard deviation of the three biological replicates

Differentially abundant metabolite analysis

Phytohormones, as essential regulators of plant responses to biotic stresses, play a significant role in the resistance mechanisms against brown spot disease in C. arabica L. [56]. To analyze the dynamic regulatory network of hormones and related metabolites under this stress, the present study systematically determined the changes in the contents of three major stress-related metabolites: hormones, flavonoids, and sugars, in C. arabica L. plants subjected to brown spot disease stress (9 days) compared to the CK using targeted secondary metabolomics techniques. The results indicated (Fig. 13B, Table S7) that ten phytohormones and their derivatives exhibited significant differential accumulation patterns in the treatment groups, including ABA, JA, and its active binding form jasmonic acid-isoleucine (JA-Ile), the metabolic intermediate 4,5-dihydro-9,10-dihydrojasmonic acid (4,5-DH-9,10-DH-JA), SA, 4-aminosalicylic acid (PAS), chickpeanin A (BCA), dihydrobaicalein (DHB), DL-glyceric acid (DL-GA), and maltose monohydrate (Mal-H₂O). These levels responded synergistically to pathogen infestation through specific up or downregulation. The principal components analysis (PCA) (Fig. 13A) further revealed significant differentiation characteristics in the metabolic profiles between samples. The cumulative contribution rate of the first two principal components (PCA1 = 95.70%, PCA2 = 2.20%) reached 97.90%, indicating that the two-dimensional space has fully captured the core variation information of the metabolic group, and the distance between groups is positively correlated with the degree of metabolic difference, thereby demonstrating the metabolic reprogramming effect induced by stress.

Fig. 13.

PCA and heatmap analysis for the identification of metabolites in C. arabica L.. A PCA scoring plot of 10 phytohormones and derivatives. Different colors represent different experimental groups, and the horizontal and vertical coordinates indicate the first principal component PC1 and the second principal component PC2, respectively. B Heatmap of 10 metabolites in C. arabica L.. Horizontal coordinates represent sample names, vertical coordinates represent metabolite information, and different colors represent different trends of metabolites, respectively

Transcriptome KEGG enrichment and joint analysis with metabolites

KEGG pathway enrichment analysis based on transcriptomic data was performed to investigate the molecular response to brown spot biotic stress. A total of seven differentially expressed genes were significantly enriched in the BR and SA signaling pathways (Fig. 14). Within the BR signaling pathway, three genes—CaSBP15, CaSBP19, and CaSBP22—were significantly upregulated under stress conditions. The expression level of CaSBP15 increased from an FPKM value of 2.37 in the control group to 7.62 in the treatment group, corresponding to a log₂FC of 1.69 (p < 0.05). Similarly, CaSBP19 showed an increase in expression from 0.30 to 0.78 (log₂FC = 1.38, p < 0.05), while CaSBP22 increased from 0.42 to 1.27 (log₂FC = 1.60, p < 0.05). These results indicate a consistent transcriptional upregulation of BR related genes in response to brown spot stress. In contrast, no significant accumulation of BR related metabolites was detected in the metabolomic analysis. In the JA signaling pathway, both JA and its bioactive conjugate JA-Ile exhibited decreased metabolite levels under stress conditions. Additionally, genes associated with the SA signaling pathway were significantly enriched at the transcriptomic level.

Fig. 14.

KEGG signaling pathway of BR, JA, and SA in conjunction with differentially significant genes. The red dashed box shows the heat map of differentially significant gene expression and the green dashed box shows the heat map of metabolite content

Discussion

Coffee is one of the world’s three most widely consumed beverages and is valued for its distinctive flavor and bioactive properties [1]. As a major commercial crop, C. arabica L. is cultivated in more than 80 countries, with a global planting area of approximately 10.6 million hectares [57]. In China, C. arabica L. is predominantly grown in Yunnan Province, which represents the country’s principal coffee-producing region [58]. Recent statistics indicate that the planting area of C. arabica L. in Yunnan has stabilized at over 1.7 million mu and continues to increase annually. In 2024, exports of Yunnan-grown C. arabica L. reached 32,500 tons, representing a year-on-year increase of 358% [53], with products exported to 29 countries and regions, including the Netherlands and Germany [59]. In recent years, climate change has substantially altered environmental conditions in major C. arabica L. production regions worldwide. Rising temperatures and increased humidity have intensified disease pressure, particularly from fungal pathogens, thereby posing significant challenges to coffee yield stability and production sustainability. Among these diseases, brown spot has emerged as a major constraint affecting the growth and productivity of C. arabica L. [7]. Consequently, there is an urgent need to select new varieties of C. arabica L. that exhibit resistance to disease stress, to investigate disease resistance genes, and to elucidate the mechanisms underlying disease resistance. Phytohormones are pivotal in regulating plant responses to disease resistance and orchestrating defense mechanisms through a complex signaling network [60]. Various hormones mediate specific responses to different types of pathogens (e.g., biotrophic or necrotrophic) and dynamically modulate disease resistance strategies through synergistic or antagonistic interactions. SBP genes contribute to biotic stress responses by participating in the hormone-signaling pathways of plants [61]. In this study, we systematically identified the CaSBP gene family in C. arabica L. and explored its expression patterns under brown spot disease stress. Additionally, we examined the changes in plant hormone levels in response to pathogen stress, aiming to establish a theoretical foundation for disease-resistance breeding in C. arabica L. Genome-wide analysis revealed 22 CaSBP gene members. A comparative analysis with jujube (16) [62], sweet orange (15) [63], passion fruit (14) [51], loquat (28) [40], and A. thaliana (16) [10] indicated that the size of this TF family correlates with the genome size of the respective species. Phylogenetic analysis classified the family into three subfamilies. Notably, the CaSBP genes in the Group I subfamily clustered closely with the A. thaliana SBP genes, suggesting potential unique biological functions. In contrast, the largest number of Group III members formed a tightly knit cluster with other CaSBP genes and homologous genes from the model plant A. thaliana, indicating that this subfamily has retained a highly conserved character [64]. Furthermore, the validity of this classification was corroborated by analyses of conserved motifs, gene structures, and structural domains. Five members of the C. arabica L. SBP gene family are enriched in the Brassinosteroid signaling pathway, three of which belong to the Group I subclade. AtSBP3 plays a biological role in regulating the activity of copper ion transmembrane transport proteins and inorganic cation transmembrane transporters [65]. Copper is essential for plant cells to maintain crucial processes, including photosynthesis, reactive oxygen species (ROS) scavenging, cell wall lignification, and hormone sensing. Specifically, copper ions scavenge ROS and contribute to cell wall lignification and biological functions that align with those of BRs. The genes CaSBP14, CaSBP15, and CaSBP16, which belong to the same subfamily, typically exhibit high similarity in their biological functions. Additionally, AtSBP7 is involved in flowering, cell differentiation, regulation of floral development, and disease resistance [66]. Therefore, it is consistent that genes within the same subfamily usually share similarities in their biological functions. However, this similarity is not absolute and must be analyzed within the context of sequence similarity, structural features, and gene homology.

Phylogenetic clustering analysis revealed a clear association between conserved motif composition and subfamily classification within the CaSBP gene family. Genes sharing similar motif architectures tended to cluster within the same evolutionary branches, supporting the reliability of the subfamily delineation and indicating structural consistency among closely related members [67]. Several CaSBP proteins displayed modular motif arrangements with highly similar combinatorial patterns, suggesting conserved structural organization within specific clades. Analysis of conserved motifs showed that motif1, motif2, and motif4 were widely distributed across most CaSBP members, indicating that these motifs represent core structural components of the SBP family. Their prevalence across different subfamilies suggests that they have been strongly conserved during evolution and are likely associated with fundamental SBP protein functions. In addition, gene structure analysis revealed considerable structural diversity among CaSBP genes, with intron numbers ranging from 1 to 10 and exon numbers varying from 1 to 11, reflecting structural differentiation within the gene family. The observed structural diversity among CaSBP genes is reflected in their exon–intron organization. Members of this family exhibited variable gene structures, ranging from relatively simple architectures with fewer introns to more complex structures containing multiple exons and introns. Such variation indicates structural differentiation within the CaSBP gene family. In addition, differences in exon number among CaSBP genes suggest the potential for alternative splicing, which has been widely reported in plant transcription factor families and is known to increase transcript and protein diversity under different conditions. Structural domain analysis further showed that CaSBP10 and CaSBP12 contain two identical Ank_2 domains and cluster within the same subfamily and phylogenetic branch. These two genes also display complete conserved motif compositions compared with other family members. The shared domain architecture and motif profiles indicate a high degree of structural conservation between CaSBP10 and CaSBP12, suggesting that they may have originated from a common ancestral gene. The presence of Ank_2 domains, together with conserved SBP motifs, highlights the structural complexity of these two proteins within the CaSBP family.

Genomic studies have shown that gene duplication is widespread in plant genomes, and a substantial proportion of annotated genes retain at least one duplicated copy [68]. Gene duplication is therefore considered an important component of genome evolution and a major source of genetic redundancy. In this study, phylogenetic and chromosomal analyses identified nine duplicated gene pairs within the C. arabica L. SBP gene family, accounting for 40.9% (9/22) of all CaSBP genes. These duplicated genes were primarily classified as tandem duplication events based on their physical proximity on chromosomes. The occurrence of multiple duplicated pairs indicates that duplication has contributed to the expansion of the CaSBP gene family. Chromosomal localization further showed that chromosome 3 contained the highest number of duplicated CaSBP genes, while chromosomes 5c and 5e together harbored 66.7% of the duplicated gene pairs. The non-random distribution of duplicated genes across chromosomes suggests that duplication events within the CaSBP family are unevenly distributed in the C. arabica L. genome. Further analyses will be required to determine whether these duplicated genes have undergone structural or functional divergence. The substantially higher number of syntenic SBP gene pairs identified between C. arabica L. and apple compared with A. thaliana suggests that, at the genome collinearity level, C. arabica L. shares a closer evolutionary relationship with apple. This pattern may reflect differences in lineage-specific genome evolution and whole-genome duplication histories among these species.

Cis-regulatory elements in the promoter region are crucial for the transcriptional regulation of genes, particularly in plant responses to environmental stresses [69]. The activation of resistance genes relies on the functional initiation of these elements [70]. In this study, we analyzed the 2000 bp regulatory region at the 5' end of the C. arabica L. CaSBP gene and identified several functional elements associated with stress tolerance, primarily including light signaling elements, phytohormone response elements, and adversity stress response elements. Quantitative analysis revealed that hormone-related elements constituted 55.37% of the total number of elements, with ABA and MeJA response sites being the most densely distributed, followed by GA and SA response elements (Fig. 8). Molecular mechanistic studies indicate that ABA and MeJA form a synergistic regulatory network in response to biotic stress: the former enhances physical defenses by promoting callose deposition in vascular tissues, while the latter activates secondary metabolite biosynthesis pathways, such as terpenes, thereby establishing a multilayered resistance system [71]. Notably, KEGG pathway enrichment analysis demonstrated significant clustering of DEGs in the plant hormone signal transduction (PST) pathway (Fig. 14), confirming from a systems biology perspective that the CaSBP TF may enhance plant disease resistance by regulating the dynamic balance of endogenous hormones and their signaling networks. These findings provide a new theoretical basis for elucidating the molecular mechanisms of the SBP gene family in brown spot resistance in C. arabica L.

Phytohormones, as key signaling molecules, are intricately involved in the defense response to biotic stresses, such as the C. arabica L. brown spot pathogen, and play a central regulatory role in the adaptation to abiotic adversities, including drought and salt damage (Zhang et al. 2025). They coordinate plant growth and development with resistance responses through a dynamic equilibrium mechanism [72]. This bidirectional regulation enables the hormone signaling network to precisely adjust plant physiology in response to environmental changes, activating specific defense pathways while maintaining basal metabolism [73]. In this study, we determined the hormone contents of C. arabica L. seedlings subjected to brown spot stress. The levels of ABA and JA were significantly downregulated compared to the control group. The notable downregulation of ABA may result from the activation of ABA-degrading enzymes, such as ABA 8'-hydroxylase, by the pathogen or the secretion of ABA catabolizing metabolites, which subsequently reduces ABA levels [74]. The duration of brown spot stress in this experiment was 9 days, which is considered a prolonged stress period; JA may be transiently upregulated at the onset of infestation, but its synthesis is inhibited as the pathogen colonizes.

BRs, a class of polyhydroxysteroidal plant growth regulators, play a pivotal role in plant growth, development, and stress responses [75]. In this study, we demonstrate that infestation by the brown spot pathogen specifically induces a cascade response within the BR signaling pathway (Fig. 15). The BR molecule initiates reciprocal phosphorylation of the kinase structural domains in the plasma membrane receptor BRI1 and its coreceptor BAK1 by recognizing the extracellular structural domains of these kinases. CaSBP (14, 15, 16, 19, 22) regulates downstream target genes by binding to BRI1, a crucial component of the BR signaling pathway. This phosphorylation event induces a conformational change in the downstream signaling factor BSK1, which subsequently activates the phosphatase BSU1 through protein interactions. The activated BSU1 inhibits the activity of the kinase BIN2 via dephosphorylation, thereby alleviating its phosphorylated modification of the transcriptional regulators BZR1/BES1. In this process, part of the phosphorylated form of BZR1/BES1 is degraded via the ubiquitination pathway, while the unmodified BZR1/BES1 is translocated to the nucleus to regulate transcription by binding to the promoter regions of target genes. Notably, BR signaling directly regulates the expression of CESA genes through intranuclear BZR1/BES1. The translated CESA protein is then transported to the plasma membrane via the endoplasmic reticulum-Golgi pathway, where it assembles into a dynamic hexameric complex structure known as the Cellulose Synthase Complex (CSC). This complex moves directionally along the cytoskeleton, catalyzing the extension of β−1,4-glucan chains and ultimately leading to the orderly deposition of newly synthesized cellulose microfilaments into the cell wall matrix. This mechanism of cell wall strengthening, mediated by BR signaling, significantly enhances the mechanical strength of the plant cell wall and creates an important physical barrier against invasion by pathogenic bacteria. This study reveals a novel mechanism by which BR signaling is involved in plant immunity through the precise regulation of cell wall biosynthesis, providing a theoretical basis for analyzing plant-pathogen interactions.

Fig. 15.

Mechanism of brown spot resistance by interaction of rapeseed sterols with jasmonic acid BRI: Brassinosteroid Insensitive 1; BAK1:BRI1-Associated Receptor Kinase 1; BSK1:BR-Signaling Kinase 1; BSU1:BRI1 Suppressor 1; BIN2: Brassinosteroid Insensitive 2; BZR1: Brassinazole Resistant 1; BES1:BRI1-EMS Suppressor 1; JA-Ile: jasminol isoleucine; JAR1: jasmonate-aminoacid synthase; ROS: reactive oxygen species; CESA: cellulose synthase; CSC: cellulose synthase complex

JA, a class of endogenous signaling molecules, coordinates the allocation of resources between plant growth and development and stress responses by dynamically regulating gene expression networks [52, 53]. In this study, we found that brown spot pathogen infection induced a chloroplast membrane lipid metabolism pathway that contributed to the substantial synthesis of linolenic acid-derived JA precursors. The newly synthesized JA molecules are catalyzed by jasmonic acid amino ligase (JAR1) in the cytoplasm and covalently bind to isoleucine to form the biologically active JA-isoleucine complex (JA-Ile) [76]. This active signaling molecule interacts conformationally with the nucleus-localized receptor module COI1-JAZ through spatially specific recognition, which triggers the ubiquitination modification of JAZ repressor proteins and their degradation via the 26S proteasome pathway. Consequently, this process results in the release of MYC2-like transcription factors originally bound by JAZ and initiates the transcription of downstream defense-related genes. MYC2 transcription factors induce callose deposition in the nucleus (activating the CALS gene), which works in concert with BR to form a physical barrier against pathogen invasion.

BR and JA interact under C. arabica L. brown spot stress to coregulate plant growth and defense responses. BZR1/BES1 and MYC2 coactivate pathogenesis-related (PR) proteins, such as chitinase, and both are involved in the synthesis of secondary metabolite genes, including enzymes of the phenylpropanoid pathway, within the nucleus. This collaboration enhances the plant's chemical defenses. While BR promotes cell elongation through the activation of BZR1/BES1, it works synergistically with JA to boost the expression of defense genes under pathogen stress, thereby preventing susceptibility that may arise from overgrowth. Furthermore, molecular mechanism analyses of pathway interactions reveal that BR regulates reactive oxygen species (ROS) levels by modulating antioxidant enzymes, such as superoxide dismutase (SOD, EC1.15.1.1) and catalase (CAT, EC1.11.1.6). In contrast, JA induces a ROS burst to strengthen pathogen resistance, thus synergistically preventing oxidative damage and sustaining immune intensity [77].

In this study, we systematically analyzed the function of the SBP transcription factor family in C. arabica L. in response to biotic stress caused by brown spot disease. Our findings revealed that five members of this family—CaSBP14, CaSBP15, CaSBP16, CaSBP19, and CaSBP22—exhibited significant changes in expression during brown spot infestation. Detailed analysis indicated that these genes function as upstream regulatory elements, activating the expression of cell wall synthesis-related genes, such as CESA, by engaging in the BR signaling pathway cascade. This activation enhances the mechanical strength of the cell wall, thereby providing resistance against pathogenic bacterial invasion. Based on the observed gene expression patterns and phenotypic correlations, these CaSBP members can serve as targets for molecular marker-assisted screening or gene editing, offering genetic resources for the development of new coffee varieties with high resistance to brown spot disease. Additionally, our promoter cis-element analysis revealed that the regulatory regions of certain CaSBP genes, such as CaSBP15 and CaSBP19, contain drought-responsive elements, including MYB-binding sites, highlighting their potential role beyond biotic stress response. Combining the mechanism of coordinating water use efficiency (WUE) via BR signaling in SBP family members, such as A. thaliana SPL9 [78], we propose that C. arabica L. SBP genes may synergistically respond to both biotic and abiotic stresses through interactions within the BR-JA signaling network. This network plays a dual role in cell wall remodeling and the regulation of stomatal movements. This finding offers a novel perspective for analyzing the mechanisms underlying multistress resistance in coffee. Future studies will focus on verifying the molecular regulatory pathways of CaSBP genes under drought conditions and their interactions with disease resistance pathways.

Conclusion

In this study, a total of 22 SBP transcription factor genes were systematically identified from the Coffea arabica genome and classified into three evolutionary subfamilies based on phylogenetic analysis. Chromosomal localization, duplication patterns, and comparative analyses with A. thaliana revealed the evolutionary conservation and diversification of CaSBP genes. Expression profiling under brown spot infection demonstrated that seven CaSBP genes responded significantly to pathogen stress, and qRT-PCR validation confirmed the reliability of the transcriptomic data. Integrated transcriptomic and metabolomic analyses further indicated that brown spot infection was accompanied by pronounced changes in phytohormone levels, including JA, SA, and ABA, suggesting the involvement of hormone-mediated regulatory networks in disease responses. Functional annotation and enrichment analyses showed that five differentially expressed CaSBP genes were significantly associated with the BR signaling pathway. Based on phylogenetic relationships, expression patterns, and functional annotation, three genes—CaSBP14, CaSBP15, and CaSBP16—were identified as key candidate regulators potentially involved in brown spot resistance. Overall, this work provides the first comprehensive characterization of the SBP TF family in C. arabica L. and offers new insights into their potential roles in hormone-related defense responses during pathogen infection. These findings establish a valuable molecular framework for future functional studies and provide a theoretical basis for the genetic improvement and breeding of brown spot resistant coffee cultivars.

Authors’ contributions

Hailin Ren: Conceptualization, Investigation, Formal analysis, Writing-original draft preparation, and Software. Ke Zhao: Conceptualization, Writing-original draft preparation, and Software. Mengsi Duan: Methodology, Formal analysis, and Visualization. Juejuan Zhu: Experimental material and Field investigation. Xiaozhen Liu: Reviewing and editing, Supervision, Funding acquisition. Hanyao Zhang: Methodology, Writing-review and editing, Funding acquisition. All the authors have read and approved the final manuscript.

Funding

This study is supported by the Yunnan Academician (Expert) Workstation Project (202305AF150020), the Yunnan First-class Construction Discipline of Forestry Science of Southwest Forestry University and the Start-up Fund Project of Doctoral Research in Southwest Forestry University, the China Agriculture Research System (CARS-21), and the China Agriculture Research System (CARS-21-05B). The funders had no role in the design of the study, the collection, analysis, or interpretation of the data, or the writing of the manuscript.

Data availability

The C. arabica L. genome data were obtained from the Coffee Genome Hub (https://coffee-genome-hub.southgreen.fr/). RNA-seq data related to Cercospora coffeicola stress can be accessed via the following link: https://www.ncbi.nlm.nih.gov/bioproject/?term=PRJNA1206329. Additionally, protein sequences and genomic data for the 22 SBP gene families in A. thaliana were downloaded from the TAIR website (https://www.Arabidopsis.org/). The genomic data for Malus domestica were retrieved from the GDR website (https://www.rosaceae.org/).

Declarations

Ethics approval and consent to participate

Plant material of coffea arabica used in this study was collected from the greenhouse facilities of Southwest Forestry University, which are owned and managed by the institution for research purposes. All sampling was conducted in compliance with the institution’s internal research guidelines, and no specific permissions or licenses were required as the collection was carried out within the institution’s own controlled research environment for scientific investigation.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.