Introduction

The interaction between sap-sucking insects and plants is often described as a co-evolutionary arms race. In this dynamic relationship, insects continuously optimize resource acquisition through the secretion of salivary effectors, whereas plants evolve a diverse array of defense mechanisms to deter herbivory [1, 2]. A critical early step in this defensive response is the recognition of herbivory-induced damage. This recognition occurs through the perception of herbivore- or damage-associated molecular patterns by pattern recognition receptors located on the plant cell membrane, triggering a basal defense response known as pattern-triggered immunity (PTI) [1, 3]. PTI encompasses a series of interconnected responses, including reactive oxygen species accumulation, mitogen-activated protein kinase cascade activation, phytohormone signaling pathway modulation, and extensive transcriptional reprogramming. For instance, infestation by the green peach aphid, Myzus persicae, induces the expression of genes and transcription factors involved in the triterpenoid biosynthesis pathway, which in turn promotes triterpenoid accumulation in peach and enhances plant resistance [4]. Similarly, transcript levels of Bphi008a are upregulated in rice upon infestation by the brown planthopper, Nilaparvata lugens, thereby contributing to the activation of defense responses through the MAPK signaling cascade [5]. These studies establish that the transcriptional activation of defense-related genes is a well-characterized component of the plant PTI response against insect herbivores.

Accumulating evidence indicates that post-transcriptional regulation plays a crucial role in enhancing plant resistance to biotic stress [6,7,8]. Among these regulatory mechanisms, alternative splicing (AS) is a common process in eukaryotes that generates multiple messenger RNA (mRNA) variants from a single pre-mRNA, thereby increasing transcript complexity and protein diversity [9]. AS is highly prevalent in plants; for example, more than 60% of intron-containing genes in Arabidopsis thaliana and approximately 66% in Nicotiana attenuata undergo AS [6, 10]. This process contributes to various physiological functions, including growth and development. Recent studies have further highlighted the importance of AS in plant defense against pathogens and insects, primarily by modulating the expression of pattern recognition receptor genes, resistance (R) genes, genes involved in hormone signaling pathways, and those associated with PTI and effector-triggered immunity [7, 9, 11]. For instance, in cotton, the splicing factor GauSR45a is upregulated and alternatively spliced upon Verticillium dahliae infection; it binds to the GAAGA motif commonly found in the pre-mRNA of genes essential for PTI and effector-triggered immunity, including BRI1-associated kinase 1, thereby enhancing cotton resistance to V. dahliae [12].

The Asian citrus psyllid, Diaphorina citri Kuwayama (Hemiptera: Liviidae), is the most efficient insect vector of Candidatus Liberibacter asiaticus (CLas), a bacterial pathogen that causes huanglongbing, a devastating disease that affects citrus plants [13, 14]. The control of D. citri primarily depends on pesticide application. However, the development of insecticide resistance in this pest underscores the urgent need for alternative management approaches [15]. This necessitates a more detailed molecular understanding of the interactions between D. citri and its host plants. To date, several salivary effector proteins involved in these interactions, such as DcE1 and DcE13, have been identified as suppressors of plant defense mechanisms, enabling D. citri to successfully feed on the phloem [16, 17]. Nevertheless, citrus genes that respond to D. citri infestation, particularly those conferring natural resistance and recognizing herbivore-associated molecular patterns, remain largely uncharacterized.

Therefore, in this study, we characterized the transcriptional and post-transcriptional defense responses in citrus following infestation by D. citri. Specifically, we: (1) identified differentially expressed genes (DEGs) in citrus shoots at 24 and 48 h post-infestation using RNA sequencing (RNA-seq); (2) performed differential alternative splicing (DAS) analysis to detect differentially spliced genes (DSGs) in infested shoots; and (3) conducted an integrated analysis of DEGs and DSGs to reveal the coordinated transcriptional and splicing changes induced by D. citri. Finally, we validated the expression of selected jasmonic acid (JA) related genes to infer their potential roles in citrus immune defense mechanisms.

Materials and methods

Plant and insect rearing

Newhall navel orange (Citrus sinensis Osbeck cv. Newhall) seedlings were grown in a greenhouse under controlled conditions, with temperatures maintained at 27–28 °C and relative humidity at 60–65% under a 14-hour light/10-hour dark photoperiod. The D. citri colony used in this study was originally established in 2015 from field-collected populations in Nankang District, Jiangxi Province. The colony was reared on Murraya paniculata seedlings in insect-rearing cages under the same controlled greenhouse conditions described above. To ensure the absence of CLas, samples from the D. citri colony were routinely subjected to DNA extraction and tested by qPCR using CLas-specific primers [18]; only CLas-free psyllids were used for plant infestation experiments.

Sample preparation and RNA extraction

Citrus shoots were collected under three experimental conditions: uninfested controls (CK), 24 h post-feeding (hpf24), and 48 h post-feeding (hpf48) following D. citri infestation. These time points are widely used in transcriptomic studies of plant responses to phloem-feeding insects, such as aphids and whiteflies [19, 20]. For the infestation treatments, approximately 200 adult D. citri individuals were introduced onto each healthy citrus shoot. Each shoot was immediately enclosed in a fine-mesh bag (18 cm height × 13 cm width) to prevent insect escape while allowing D. citri free access to the shoot throughout the experiment. Active feeding was confirmed by visual inspection, with D. citri stably settled on the shoot surface and exhibiting a typical stationary feeding posture.

Citrus samples were harvested at 24 h and 48 h after psyllid placement, flash-frozen in liquid nitrogen, and stored at − 80 °C until RNA extraction. Each treatment consisted of three biological replicates, with each replicate comprising three shoots harvested from independent citrus seedlings. These replicates were subsequently used for RNA extraction, RNA-seq, and gene expression analysis. Total RNA was extracted using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. RNA concentration, quality, and integrity were evaluated using a NanoDrop One spectrophotometer (Thermo Fisher Scientific, MA, USA).

Complementary DNA (cDNA) library construction and Illumina sequencing

cDNA library construction and sequencing were outsourced to Personal Biotechnology Co., Ltd. (Shanghai, China). For each sample, 3 µg of total RNA was used for mRNA purification using poly(T)-coated magnetic beads. The purified mRNA was fragmented at elevated temperatures using Illumina Fragmentation Buffer. First-strand cDNA synthesis was performed using random oligonucleotides and SuperScript II reverse transcriptase (Invitrogen), followed by second-strand synthesis using DNA Polymerase I and RNase H (Invitrogen). The resulting double-stranded cDNA underwent end repair and 3′-adenylation prior to ligation with Illumina paired-end adapters. Fragments of 400–500 bp were size-selected using the AMPure XP system (Beckman Coulter, CA, USA). Adapter-ligated fragments were enriched by 15-cycle PCR, followed by purification and quantification using an Agilent 2100 Bioanalyzer (Agilent Technologies, Inc., California, USA). The final libraries were subjected to paired-end sequencing using an Illumina NovaSeq 6000 platform.

DEG and DAS analyses

Illumina sequencing reads were processed for quality control using fastp (v0.22.0) to remove low-quality reads and adapter-contaminated sequences. High-quality reads were aligned to the C. sinensis reference genome (v3.0; http://citrus.hzau.edu.cn) using HISAT2 (v2.1.0). HTSeq (v0.9.1) was used to count the reads mapped to each gene. Gene expression quantification was performed by calculating the Fragments Per Kilobase of transcript per Million mapped reads using StringTie software [21]. DEGs between CK and D. citri-infested citrus shoot samples were identified using DESeq2 (v1.38.3), with a statistical significance threshold of P < 0.05 and an absolute log₂ fold change ≥ 1.0. DAS events were detected using rMATS (v4.0.1) [22]. For DAS analysis, percent spliced-in (PSI) values were calculated, and significant splicing changes were defined as those with |ΔPSI| > 0.05 and either P-value < 0.05 or a false discovery rate < 0.05, as described by Huang et al. [23].

Gene Ontology (GO) enrichment analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis

Genes containing at least one DAS event were defined as DSGs. All genes were functionally annotated using GO terms, and the number of significantly enriched genes associated with each term was determined. GO enrichment analyses were conducted separately for DEGs, DSGs, and the overlapping gene set shared between DEGs and DSGs, using clusterProfiler (v4.6.0). For GO enrichment analysis, statistical significance was assessed using the hypergeometric test, and the Benjamini–Hochberg procedure was applied to control the false discovery rate (FDR) for multiple comparisons. An adjusted P-value < 0.05 was set as the threshold for significance. KEGG pathway enrichment was performed using the clusterProfiler package (v4.6.0), with the same FDR-adjusted significance cutoff (adjusted P-value < 0.05).

Quantitative real-time PCR (qRT-PCR) validation

Total RNA was extracted from CK, hpf24, and hpf48 samples using TRIzol reagent, as previously described. For each sample, 1 µg of total RNA was reverse transcribed using EasyScript® One-Step gDNA Removal and cDNA Synthesis SuperMix (TransGen Biotech). qRT-PCR reactions were performed in a final volume of 20 µL, containing 4 µL of 20-fold diluted cDNA template, 0.4 µM of each primer, and 10 µL of PerfectStart® Green qPCR SuperMix (TransGen Biotech). Amplification was performed using a Roche LightCycler® 96 Real-Time PCR system. Two commonly used reference genes, CsActin and CsGAPDH, were used as internal controls to normalize the data. The primer sequences are listed in Supplementary Table S1. The relative expression levels of the selected genes and their respective splice variants across all samples were calculated using the 2−ΔΔCt method [24].

Statistical analysis

The relative expression levels of JA-related genes in the hpf24 and hpf48 samples were statistically compared with those in CK citrus shoots using Student’s t-test. All statistical analyses were performed using GraphPad Prism (version 6.0; GraphPad Software Inc., La Jolla, CA, USA).

Results

Transcriptome overview

A total of nine RNA samples were sequenced from citrus shoots, generating an average of 43.04 ± 4.47 million raw reads per sample. After adapter trimming and removal of low-quality reads, the clean data output was 6.39 ± 0.54 Gb for CK samples, 5.44 ± 0.19 Gb for hpf24 samples, and 6.22 ± 0.68 Gb for hpf48 samples (Table S2). The GC content across samples ranged from 43.41% to 43.89%, whereas the Q30 quality scores varied between 96.67% and 97.17% (Table S2). More than 92.31% of the clean data were successfully mapped to the citrus reference genome, with uniquely mapped reads ranging from 33.18 to 44.45 million per sample (Table S3). These results indicate that both the sequencing data and reference genome were of sufficient quality for subsequent analysis.

Isolation and functional enrichment analysis of DEGs in D. citri-infested citrus shoots

Principal component analysis revealed a clear separation of gene expression profiles between D. citri-infested and CK plants. In contrast, the expression profiles of infested plants at hpf24 partially overlapped with those at hpf48 (Figure S1). Compared with the CK groups, a total of 2,273 DEGs were identified in the hpf24 shoots, including 1,435 upregulated and 838 downregulated genes (Fig. 1A; Table S4). In hpf48 shoots, 3,049 DEGs were detected, comprising 2,069 upregulated and 980 downregulated genes. When comparing hpf48 with hpf24, 845 genes were upregulated and 188 genes were downregulated in the hpf48 shoots (Fig. 1A; Table S5). Notably, 184 DEGs were commonly shared across all three pairwise comparisons: ‘hpf24 vs. CK’, ‘hpf48 vs. CK’, and ‘hpf48 vs. hpf24’ (Fig. 1B).

Fig. 1
Fig. 1
Full size image

Differentially expressed genes (DEGs) in citrus shoots infested with D. citri at 24 h post feeding (hpf24) and 48 h post feeding (hpf48). A The number of up-regulated and down-regulated DEGs in different comparison groups. B Venn diagram of DEGs among comparison groups. CK, uninfested citrus shoots

To further elucidate the functional roles of these upregulated DEGs, GO and KEGG enrichment analyses were conducted. GO enrichment analysis identified the most significantly enriched functional categories among the upregulated DEGs in D. citri-infested citrus as ‘response to stimulus’, ‘biological regulation’, and ‘signaling’. At hpf24, 405 and 337 upregulated DEGs were associated with ‘response to stimulus’ and ‘biological regulation,’ respectively, whereas at hpf48, 570 and 492 upregulated DEGs were assigned to these categories (Fig. 2A; Figure S2). In contrast, the downregulated DEGs in D. citri-infested citrus were predominantly enriched in ‘regulation of biosynthetic process’ and ‘regulation of cellular biosynthetic process’ (Figure S3). KEGG enrichment analysis revealed that ‘plant hormone signal transduction’ was the most significantly enriched pathway among the upregulated DEGs in D. citri-infested citrus, with 34 DEGs identified at hpf24 and 43 DEGs at hpf48. This was followed by the pathways ‘phenylpropanoid biosynthesis,’ with 25 and 36 DEGs, respectively, and ‘α-linolenic acid metabolism,’ with 15 and 16 DEGs, respectively (Fig. 2B; Figure S4). However, no significantly enriched KEGG pathways were detected among downregulated DEGs in either the ‘hpf24 vs. CK’ or ‘hpf48 vs. CK’ comparisons (adjusted P > 0.05; Figure S5).

Fig. 2
Fig. 2
Full size image

GO and KEGG enrichment analysis of upregulated DEGs in citrus shoots following D. citri infestation. A Representative level-2 GO functional classifications of upregulated DEGs. B Top five KEGG enrichment pathways of upregulated DEGs. hpf24, citrus shoots infested with D. citri at 24 h post feeding; hpf48, citrus shoots infested with D. citri at 48 h post feeding. “ns” indicates no significant difference; asterisks denote statistically significant differences as determined by a hypergeometric test (**, adjusted P < 0.01)

JA biosynthesis genes were upregulated in response to D. citri feeding

JA signaling is widely recognized as the primary defense pathway that mediates plant resistance against sap-sucking insects. Given that α-linolenic acid is the key precursor for JA biosynthesis, and coupled with the observed significant enrichment of both ‘plant hormone signal transduction’ and ‘α-linolenic acid metabolism’ pathways in the KEGG database following D. citri feeding, this study performed a comprehensive transcriptomic assessment of genes associated with the JA biosynthetic pathway. In citrus plants, 19 key genes involved in JA biosynthesis and signaling, spanning the entire enzymatic cascade from α-linolenic acid to bioactive JA derivatives, were significantly upregulated after D. citri feeding (Table S6; Fig. 3). The first three enzymatic steps of JA biosynthesis occur within chloroplasts and collectively convert α-linolenic acid to 12-oxo-phytodienoic acid. These steps are catalyzed by lipoxygenase (LOX), allene oxide synthase (AOS), and allene oxide cyclase (AOC). Notably, the expression levels of five LOX, one AOS, and two AOC genes were significantly upregulated at 24 and/or 48 h post infestation with D. citri. Subsequently, 12-oxo-phytodienoic acid is transported from chloroplasts to peroxisomes, where it undergoes three further enzymatic modifications mediated by 12-oxophytodienoate reductase (OPR), OPC-8:0-CoA ligase (OPCL), and acyl-CoA oxidase (ACX), which ultimately produce JA. In this study, four DEGs encoding OPR, one encoding OPCL, and three encoding ACX were identified (Table S6; Fig. 3).

Fig. 3
Fig. 3
Full size image

Differential expression profiles of genes involved in the jasmonic acid (JA) biosynthesis pathway. A Schematic representation of the JA biosynthesis pathway, including key enzymes such as lipoxygenase (LOX), allene oxide synthase (AOS), allene oxide cyclase (AOC), 12-oxophytodienoate reductase (OPR), OPC-8:0 CoA ligase (OPCL), acyl-CoA oxidase (ACX), and jasmonoyl-isoleucine synthetase (JAR1). B Heat map of genes associated with JA biosynthesis. Citrus samples were collected from plants under three experimental conditions: uninfested controls (CK), and at 24 h post-D. citri feeding (hpf24) and 48 h post-D. citri feeding (hpf48). Values in the heatmap represent log₂-fold changes of gene expression relative to CK for each treatment. Statistical significance was determined using DESeq2: “ns” indicates no significant difference, and asterisks denote statistically significant differences (*P < 0.05; **P < 0.01)

JA-isoleucine (JA-Ile), the most bioactive form of JA, is synthesized through the conjugation of JA and isoleucine by jasmonoyl-isoleucine synthetase (JAR1). Notably, the expression level of DEG Cs_ont_5g004250, which encodes JAR1, was significantly upregulated compared with that of control samples (Table S6; Fig. 3). JA-Ile accumulation induces the ubiquitin-mediated proteasomal degradation of jasmonate ZIM-domain (JAZ) repressor proteins. After the degradation of JAZ repressors, the basic helix–loop–helix transcription factor MYC2 acts as a principal regulator by orchestrating the transcriptional activation of JA-responsive downstream genes. Consistently, two DEGs (Cs_ont_7g010850 and Cs_ont_5g005120) encoding MYC2 were significantly upregulated in response to D. citri herbivory. These findings suggest that D. citri feeding activates the ‘α-linolenic acid metabolic’ pathway and promotes the biosynthesis of JA and JA-Ile, while simultaneously enhancing JA signaling through MYC2 induction.

DAS analysis of D. citri-infested citrus shoots

To investigate DAS events induced by D. citri, RNA-seq data from CK and D. citri-infested plants were analyzed using rMATS software. Genes hosting at least one DAS event were defined as DSGs. By comparing AS patterns between CK and infested plants at hpf24 and hpf48, we detected 422 DAS events (363 DSGs) in the hpf24 samples and 717 DAS events (580 DSGs) in the hpf48 samples (Fig. 4). AS events were classified into five canonical types: exon skipping (ES), intron retention (IR), alternative 5′ splice site (A5SS), alternative 3′ splice site (A3SS), and mutually exclusive exons (MXE). In both the hpf24 and hpf48 samples, IR was the predominant DAS event type, with 287 and 521 events, respectively; followed by ES (69 and 85 events), A5SS (33 and 58), A3SS (33 and 51), and MXE (0 and 2) (Fig. 4). GO enrichment analysis showed that the category ‘regulation of nuclear division’ comprising nine DSGs was significantly enriched in the hpf48 samples (Figure S6), whereas no biological pathways were significantly enriched among the DSGs according to KEGG pathway enrichment analysis (Figure S7).

Fig. 4
Fig. 4
Full size image

Alternative splicing patterns and the number of differentially spliced events in citrus shoots infested with D. citri at 24 h post feeding (hpf24) and 48 h post feeding (hpf48). A Schematic illustration of the five major types of alternative splicing. B Number of differential splicing events for each alternative splicing type in response to D. citri infestation. ES, exon skipping; A5SS, alternative 5′ splice site; A3SS, alternative 3′ splice site; MXE, mutually exclusive exons; IR, intron retention

Integrative analysis of gene expression and AS induced by D. citri

To investigate the relationship between D. citri-induced changes in gene expression and transcript splicing, we compared the sets of DEGs and DSGs in the hpf24 and hpf48 samples. As shown in Fig. 5, there was a partial overlap between the DEG and DSG datasets: of the 363 DSGs detected in the hpf24 sample, 69 genes were also differentially expressed (Fig. 5A; Table S7), and among the 580 DSGs identified in the hpf48 sample, 137 genes were shared with the DEG dataset (Fig. 5B; Table S8).

Fig. 5
Fig. 5
Full size image

Comparison of differentially expressed genes (DEGs) and differentially spliced genes (DSGs) in citrus shoots following D. citri infestation. A Venn diagram showing the overlap between DEGs and DSGs at 24 h post-infestation (hpf24). B Venn diagram showing the overlap between DEGs and DSGs at 48 h post-infestation (hpf48). CK, uninfested citrus shoots

GO functional analysis of the overlapping genes between DEGs and DSGs revealed distinct profiles at the two time points. At hpf24, significant enrichment was observed for the GO terms including ‘peptidoglycan metabolic process’ (2 genes), ‘glycosaminoglycan catabolic process’ (2 genes), and ‘beta-N-acetylhexosaminidase activity’ (2 genes), among others. At hpf48, however, the overlapping genes were predominantly enriched in three categories: ‘protein modification process’ (26 genes), ‘phosphorus metabolic process’ (25 genes), and ‘phosphate-containing compound metabolic process’ (25 genes) (Figure S8). KEGG pathway enrichment analysis identified that ‘plant hormone signal transduction’ was the most enriched pathway among the DEG–DSG overlapping genes, encompassing six overlapping genes at hpf48 (Figure S9). These six genes are functionally distributed across multiple phytohormone signaling cascades: auxin (Cs_ont_4g022490, encoding an auxin-responsive protein), ethylene (Cs_ont_2g030110, encoding CsEIN3), JA (Cs_ont_4g022880, encoding a TIFY domain-containing protein), and cytokinin (Cs_ont_2g013750, encoding CsRR21), as well as two serine/threonine-protein kinases putatively involved in abscisic acid (Cs_ont_5g043390, CsSAPK3) and brassinosteroid (Cs_ont_5g047680, CsBSK7) signaling, respectively.

Expression validation of DEGs upregulated in JA biosynthesis genes

To further validate our RNA-seq expression profiling data, we conducted qRT-PCR assays on nine JA-related DEGs that were identified as being upregulated in response to D. citri infestation based on the Fragments Per Kilobase of transcript per Million mapped reads values. These genes included CsLOX2-1, CsAOS, CsAOC3, CsOPR2-1, CsOPR3, CsOPCL1, CsACX1, CsJAR1, and CsMYC2-1 (Fig. 6; Table S6). The qRT-PCR results were largely consistent with the RNA-seq data, showing significant increases in the expression of all nine genes at 24 and/or 48 hpf. For instance, CsLOX2-1 expression increased 5.97-fold at 24 hpf and 3.85-fold at 48 hpf, CsAOS expression increased 1.67-fold at 24 hpf and 2.82-fold at 48 hpf, and CsJAR1 expression increased 7.19-fold at 24 hpf and 9.80-fold at 48 hpf (Fig. 6). This validation not only reinforced the reliability of our RNA-seq data, but also provided quantitative evidence for the coordinated activation of the JA biosynthetic and signaling cascade in citrus plants during D. citri infestation.

Fig. 6
Fig. 6
Full size image

The relative expression of JA-related genes in response to D. citri infestation. Expression levels of the selected genes in the uninfested citrus shoots (CK) were assigned an arbitrary value of 1. The expression levels in citrus shoots infested with D. citri at 24 h post feeding (hpf24) and 48 h post feeding (hpf48) are presented relative to the mean expression of CK. All data are represented as means ± standard error of the mean (SEM) from three biological replicates. Asterisks indicate significant differences between the infested (hpf24 and hpf48) and CK samples (*, P < 0.05; **, P < 0.01; Student’s t-test)

Discussion

Citrus plantations are currently threatened by D. citri, the primary insect vector of huanglongbing pathogens, which causes the most destructive diseases affecting citrus worldwide. Therefore, understanding the mechanisms underlying citrus defense responses to D. citri infestation is crucial for developing effective control strategies. To this end, we performed a comprehensive transcriptomic analysis to investigate the transcriptional responses of citrus plants to D. citri infestation. By further analyzing the differential expression of JA pathway genes and AS events, we elucidated the complex regulatory networks underlying citrus defense against this insect vector.

When plants are attacked by herbivores, phytohormones accumulate as signaling molecules and trigger the expression of downstream defense-related genes, thereby activating the plant immune response [3, 25, 26]. A previous study has found that JA signaling pathway was involved in the citrus defense against D. citri [27]. Exogenous application of JA enhanced citrus resistance to D. citri, whereas D. citri effectors DcE1 and DcE13 suppressed the expression of JA-responsive marker genes, thereby facilitating feeding on host plant phloem [16, 17, 27]. JA is biosynthesized from α-linolenic acid, a precursor produced via the octadecanoid pathway, which proceeds sequentially through plant chloroplasts, peroxisomes, and the cytoplasm. In this study, feeding by D. citri significantly upregulated genes associated with α-linolenic acid metabolism, JA biosynthesis, and JA signaling pathways. These include genes involved in the conversion of α-linolenic acid to JA (LOX, AOS, AOC, OPR, OPCL, and ACX), the JAR1 gene responsible for converting JA into its bioactive form JA-Ile, and the JA signaling regulator MYC2. Interestingly, many other sap-sucking insects, such as aphids and whiteflies, have also been shown to activate JA-mediated signaling pathways in their host plants [19, 28, 29].

While JA signaling is well recognized as a key component of plant defense against herbivores, we acknowledge that endogenous JA and/or JA-Ile levels were not directly quantified in this study. Nevertheless, previous studies in citrus and other plant species have demonstrated that upregulation of LOX, AOS, AOC, and OPR is consistently accompanied by significant increases in JA and JA-Ile upon herbivory or wounding [27, 30, 31], supporting the inference that the transcriptional reprogramming observed in our dataset likely elevated JA and JA-Ile levels in D. citri-infested citrus shoots. It is worth noting that transcriptomic responses to D. citri infestation can vary across studies due to methodological differences, such as tissue type and citrus cultivar. In a previous leaf-focused analysis by Sun et al. [32], several defense-related pathways, including peroxisome, α-linolenic acid metabolism, and phenylpropanoid and terpenoid biosynthesis, were reported to be activated upon D. citri infestation. In contrast, our study provides a complementary perspective by examining D. citri-infested citrus shoots and integrating transcriptional and post-transcriptional analyses. Specifically, we highlight the important role of JA-mediated defense in the citrus–D. citri interaction, along with evidence for dynamic, time-dependent regulation via AS at 24 and 48 h post-infestation.

In addition to the activation of JA‑mediated defense signaling, our transcriptomic data revealed widespread transcriptional suppression, with 838 and 980 genes downregulated at hpf24 and hpf48, respectively. GO enrichment analysis further showed that these downregulated genes were significantly enriched in ‘regulation of biosynthetic process’ and ‘regulation of cellular biosynthetic process’ categories, which encompass genes involved in photosynthesis and energy metabolism. The downregulation of biosynthesis-related genes appears to be a conserved adaptive response to biotic stress; indeed, a meta‑analysis of 46 plant–pest interaction studies demonstrated that herbivore infestation consistently reduces plant photosynthesis, growth, and reproduction [33, 34]. This pattern is consistent with a well‑documented resource reallocation strategy between plant growth and defense, in which plants suppress primary metabolism to prioritize defense under biotic stress [35, 36]. In this context, the concurrent upregulation of JA-related defense genes and downregulation of biosynthetic genes in our study may reflect highly coordinated, genome-wide transcriptional reprogramming that balances metabolic costs and defense. We propose that this shift allows citrus plants to temporarily reduce photosynthetic investment, redirecting energy and substrates toward defensive compound biosynthesis while maintaining survival under D. citri infestation.

AS increases the diversity of transcripts and proteins, enabling plants to respond dynamically to biotic stresses via post-transcriptional regulatory mechanisms. Advances in bioinformatics have significantly enhanced the detection and characterization of AS events, with RNA-seq being the most widely used technique. This approach involves sequencing the transcriptome and using splice-aware algorithms to align transcriptomic data with a reference genome, thereby facilitating the identification of splicing variations and intron locations [6, 37]. By applying this method and comparing AS patterns between CK and infested plants, we identified 422 DAS events corresponding to 363 DSGs in the hpf24 samples and 717 DAS events corresponding to 580 DSGs in the hpf48 samples. All five canonical AS types, IRs, ESs, A3SSs, A5SSs, and MXEs were detected in the DAS events. Notably, IRs were the most prevalent, accounting for > 72% of the total DAS events. This predominance of IR aligns with previous reports in both pathogen- or herbivore-infested and uninfested plants, including cotton, tobacco, tomato, and Eucalyptus grandis, where IR has been identified as the most frequent AS event [6, 12, 38, 39]. For instance, during early infection of tomato by Meloidogyne incognita, 933 DAS events were identified 11 days post-inoculation. Among these, 180 of 337 DAS events significantly enriched in galls corresponded to IR, and 377 of the 596 events enriched in neighboring root tissues were classified as IR‑type events [39].

AS operates largely independently of transcriptional regulation, providing an additional layer of control in response to biotic and abiotic stress. For instance, in hypoxia-treated rice seeds, 1,741 DAS events from 811 genes were identified, with 95% of these DSGs showing no overlap with DEGs [40]. Consistently, a limited overlap between DEGs and DSGs, along with their involvement in distinct biological processes, has been observed across diverse plant–stress systems. These include rice exposed to cadmium, cold, flooding, and drought stresses [41,42,43]; OsSR45-mediated resistance to N. lugens in rice [11]; foxtail millet subjected to salt stress [44]; tomato leaves infected with Phytophthora infestans [45]; and oilseed rape (Brassica napus) infected with the fungal pathogen Sclerotinia sclerotiorum [46]. In this study, a combined analysis of gene expression and AS revealed that 69 genes overlapped between DEGs and DSGs at hpf24, and 137 overlapping genes were identified at hpf48. The limited overlap between the DEG and DSG groups suggests that AS operates largely independently of transcriptional regulation during D. citri infestation. Nevertheless, these overlapping genes likely represent key regulatory nodes where transcriptional and post-transcriptional mechanisms converge to fine-tune defense responses. However, it should be noted that the DAS events, particularly those among the DEG–DSG overlapping genes, were not experimentally validated by RT-PCR using isoform-specific primers, a step that would help to infer their functional relevance during D. citri infestation.

In summary, this study elucidates the coordinated transcriptomic and spliceomic reprogramming in citrus following D. citri infestation, highlighting the activation of JA-mediated defenses and widespread AS events that largely operate independently of transcriptional changes. These findings advance our understanding of the multilayered regulatory networks underlying plant-insect interactions and underscore the importance of post-transcriptional regulation in defense responses. Future functional characterization of key DEGs within the JA pathway, and experimental validation of the overlapping genes between DEGs and DSGs will be essential to decipher their precise roles.