Transcriptome and small RNA sequencings reveal the response of tobacco to aphid infestation

Abstract

Aphids threaten many economically important crops by extracting plant sap and efficiently transmitting plant viruses. To elucidate the molecular mechanisms underlying plant-aphid interactions, we performed mRNA, circRNA, and microRNA sequencing to investigate the response of tobacco plants to infestation by the aphid Myzus persicae. Our results revealed that aphid infestation significantly upregulated 1,091 genes and downregulated 407 genes. Notably, differentially expressed genes were enriched in several key pathways, including the MAPK signaling pathway, α-linolenic acid metabolism, glutathione metabolism, and plant-pathogen interactions. Analysis of circRNA expression identified 53 circRNAs with significant changes following aphid infestation. However, neither their predicted miRNA targets nor their host genes exhibited altered expression levels, suggesting that circRNAs may play previously uncharacterized roles in plant responses to aphid attack. Approximately 400 miRNAs were predicted in each library, with nearly two-thirds identified as novel. Among them, 22 miRNAs were differentially expressed in response to aphid infestation, but only 8 were predicted to target mRNAs, indicating a selective role of miRNAs in regulating gene expression during the aphid response. In summary, our findings shed light on the transcriptional and post-transcriptional regulatory mechanisms activated in tobacco during aphid infestation. Our findings provide a basis for future studies of aphid-responsive regulatory mechanisms in plants.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12864-025-12361-8.

Background

Aphids preferentially inhabit the abaxial surface of leaves and plant stems, using their piercing-sucking mouthparts to extract phloem sap. Although small in size, aphids reproduce rapidly, making them challenging to control. Their infestations can significantly impact the number of inflorescences on host plants and suppress floral bud growth [1]. Additionally, aphids serve as vectors for various plant viruses, including potato virus Y (PVY), cucumber mosaic virus (CMV), barley yellow dwarf virus (BYDV), and potato leaf roll virus (PLRV) [2–4], posing a serious threat to the production of many cash crops. Currently, a range of methods has been proposed to combat aphids, including chemical pesticides [5], natural alarm pheromones [6], and biological control through the use of natural predators [7, 8]. From the perspective of plant improvement, developing insect-resistant cultivars through hybridization or genetic engineering [9, 10] is a promising avenue for pest control. Leveraging the natural defense systems of plants to create an insect-resistant network could significantly minimize environmental impacts associated with pest management.

Plants have evolved a variety of defense mechanisms to counteract insect infestations. First, plants can upregulate the expression of specific response genes to enhance physical barriers [11, 12] or promote the biosynthesis of toxic secondary metabolites [13–15]. For instance, infestation by the generalist herbivore Spodoptera littoralis induces the expression of MYC2/3/4 transcription factors, which in turn enhances glucosinolate biosynthesis and strengthens plant resistance [16]. Similarly, transcript levels of OsERF3 are upregulated in rice stems upon mechanical wounding or infestation by the rice striped stem borer (SSB), contributing to resistance by regulating the expression of trypsin protease inhibitors [17]. Second, plants modulate gene expression at the post-transcriptional level through various non-coding RNAs, including microRNAs (miRNAs) [18], circular RNAs (circRNAs), and long non-coding RNAs (lncRNAs) [19, 20]. In alfalfa, microRNA396 (miR396) enhances resistance to S. litura larvae by downregulating genes involved in lignin, flavonoid, and glucosinolate biosynthesis [21]. Third, plants control the accumulation and activity of resistance-related proteins through post-translational modifications, such as phosphorylation [22–24], glycosylation [25] and ubiquitination [26, 27].

Apart from the well-known mRNA, tRNA, and rRNA, plants produce a variety of RNAs transcribed from the genome that carry genetic information but do not encode proteins. These RNAs are collectively referred to as ncRNAs [28, 29]. Based on their length, ncRNAs are classified using a 200-nt threshold: those shorter than 200 nucleotides are considered as small ncRNAs, which include microRNA (miRNA), small interfering RNA (siRNA), piwi-interacting RNA (piRNA), trans-activating CRISPR (tracr) RNA, signal recognition particle RNA (7SL), small Cajal body-specific RNA (scaRNA) and others [30, 31]. lncRNAs are transcripts longer than 200 nucleotides, generally exhibit low sequence conservation across species, and lack protein-coding potential [32–34].

Aphids are highly destructive agricultural pests that damage a wide range of crops worldwide. By piercing the phloem and extracting sap, they weaken plants and produce honeydew, which promotes sooty mold growth. In addition to direct feeding damage, aphids serve as vectors for numerous plant viruses [35]. Previous studies have investigated the transcriptomic responses of various plants, including rose [36], cowpea [37], rapeseed [38], sorghum [39], alfalfa [40], to aphid infestation. However, the small RNA-mediated response to aphids remains largely unexplored. Tobacco (Nicotiana tabacum) is an important economic crop in our country. To investigate its response to aphid feeding, we performed whole-transcriptome analysis to characterize RNA and small RNA changes in tobacco (Nicotiana tabacum) in response to aphid infestation. By further investigating post-transcriptional regulation mediated by microRNAs (miRNAs) and circular RNAs (circRNAs), we uncovered the complex regulatory networks underlying tobacco’s defense responses to aphids. This work provides new insights into the molecular mechanisms of plant–insect interactions and lays a foundation for developing novel strategies to enhance crop resistance.

Results

Transcriptomic characteristics of tobacco after aphid infestation

To investigate the molecular response of tobacco to aphid infestation, we performed RNA-seq analysis on control (CK) and aphid-infested (Aphid) samples. Each sample yielded more than 17 Gb of raw data, with mapping rates exceeding 93% (Table S1). Principal component analysis (PCA) showed that PC1 explained 43.5% of the total variance, indicating that aphid feeding is the main factor driving transcriptomic differences (Figure S1A). Hierarchical clustering heatmaps also confirmed that biological replicates within the same treatment group clustered closely together, demonstrating good reproducibility (Figure S1B).

A total of 66,812 genes were detected across all samples. Differential expression analysis identified 1,091 genes as significantly upregulated and 407 genes as downregulated in response to aphid infestation (Fig. 1A, Table S2). Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis revealed that these differentially expressed genes (DEGs) were significantly enriched in several functional pathways, including the MAPK signaling pathway, alpha-linolenic acid metabolism, glutathione metabolism, ribosome biogenesis in eukaryotes, DNA replication, and plant-pathogen interaction (Fig. 1B). The salicylic acid (SA) signaling pathway, known for its critical role in plant defense against phloem-feeding insects [41], appeared to be activated upon aphid attack. Notably, we observed increased expression of two salicylate carboxymethyltransferase-like (SAMT-like) genes and one pathogenesis-related protein 1 (PR1) gene following aphid infestation (Fig. 1C), which are potentially indicative of SA pathway activation. Furthermore, several WRKY transcription factor, well-established roles in regulating plant defense responses to aphid [42–44], including those mediated by salicylic acid, also showed altered expression levels following infestation (Fig. 1D), suggesting their involvement in modulating tobacco’s defense responses to aphid feeding.

Fig. 1.

Differentially expressed genes (DEGs) induced by aphids.A Volcano plot of DEGs. FDR < 0.05, |log₂Foldchange| > 1. B Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis for DEGs (top20). Sort by p value. C The relative expression level of SAMTs (Nta22g13590, Nta04g07620) and PR1 (Nta01g05390) in RNA-seq. D The relative expression level of WRKYs in RNA-seq

CircRNAs

In addition to mRNA-level changes, we next investigated the expression landscape of circular RNAs (circRNAs), which are emerging regulators of gene expression in plants. We identified 114 circRNAs, the majority of which originated from exonic regions of genes (Fig. 2A). Their chromosomal distributions are shown in Fig. 2B, and most circRNAs ranged in length from 200 to 1200 bp (Fig. 2C). A total of 114 circRNAs were identified, including 26 upregulated and 27 downregulated circRNAs (Fig. 2D and Table S3). circRNAs regulate gene expression via multiple mechanisms, including direct interaction with their host genes or acting as miRNA sponges to indirectly influence gene expression [45]. Among the 114 circRNAs, 22 were predicted to directly target host genes (Fig. 2E, Table S4). However, none of these host genes exhibited differential expressions following aphid infestation (Fig. 2E).

Fig. 2.

Differentially expressed Circular RNAs (circRNAs) induced by aphids.A The distribution of circRNAs in different regions of genes. B The distribution of circRNAs on the chromosomes. C The length of circRNAs. D Volcano plot of differential expressed circRNAs. p < 1, |log₂Foldchange| > 0.5849625007211562. E The Venn diagram of circRNA-targeted host genes and differential expressed mRNA

Some circRNAs may have the potential to be translated into peptides or proteins. In our analysis, three circRNAs -Chr11:213213119|213,228,582, Chr09:39723389|39,731,269, and Chr11:71653710|71,658,532- were predicted to possess protein-coding potential. Notably, Chr11:71653710|71,658,532 was predicted to encode a peptide containing both RRM_1 and SPOC domains and was found to be upregulated following aphid infestation. However, these predictions are speculative, and the potential role of this circRNA in tobacco resistance to aphid feeding remains to be experimentally validated.

Overview of MiRNA

MicroRNAs (miRNAs) regulate gene expression post-transcriptionally by directing mRNAs toward degradation or translational repression [46]. To characterize miRNA dynamics in response to aphid infestation, we profiled miRNA expressions. Each sample yielded approximately 2.1 to 4.54 million raw reads (Table S5), with aphid-infested samples producing approximately 13% more clean reads than the control (CK) group (Fig. 3A). After alignment to the Rfam and Repbase databases, RNA species in each library were annotated (Fig. 3B). Length distribution analysis of small RNAs (sRNAs) revealed that aphid infestation increased the proportion of 22-nt sRNAs (Fig. 3C-D).

Fig. 3.

Small RNAs (sRNAs) triggered by aphid. A The proportion of clean reads for each sample. B RNA types present in each library. C sRNA length distribution of CK. D sRNA length distribution of aphid infestation treatment

In total, 384 miRNAs were identified, including 253 novel and 131 known miRNAs. Known miRNAs were 20–24 nucleotides in length (Fig. 4A), whereas novel miRNAs ranged from 18 to 25 nucleotides (Fig. 4B). The nucleotide bias of plant miRNAs, particularly at the 5^’ end, influences their loading into specific AGO complexes and affects target recognition and silencing efficiency [47, 48]. Thus, we next performed nucleotide bias analysis and found that a strong enrichment of adenine (A) at the first nucleotide position of novel miRNAs (Figs. 4C-D). Further examination of first-nucleotide composition across different miRNA lengths revealed that, compared with 20-nt known miRNAs, novel miRNAs had a higher proportion of uracil (U) at the first position. An A bias was particularly pronounced in 24-nt miRNAs, while 18-, 23-, and 25-nt miRNAs preferentially started with either U or A. Interestingly, 19-nt miRNAs predominantly began with U (Figs. 4E-F).

Fig. 4.

Overview of microRNA (miRNA) after aphid infestation. A The length of known miRNAs. B The length of novel miRNAs. C Nucleotide bias analysis of known miRNA. D Nucleotide bias analysis of novel miRNA. E Nucleotide bias analysis for known miRNA with different length. F. Nucleotide bias analysis for novel miRNA with different length

miRNA-mRNA and miRNA-circRNA

A total of 22 differentially expressed miRNAs were identified after aphid infestation, comprising 7 known and 15 novel miRNAs (Fig. 5A& Table S6). Among them, the mature sequence of nta-miR477a was identical to that of ghr-miR477 [49], but distinct from Csn-miR477 [50] (Fig. 5B). In addition, high sequence similarity was observed between the precursors of novel_miR_44975 and novel_miR_39647, as well as between novel_miR_40256 and novel_miR_2762 (Fig. 5C&D).

Fig. 5.

miRNAs triggered by aphid.A Volcano plot of differential expressed miRNAs. p < 0.05, |log₂Foldchange| > 1. B Sequence alignment of miR477 in different species. The highlighted blue color indicates the different bases. C-D The sequence and structure of the novel miRNA. Red indicates mature miRNA sequences, and highlighted blue indicates the different bases. E Comparison of the sequences of miR477a with PAL and CBP60A. The highlighted blue color indicates the different bases (F) novel_miR44975 and novel_miR39647 regulate the expression of Nta03g13750

Target prediction revealed candidate genes for two known and six novel miRNAs (Table S7). nta-miR477a was predicted to target five genes, including two replication factor C subunit 3-like genes (Nta01g21720, Nta02g33660), one DELLA protein RGL1-like (Nta05g20150), and one GRAS family protein RAM1-like (Nta06g18220) (Table S7). However, none of these target genes exhibited significant differential expression after aphid infestation. We further examined whether nta-miR477a could regulate the previously reported genes PAL [50] and CBP60A [49]. Three PAL genes (Nta05g23200, Nta20g11550 and Nta19g14210) were predicted as potential targets of nta-miR477a (Fig. 5E).

Notably, the target gene Nta03g13750 (Floral homeotic protein APETALA 2), shared by novel_miR_44975 and novel_miR_39647, showed significant expression changes after aphid infestation. Luciferase (LUC) activity assays further confirmed that both novel_miR_44975 and novel_miR_39647 suppressed the expression of Nta03g13750 (Fig. 5F).

circRNAs, which are rich in miRNA binding sites, can function as miRNA sponges to regulate the downstream gene expression. Among the identified miRNAs, only three non-differentially expressed miRNAs were predicted to interact with circRNAs. Specifically, circRNA scaffold50:484138|500,708, targeted by nta-miR6164a, nta-miR6164a, and novel_miR_485, exhibited significant differential expression (Table S8).

Discussion

In this study, we systematically characterized the transcriptional and post-transcriptional regulatory landscape of tobacco (Nicotiana tabacum) in response to green peach aphid (Myzus persicae) infestation. By integrating mRNA, circRNA, and miRNA expression profiles, we uncovered a complex and multilayered regulatory network that underlies plant defense mechanisms against phloem-feeding insects.

Transcriptional responses to aphid feeding

Transcriptome profiling revealed extensive transcriptional reprogramming, with 1,091 genes upregulated and 407 genes downregulated after aphid attack (Fig. 1A). Enrichment analysis indicated that many DEGs are involved in classical defense-related pathways responding to the resistance to phloem-feeding insect, such as the MAPK signaling cascade [51], alpha-linolenic acid metabolism, and the salicylic acid (SA) pathway [52, 53] (Fig. 1B). The pronounced upregulation of PR1 and SAMT-like genes strongly suggests activation of SA-mediated defense, which is well established in resistance to both pathogens [54, 55] and phloem-feeding insects [41, 56]. Moreover, the differential expression of several WRKY transcription factors highlights their potential roles as key regulatory nodes orchestrating downstream defense gene expression. Similar to our findings, more than 2,000 differentially expressed genes (DEGs) were identified in cotton after 24 h of aphid or whitefly infestation, among which WRKY33, WRKY21, WRKY20, WRKY1, WRKY35, and WRKY3 were regulated by both aphid and whitefly attacks [57].

Involvement of circrnas in aphid-responsive regulation

Several studies have demonstrated that insect infestation can alter circRNA expression profiles [58–60]. For instance, in peach, green peach aphid infestation was shown to affect the expression of 10 circRNAs in leaves [61]. In our study, 53 circRNAs were differentially expressed in response to aphid attack (Fig. 2D). Although 22 circRNAs were predicted to act on their host genes, the expression levels of these host genes remained largely unchanged (Fig. 2E). This suggests that circRNA-mediated regulation may occur through alternative mechanisms, such as miRNA sponging or translation into functional peptides.

A variety of computational tools have been developed to predict the protein-coding potential of circRNAs [62–64]. However, only a limited number of protein-coding circRNAs have been experimentally validated to date [65–67]. In plants, only one functional peptide-coding circRNA has been reported: rice circWRKY9 (also referred to as WRKY9-88aa), which encodes an 88-amino acid peptide and confers broad-spectrum resistance against rice stripe mosaic virus (RSMV), blast disease, and bacterial leaf blight [68]. In our study, the circRNA Chr11:71653710|71,658,532 was notably upregulated following aphid infestation and was predicted to encode peptides containing both RRM_1 and SPOC domains, typically associated with RNA binding and transcriptional repression. This finding raises the intriguing possibility that this circRNA may contribute directly to gene regulation beyond their conventional noncoding functions, although experimental validation will be essential to confirm this role.

MiRNA dynamics and regulatory networks

Small RNA sequencing revealed substantial miRNA diversity, with 384 miRNAs identified, including a large proportion of novel candidates. Aphid infestation altered the abundance of 22 differentially expressed miRNAs (Table S6), suggesting that these small RNAs play pivotal roles in shaping the host transcriptome during stress adaptation. One key example is nta-miR477a, which was upregulated and predicted to target genes related to cell cycle regulation (e.g., replication factor C subunit 3-like gene) [69, 70] and hormone signaling (e.g., DELLA/GRAS proteins) [71, 72]. However, none of these predicted targets were significantly differentially expressed, implying that nta-miR477a may act primarily through translational repression rather than mRNA degradation. In addition, our analysis indicated that nta-miR477a may also target the previously reported PAL genes (Fig. 4E).

Interestingly, two novel miRNAs (novel_miR_44975 and novel_miR_39647) shared the same target, Nta03g13750, which was significantly downregulated after infestation. Functional validation via luciferase assays confirmed repression of this target (Fig. 4F), supporting the notion that these two miRNAs may function redundantly or cooperatively in aphid-induced defense regulation.

circRNA-miRNA crosstalk

The circRNA-miRNA regulatory axis plays an important role in plant defense and development. For example, Os-circANK functions as a sponge for miR398b to modulate bacterial blight infection in rice [73], while ciMIR156dS, derived from pri-miR156d, negatively regulates miR156 levels and participates in the transition from the vegetative to the reproductive stage [74]. Wang et al. predicte circ16–miR482a and circ116–miR319a pairs may be associated with peach jasmonic acid (JA) pathway to regulate plant resistance to green peach aphid [59]. In our study, potential circRNA-miRNA interactions were also investigated. Among the differentially expressed circRNAs, scaffold50:484138|500,708 was predicted to be targeted by three non-differentially expressed miRNAs. Although this interaction may not represent a central aphid-responsive module, it points to the existence of a complex and condition-specific miRNA-circRNA regulatory axis. These findings highlight the importance of exploring the spatial and temporal dynamics of circRNA–miRNA interactions during plant-insect interactions.

Conclusions and perspectives

Together, our findings reveal a basis for future studies of aphid-responsive regulatory mechanisms in plants involving transcription factors, circRNAs, and miRNAs in tobacco’s defense against aphid attack. The activation of the SA pathway, dynamic modulation of WRKY transcription factors, and post-transcriptional regulation via miRNAs and circRNAs reflect a coordinated defense strategy. Future studies employing approaches such as RNA pull-down, CLIP-seq, and functional validation assays will be essential to verify these interactions and elucidate their biological significance.

Methods

Growing conditions of plants and aphids

Seeds of tobacco (Nicotiana tabacum cv. SR1) were sown and germinated in a greenhouse maintained at 26 ± 1 °C with a 14-hour light/10-hour dark photoperiod and 65% relative humidity. Ten days after germination, seedlings were transplanted into plastic pots and grown for an additional three weeks before being used in experiments. Virus-free Myzus persicae colonies were reared on N. tabacum cv. SR1 [75] in an insectary under the same environmental conditions.

Sample collection

Approximately 500 age-synchronized (5–7-day-old) aphids were carefully transferred onto each 21-day old tobacco plant and allowed to feed freely for 24 h. Leaf samples were subsequently collected from both the aphid-infested plants and the uninfested control group (CK) for subsequent sequencing analysis. Each treatment included three biological replicates, with each replicate consisting of two individual plants.

RNA-seq

Total RNA was extracted using Trizol reagent (Invitrogen, CA, USA) according to the manufacturer’s instructions. RNA purity and integrity were assessed with a NanoDrop 2000 spectrophotometer (NanoDrop Technologies, Wilmington, DE, USA) and a Bioanalyzer 2100 system (Agilent Technologies, CA, USA). RNA contamination was evaluated by electrophoresis on a 1.5% agarose gel. Ribosomal RNA was subsequently removed with the Ribo-off rRNA depletion kit for Plant (Vazyme, N409-01).

For mRNA, cirRNA and lncRNA sequencing, libraries were prepared using the VAHTS Universal V10 RNA-seq Library Prep Kit (Vazyme, NR606-02). Library quality was verified using a Qubit 2.0 Fluorometer and an Agilent 2100 Bioanalyzer. Equal amounts of high-quality libraries were pooled and sequenced on the Illumina NovaSeq X Plus platform. For small RNA sequencing, the RNA molecules in a size range of 18-30nt were enriched by polyacrylamide gel electrophoresis (PAGE). Then the 3’ adapters were added and the 36-44nt RNAs were enriched. The 5’ adapters were then ligated to the RNAs as well. The ligation products were reverse transcribed by PCR amplification and the 140–160 bp size PCR products were enriched to generate a cDNA library and sequenced using Illumina NovaSeq X Plus.

Data analysis for mRNA

The raw data were filtered using fastp (0.23.2) [76] with the following parameters “--cut_front–cut_tail–cut_window_size 4-a auto–cut_mean_quality 20–length_required 36” to obtain the clean data. The filtered reads were then aligned to Nicotiana tabacum NtaSR1_genome_1.0 (http://lifenglab.hzau.edu.cn/Nicomics/Download/index.php) using HISAT2 (v2.2.1) [77] with the parameters “–phred33–no-mixed–no-discordant”. Gene expression levels were quantified using featureCounts(v2.0.1) [78] and and differential gene expression analysis was performed using DESeq2 [79] (1.16.1). GO and KEGG enrichment analyses of differentially expressed genes was employed using clusterProfiler (4.2.0) [80] and KOBAS [81], respectively.

Data analysis for circrna

Raw sequencing data were filtered using fastp and the resulting clean reads were aligned to the reference genome using HISAT2 [77]. CircRNA identification was performed using three tools: find_circ [82], CIRI [83] and CIRCexplorer2 [84]. circRNAs predicted by at least two of these tools and supported by more than two back-spliced reads were considered candidate circRNAs for further analysis. After normalization, circRNAs expression levels were quantified using edgeR [85], with differentially expressed circRNAs defined by a threshold of |log2(FoldChange)|>0.585. miRanda [86] was employed to predict potential miRNA targets of the identified circRNAs. In addition, the coding potential of circRNAs was evaluated using IRESfinder [87], CPC2 [63] and Pfam [88].

Annotation and analysis of MiRNA

The raw sequencing data were filtered using fastp (0.23.2) to obtain high-quality clean reads. These reads were first aligned to the Rfam [88] and Repbase [89] databases using Bowtie to remove non-coding RNAs (ncRNAs). The remaining reads were then mapped to the reference genome with Bowtie to identify microRNAs (miRNAs). For known miRNA identification, reads aligned to the reference genome were further matched against mature miRNA sequences in the miRBase database (v22) [90]. Reads meeting the alignment criteria were classified as known miRNAs. Novel miRNAs were identified using miRDeep2 [91]. miRNA target genes were identified using psRobot [92]. miRNA expression levels were quantified using miRDeep2, and differential expression analysis was conducted using edgeR, with significant thresholds set at |log₂(Fold Change)| >1.0 and p < 0.05.

Luciferase activity analysis

Transient protein expression in N. benthamiana was performed. To determine the activity of luciferase, we sprayed 0.5 mM luciferin (Promega) on the leaves and recorded signals using an HRPCS5 camera (Photek). Firefly luciferase produces luminescence spanning a broad spectrum from blue to green, yellow, red, and even white, with brighter signals corresponding to higher levels of gene expression. The primers used in this study can be found in Table S9.

Authors’ contributions

Conceptualization, writing—original draft preparation, J. L. and P.L.; investigation, J.L. and Y.Y.; writing—review and editing, X.W. and P. L.; funding acquisition, J.C. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by National Natural Science Foundation of China (U23A6006).

Data availability

The raw data presented in this study are available on the websites: [https://www.ncbi.nlm.nih.gov/bioproject/PRJNA1267184](https:/www.ncbi.nlm.nih.gov/bioproject/PRJNA1267184).

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Conflicts of interest

The authors declare no conflicts of interest.