Phytophthora nicotianae, a typical oomycete, is known for its broad host range (Meng et al., 2021). The pathogen attacks the plant at the collar region and below the soil. Microclimate with humid and moist shade acts as a favorable condition for both the growth of betelvine and P. nicotianae. The characteristic symptom of disease is the disappearance of lustre of leaves, followed by wilting and drooping of vine due to loss of chlorophyll, ultimately leading to drying up of vine, meanwhile underground parts of the plant rot completely. The extent of losses may vary from 5 to 90 per cent due to foot rot (Dasgupta et al., 2008) and 20 to 40 per cent in case of leaf rot, leading to almost total crop failure (Dasgupta et al., 2000).

Betelvine is a vegetatively propagated crop, which preserves high level of heterozygosity that is necessary for more hybrid vigor and any favourable spontaneous and induced mutants can easily be identified and propagated (Glemin et al., 2006). Unfortunately few countable efforts have been made till now in identifying resistant sources or developing resistant hybrids for foot rot.

Piper colubrinum is a distant relative of cultivated betelvine, showing high resistance to Phytophthora foot rot. Understanding the resistance mechanism involved in pathogen interaction can be aided by knowledge of differential expression profiles. At ICAR-Indian Institute of Horticultural Research, Bengaluru, interspecific hybrids have been developed by crossing Phytophthora resistant species P. colubrinum (Turner, 1971 and Purseglove et al. 1981) and P. betle to develop resistant sources for this devastating disease.

Plants have evolved with induced defense responses, which rely on the recognition of host-pathogen recognition and its associated molecular like DAMPs, MAMPs and PAMPs and pathogen effectors such as PTI (PAMP-triggered immunity) and ETI (effector-triggered immunity) (Jones and Dangl, 2006). Conversely, pathogens rely on host response to establish infection to express disease susceptibility genes aid mainly in infection of pathogen and disruption in these genes can lead to loss in host-pathogen compatibility (Chris and Frank, 2014). Recently, there has been increased focus on studying identifying and utilizing susceptible genes in developing durable broad-spectrum resistance (Ziidi et al., 2018). RNA-seq through exploring genomic tools are powerful tools which facilitate the identification of candidate genes for desired trait.

We earlier screened 13 interspecific hybrids for P. nicotianae infection and found IIHRPBIH9 identified as a resistant hybrid and the susceptible variety IIHRBVP170 as models (Mohankumar et al., 2022). RNA-seq was employed to analyze the gene expression profiles of leaves from both the resistant hybrid and susceptible variety at 0 and 72 hours after inoculation (hai) with P. nicotianae. Differentially expressed genes in relation to P. nicotianae infection were identified and key genes were validated. Our findings provide insights into the resistance mechanisms of betelvine response to the pathogen and provide a potentially valuable resource for future plant development.

PCA was performed using iGEAK (Choi and Ratner, 2019) to reduce data dimensionality and distinguish between the resistant and susceptible betelvine samples. The PCA results showed clear separation between resistant interspecific hybrid and susceptible variety of both inoculated and uninoculated samples, indicating substantial transcriptome variation in response to Phytophthora infection. The largest distance observed was between resistant uninoculated and susceptible uninoculated samples, suggesting significant transcriptome differences (Supplementarty Fig. 1).

Calculations of ΔΔCt values for individual genes based on the ΔCt values of susceptible, resistant, and inoculated samples obtained from their RPKM (reads per kilo base per million) values revealed that an average of 5,000 transcripts were differentially expressed in any one of the four comparisons performed at 72 hai. Between uninoculated and inoculated resistant interspecific hybrids, 8,561 genes were differentially expressed, with 4,619 up-regulated and 3,942 down-regulated. In the susceptible variety, 10,188 genes were differentially expressed, with 5,543 up-regulated and 4,645 down-regulated. Between susceptible inoculated and resistant inoculated samples, 16,547 genes were differentially expressed, with 9,010 up-regulated and 7,537 down-regulated (Supplementary Figs. 2, 3 & 4).

Comprehensive analysis of differential gene expression, go enrichment, and pathway interactions in betelvine resistance to P. nicotianae

The Venn diagram revealed significant differential gene expression in the resistant hybrid compared to the susceptible variety, indicating transcriptional changes related to disease resistance. Specifically, 19,263 genes were differentially expressed between uninoculated and inoculated resistant and susceptible betelvine. Notably, 8,561 genes were differentially expressed in resistant uninoculated (RU) versus resistant inoculated (RI), 10,188 in susceptible uninoculated (SU) versus susceptible inoculated (SI), 16,547 in susceptible inoculated versus resistant inoculated, and 12,083 in susceptible uninoculated versus resistant uninoculated (Fig. 1a and 1b). Unique up-regulated genes were: 1,635 for RU versus RI, 3,344 for SU versus SI, 4,962 for SI versus RI, and 1,500 for SU versus RU, with 71 common genes. Unique down-regulated genes were: 2,342 for RU versus RI, 1,681 for SU versus SI, 2,713 for SI versus RI, and 1,086 for SU versus RU, with 202 common genes.

GO analysis predicted the functions of differentially expressed genes in four groups: RU versus RI, SU versus SI, SI versus RI, and SU versus RU using Blast2GO. GO annotation was performed for biological processes, molecular functions, and cellular components, revealing enriched terms in the betelvine transcriptome interaction with P. nicotianae. In the RU versus RI comparison, frequent biological processes included "other cellular process," "anatomical structure development," "response to external stimulus," and "response to stress." Frequent cellular components were "chloroplast," "cytoplasm," "cytosol," and "plasma membrane," and enriched molecular functions were "protein binding," "signaling receptor activity," and "lipid binding" (Fig. 2a).

Figure 2b shows GO classification for SU versus SI, with "biosynthetic process" and "embryo development" as predominant biological processes, "nucleus," "chloroplast," "cytoplasm," and "mitochondrion" as major cellular components, and "protein binding," "molecular function," "RNA binding," and "DNA binding transcription factor activity" as key molecular functions. For SI versus RI (Fig. 2c), biological processes included "other metabolic processes," "anatomical structure development," and "response to external stimulus." Cellular components were "chloroplast," "cytosol," and "cytoplasm," and molecular functions were "protein binding" and "signaling receptor activity." In SU versus RU (Fig. 2d), predominant GO terms were "biosynthetic process," "chloroplast," and "protein binding," with "secondary metabolic process," "peroxisome," and "signaling receptor activity" being least frequent.

KEGG databases, which include genomic, biological pathway, and disease information, were used to annotate DEGs and map them to pathways, revealing that DEGs in RU versus RI, SU versus SI, and SI versus RI comparisons were involved in several pathways. In RU versus RI, DEGs were significantly enriched in defense-related pathways like superoxide radicals degradation, jasmonic acid biosynthesis, flavonoid biosynthesis, and ethylene biosynthesis I, with the highest number in superoxide radicals degradation. Moderate enrichment was seen in L-glutamine biosynthesis II, phenylalanine biosynthesis, and abscisic acid biosynthesis, while brassinosteroids inactivation and ubiquinol-9 biosynthesis showed the least enrichment (Fig. 4a). In SU versus SI, pathways like jasmonic acid biosynthesis, superoxide radicals degradation, and glutathione-mediated detoxification II were enriched, with the most genes in glutathione-mediated detoxification II and the fewest in quercetin glycoside biosynthesis (Fig. 4b). For SI versus RI, the most enriched pathways were glutathione-mediated detoxification II, phenylpropanoid biosynthesis, and superoxide radicals degradation (Fig. 4c).

Heat maps with hierarchical clustering displayed top defense-related DEGs. In the resistant inoculated sample, peroxidase 5-like protein and pathogenesis-related protein PRB1-3 like protein were highly up-regulated, while thaumatin-like protein 1, disease resistance protein (TIR-NBS-LRR class) family, and lipid transfer protein 1 were was down-regulated (Fig. 5a). However, SI versus RI comparison (Fig. 5b), glucan endo-1, 3-beta-glucosidase, putative disease resistance protein (RGA1 & RGA4), and pathogenesis-related protein 1-like protein were up-regulated in the resistant inoculated sample. In contrast, LRR and NB-ARC domains containing disease resistance protein, pathogenesis-related protein PR2 like protein, and galactose-binding protein isoform-1 were down-regulated in the resistant inoculated sample during betelvine-P. nicotianae interaction.

High-throughput RNA sequencing (RNA-seq) is a powerful tool for identifying differentially expressed genes, assessing their levels of expression, and elucidating their regulatory mechanisms. Transcriptome analysis is widely recognized for identifying plant growth or developmental stages that are both scientifically and economically significant. Consequently, it is essential for obtaining a genome-scale map that includes the structure of transcripts, their complexity, and expression levels under specific conditions (Trapnell et al., 2010). In gene expression or RNA-seq analysis, numerous pathways are involved, although biotic stress pathways are predominantly focused on identifying disease resistance-related genes.

The interspecific hybrid IIHRPBIH9 is considered a novel source of resistance to Phytophthora foot rot due to its high level of resistance based on artificial screening. In contrast, another germplasm, IIHRBV170, which is susceptible to Phytophthora foot rot disease, was also used for RNA-seq assay. To thoroughly understand the genes and pathways involved in betelvine response to P. nicotianae infections, four samples were selected for transcriptome sequencing: both uninoculated and inoculated plants were analyzed to identify differential gene expression between the resistant interspecific hybrid and the susceptible check variety of betelvine.

The expression pattern regulated by pathogen infection enables the identification of defense-related gene families involved in betelvine immune response. R protein-mediated resistance typically correlates with a hypersensitive (HR) reaction, leading to the death of both the pathogen and the surrounding plant tissue (Hammond-Kosack and Kanyuka, 2007). Some R genes associated with disease responses include the WRKY transcription factor, CC-NBS-LRR type gene, TIR-NBS-LRR gene, and putative LRR domains (Malik and George, 2018).

Using the Illumina HiSeq platform, a total of 6.62 GB of data was obtained. This included 1.99 GB from the uninoculated susceptible variety (IIHRBV170), 1.50 GB from the uninoculated resistant variety (IIHRPBIH9), 1.19 GB from the inoculated susceptible variety, and 1.93 GB from the inoculated resistant variety. A total of 25 million reads were generated for the susceptible uninoculated sample, 19 million reads for the resistant uninoculated sample, 14.8 million reads for the susceptible inoculated sample, and 24.40 million reads for the resistant inoculated sample.

Principal component analysis showed substantial variation between P. nicotianae inoculated and uninoculated samples, clearly separating them into two groups, indicating transcriptome variation between the resistant interspecific hybrid (IIHRPBIH9) and the susceptible variety (IIHRBV170). The resistant uninoculated line exhibited the greatest distance from the susceptible uninoculated line, showing significant variation between them. A similar trend was observed in the resistant and susceptible inoculated samples, with a shorter distance between the resistant uninoculated and inoculated samples compared to the susceptible variety. This finding aligns with Mohammadbagheri et al. (2021), who reported distinct genetic groups in bell pepper in response to P. capsici, and with Zhang et al. (2002) regarding P. infestans in tomato.

A total of 8,561 genes were differentially expressed between uninoculated and inoculated resistant interspecific hybrid (IIHRPBIH9) betelvine samples. Of these, 4,619 genes were up-regulated, and 3,942 genes were down-regulated. Among them, 1,635 and 2,342 genes were highly expressed as up- and down-regulated, respectively. In the uninoculated and inoculated susceptible variety (IIHRBVP170), 10,188 genes were differentially expressed, with 5,543 genes up-regulated and 4,645 genes down-regulated. Among these, 3,344 and 1,681 genes were highly expressed as up- and down-regulated, respectively. A total of 16,547 genes were differentially expressed among the samples of susceptible inoculated and resistant inoculated plants, with 9,010 genes up-regulated and 7,537 genes down-regulated. Additionally, 4,962 and 2,713 genes were uniquely expressed as up- and down-regulated, respectively. These results are consistent with Rabuma et al. (2022), who identified 57 DEGs related to defense response in bell pepper, and with Hao et al. (2016), who found higher expression levels of transcripts in resistant P. flaviflorum compared to susceptible P. nigrum L. in response to P. capsici in black pepper. Similar findings were reported by Song et al. (2022) in tobacco's response to P. nicotianae.

A Venn diagram recorded 19,263 DEGs between uninoculated and inoculated resistant interspecific hybrid (IIHRPBIH9) and susceptible variety (IIHRBV170) of betelvine. Specifically, 8,561 genes were found between RU versus RI, 10,188 genes in SU versus SI, 16,547 genes in SI versus RI, and 12,083 genes in SU versus RU. In all four comparisons, 1,635, 3,344, 4,962, and 1,500 up-regulated genes were unique to RU versus RI, SU versus SI, SI versus RI, and SU versus RU, respectively, with 71 common DEGs across all four comparisons. Similarly, 2,342, 1,681, 2,713, and 1,086 down-regulated genes were unique to RU versus RI, SU versus SI, SI versus RI, and SU versus RU, respectively, with 202 common DEGs. These results align with Rabuma et al. (2022), who reported common DEGs in leaf samples of C. annum L., and with Song et al. (2022) in tobacco.

Gene ontology (GO) enrichment analysis was performed on four groups of differentially expressed genes: RU versus RI, SU versus SI, SI versus RI, and SU versus RU. GO annotation was conducted for all three GO terminologies: biological process, molecular function, and cellular component. In the biological process category, the most frequent GO enrichments were "anatomical structure development, response to external stimulus, and response to stress." In molecular functions, "protein binding and signaling receptor activity" were most enriched. In the cellular component category, the predominant terms were "chloroplast, cytoplasm, cytosol, followed by plasma membrane." These results are consistent with Rabuma et al. (2022), who reported similar dominant GO terms in C. annum L., and with Dia et al. (2017), who highlighted the plant defense mechanism's role in recognizing cellular signal transduction in response to biological processes. Yoo and Sheen (2008) also emphasized the importance of plant hormone signal transduction in defense response.

Volcano plots were used to visualize the distribution of up- and down-regulated genes between resistant, susceptible uninoculated, and inoculated samples. Key genes identified in this study included peroxidase 5-like protein, CRT-like transporters, disease resistance-responsive proteins, disease resistance protein (TIR-NBS-LRR class) family, pathogenesis-related proteinase inhibitor family, pathogenesis-related protein 1-like protein, cytochrome P450-like protein, and ubiquitin-like superfamily protein, which were differentially up-regulated. Conversely, genes like pathogenesis-related protein 2-like protein, LRR receptor-like serine/threonine-kinase, cotton fiber protein, and UDP-glucosyltransferase 76B1-like protein were differentially down-regulated against P. nicotianae infection in both the resistant interspecific hybrid (IIHRPBIH9) and susceptible variety (IIHRBV170). Boava et al. (2011) similarly identified pathogenesis-related genes involved in citrus defense against P. parasitica.

KEGG pathway analysis revealed that most DEGs in the RU versus RI comparison were significantly enriched in defense-related compound synthesis pathways, including superoxide radicals degradation, jasmonic acid biosynthesis, flavonoid biosynthesis, ethylene biosynthesis I, and phenylalanine biosynthesis. These pathways respond to various environmental signals (biotic and abiotic stress), increasing essential enzymes' activity in the phenylpropanoid pathway. These findings align with previous work by Li et al. (2020) in C. annum L., who reported significant enrichment in secondary metabolite biosynthesis pathways. Wang et al. (2015) also highlighted the importance of secondary metabolites and terpenoids in response to P. capsici, and Jin Kim and Kook Hwang (2000) emphasized the role of ethylene biosynthesis pathways in defense.

A heat map representing the expression of defense-related genes in Phytophthora uninoculated and inoculated resistant interspecific hybrid and susceptible variety showed significant up-regulation of genes like pathogenesis-related protein PRB1-3, disease resistance responsive protein, and peroxidase 5-like protein in the resistant inoculation sample. Conversely, TIR-NBS-LRR class proteins and thaumatin-like protein 1 family were down-regulated. In the comparison between SU versus SI, genes like putative receptor-like protein kinase, heat shock protein 70 (HSP 70) family protein, and pathogenesis-related protein 2-like protein (PR 2) were up-regulated in the susceptible inoculated sample, while NB-ARC domain-containing disease resistance proteins were downregulated in the uninoculated susceptible sample. The expression level between SI and RI showed up-regulation of genes like glucan endo-1, 3-beta-glucosidase and pathogenesis-related protein 1-like protein, whereas pathogenesis-related protein PR2-like protein, LRR, and NB-ARC domains containing disease resistance protein, and galactose-binding protein isoform-1 were down-regulated in the resistant inoculation sample. Zhang et al. (2017) also reported similar up-regulation of heat shock protein HSP70 and PHE ammonia lyase 1 in soybean's defense against P. sojae.

Our study used RNA-seq and advanced bioinformatics tools to understand how betelvine responds to Phytophthora nicotianae. We found many differentially expressed genes and pathways involved in defense, both in resistant (IIHRPBIH9) and susceptible (IIHRBV170) varieties. Notable differences were seen between inoculated and uninoculated samples, and between resistant and susceptible variety, highlighting the complex interaction between plant and pathogen. Principal Component Analysis showed distinct sample clustering based on infection status and genetics, revealing significant transcriptional changes due to Phytophthora infection. Key defense genes like peroxidase 5-like protein and pathogenesis-related proteins showed significant up-regulation, aligning with previous research highlighting their importance in plant defense. Gene Ontology and KEGG pathway analyses revealed stress response, cellular processes, and defense-related pathways like superoxide radicals degradation and jasmonic acid biosynthesis as crucial during Phytophthora infection. Validation via qRT-PCR confirmed the up-regulation of selected defense genes, supporting our transcriptomic analysis. Although minor differences were observed between RNA-seq and qRT-PCR, our study provides a comprehensive understanding of betelvine defense mechanisms against Phytophthora. This insight can guide targeted strategies for enhancing disease resistance in this economically important crop. Further research is needed to explore the functional roles of identified genes and pathways, aiding their incorporation into breeding programs aimed at bolstering betelvine resilience to Phytophthora diseases.