The P. versicolora larvae and adults were obtained from Sha Lake Park in Wuhan, China and reared with fresh leaves of willows in laboratory at 28 ± 1°C, under a photoperiod of 12:12 h (light: dark) and 70 ± 5% RH (relative humidity). Strain of P. aeruginosa was stored at – 80°C and cultured in Luria-Bertani sterile liquid medium at 37°C for 12 h. OD₆₀₀ was measured to calculate the concentration of P. aeruginosa and adjusted it by sterile water to approximately 2.0 for feeding assay. Fresh willow leaves were cut into rectangle and areas were 6 cm². The bacteria solution was added onto the surface of fresh willow leaves for 60 μl and painted repeatedly for several times by plastic coating bar. Fresh willow leaves were air dried with the bacteria concentration was 10 μl/cm² and then fed to the 2nd instar P. versicolora larvae that had been deprived of food for 4 h. Synchronized groups of larvae were selected and divided into three groups (each group containing 20 individuals and serving as a biological replicate).

For all pathogenic infection bioassays, the concentrations of the bacteria solutions were consistent with those described above. The experiments were conducted in the artificial climate chamber under controlled conditions as previously described, with willow leaves replaced and survival rate recorded daily.

Three time points were selected to collected samples, first day, time to reach 10% and 50% mortality. Time to reach 10% and 50% mortality were the time point that the mortality rates of larvae were nearly 10% and 50% (Ma et al., 2021). Considering the dynamics of pathogenicity and lethality in P. aeruginosa, the infected group was sampled at day 1, day 2 (time to reach 10% mortality), day 5 (time to reach 50% mortality) after infection and the H₂O treatment group was sampled at the same time points as controls. Meanwhile, the samples of day 5 were prepared for RNA-seq and the samples of day 1, day 2 and day 5 were prepared for RT-qPCR. Five live larvae in each group were pooled as one biological replication. The whole body of larva was sampled in each treatment.

After collection, all samples were promptly frozen in liquid nitrogen and kept at a temperature of – 80°C for RNA extraction. The total RNA of all collected samples was extracted using Trizol Reagent (Invitrogen, USA) according to protocol. The quality of RNA was checked with a NanoDrop-2000 (Thermo Scientific, USA). The Illumina sequencing of the samples was performed by Majorbio (China). The cDNA library was synthetized with NEBNext ® Ultra mRNA Library Prep Kit for Illumina (NEB, USA) following manufacturer’s instructions and previous study (Liu et al., 2021).

After removing the raw reads containing the adaptor sequences, low-quality reads (Quality score < 20), and repeated reads, the clean reads were obtained. The transcriptome was assembled according to these clean reads by using Trinity 2.8.5 to generate a set of transcripts. These transcripts were annotated according to the following databases: non-redundant (NR) protein database, Swiss-prot, KEGG, KOG/COG, and a search conducted in the National Center for Biotechnology Information (NCBI). A differentially expressed gene (DEG) analysis was performed based on the significance level using the R package, DEGseq2 (p-adjust < 0.05, fold change > 2).

Different stages of P. versicolora including eggs, 1st to 3rd instar larvae, pupae, male and female adults were collected respectively. Different tissues of P. versicolora including epidermis, foregut, midgut, hindgut and Malpighian tubule were dissected from 3rd instar larvae, respectively (Li et al., 2022). Each sample was studied using three biological replicates. All samples were immediately frozen in liquid nitrogen and stored at – 80°C until used. Following the extraction of RNA, a reverse transcription reaction was carried out using 1 µg of total RNA. The HiScript III RT SuperMix for qPCR (+gDNA wiper) (Vazyme, China) was utilized for cDNA synthesis following the guidelines from the manufacturer. The housekeeping gene RPS18 was chosen as a reference gene for internal control. Reaction mixtures, reaction protocols and gene expression calculation methods for RT-qPCR have been reported before (Liao et al., 2023).

First, dsRNA template was prepared by PCR using specific primers containing the T7 promoter sequence ( Supplementary Table S1 ;10 µM) and purified with a gel extraction kit (Omega, China). Second, dsRNA was synthesized in vitro using the T7 RNAi Transcription Kit (Vazyme, China) according to the manufacturer’s instructions and previous study (He et al., 2020). Following the production of dsRNA, two continuous experiments were conducted for synergistic toxic bioassays. First, 1st P. versicolora larvae were deprived of food for 4 h before the start of RNAi. Identical amounts (96 ng, in 50 μl H₂O) of dsRNAs were applied onto fresh willow leaves, covering a consistent area of 6 cm², thus adjusting the concentration of dsRNA to 16 ng/cm². The insects were provided with an adequate amount of fresh leaf coated with dsRNA for RNAi or H₂O for control. The leaves were replaced daily for three days to ensure that the foliar-applied dsRNA remained stable and target genes were knocked down. Samples were collected on day 3 for RT-qPCR to evaluate the silencing efficiency of RNAi. Second, synchronized groups of 2nd larvae that dsRNA-treated or not were selected, divided into three groups (each group containing 10 individuals and serving as a biological replicate) for pathogenic infection bioassays, and the operations were the same as mentioned above. Samples were collected on day 4 for RT-qPCR to analysis the expression patterns of immune-related genes.

The Kaplan-Meier method was utilized to analyze survival curves. The log-rank test was employed to assess the significance of differences between the two groups. The t-test was utilized to analyze the data for bacteria induced genes expression and RNAi efficiency by RT-qPCR. The ANOVA was employed to analyze the data for the expression of immunity-related genes by RT-qPCR. A value of P < 0.05 meant significantly different. All statistical analyses were performed using SPSS version 22. Figures were drawn with GraphPad Prism 8.

The survivorship of larvae infected by P. aeruginosa was significantly lower than that of control group ( Figure 1A ), demonstrating that P. versicolora is very sensitive to this pathogen. After P. aeruginosa infection, the mortality rates of P. versicolora on day 2 and day 5 were 8.3% and 56.7%, respectively, which were close to 10% and 50%. In comparison, the P. versicolora exhibited high survivorship (up to 81.4%) at day 5 in the control group ( Figure 1A ). Therefore, including the first day after P. aeruginosa infection, three time points (day 1, day 2, day 5) were selected for further study.

The cDNA library from P. versicolora larvae, either infected or not infected with P. aeruginosa on day 5, was constructed using the Illumina NovaSeq platform for next-generation sequencing. A total of 59,629,750 clean reads were acquired, with a Q20 percentage of 97.84%. About 51,891 unigenes were discovered, with a combined length of 52453360 bp and an N50 length of 2127 bp ( Table 1 ). Statistics showed that 75.2% of the 51,891 unigenes were lower than 1000 bp in length ( Figure 1B ). In total, 20,439 unigenes were matched to entries in the NCBI NR protein database (http://www.ncbi.nlm.nih.gov/protein (accessed on 10 November 2023)) by a BLASTX search.

Overall, the differentially expressed unigenes were categorized into six functional groups based on their KEGG annotation, including environmental information processing (156), human diseases (667), genetic information processing (187), cellular processes (158), organismal systems (347), and metabolism (318). In the functional group of organismal systems, the most differentially expressed genes were found to be associated with the immune system (42) and the endocrine system (38). In the functional group of environmental information processing, the most differentially expressed genes were found to be associated with the signal transduction (62) ( Figure 1C ). The results indicating that some unigenes in these sub-categories might have a connection with chemosensory behavior and immunity in insects.

The transcriptomic studies revealed 11 candidate OBP genes, 2 of which were differentially expressed. PverOBP15 and PverOBP18 genes were up-regulated in larvae infected with P. aeruginosa, when compared to the control group ( Figure 2A ). In order to investigate whether the transcript level of PverOBP15 and PverOBP18 could be triggered by P. aeruginosa infection, the expression profiles of these genes were conducted using RT-qPCR at the three previously mentioned time points. The results revealed that the PverOBP15 gene expression was significantly increased by P. aeruginosa infection comparing to control only on day 2, but not day 1 and day 5 ( Supplementary Figure S1 ). Unlike PverOBP15, the expression of PverOBP18 was significantly downregulated on day 1. As time went on, the expression of PverOBP18 was upregulated and significance showed on day 5 ( Figure 2B ).

The relative expression of PverOBP15 and PverOBP18 transcripts in different body segments and developmental stages were investigated using RT-qPCR. PverOBP15 transcript was expressed highest in 3rd larvae and female adults, lower in pupae, and lowest in eggs, 1st larvae, 2nd larvae and male adults ( Supplemetary Figure S2A ). Unlike PverOBP15, PverOBP18 transcript was expressed highest in pupae, followed by female adults, and lower in 1st-3rd larvae and male adults with lowest in eggs (P < 0.05, LSD test) ( Figure 3A ).

In different tissues, PverOBP15 transcript was expressed highest in the fat body, followed by the foregut, and lowest in the epidermis, midgut, hindgut, and Malpighian tubule ( Supplemetary Figure S2B ). Unlike PverOBP15, the expression of PverOBP18 transcripts was highest in the epidermis, lower in the midgut, hindgut, fat body, and Malpighian tubule, and lowest in foregut (P < 0.05, LSD test) ( Figure 3B ).

For the purpose of investigating the role of PverOBP18 and PverOBP15 in response to P. aeruginosa infection, the dsRNAs for gene knockdown of PverOBP18 and PverOBP15 were synthesized in vitro and fed to P. versicolora larvae. Analyzing the efficiency of RNAi in P. versicolora, the results revealed that both PverOBP15 and PverOBP18 expression were reduced by nearly 65% after fed dsRNA for 3 days ( Supplementary Figure S3A ; Figure 4A ). Survival curves showed that the larval mortality of dsPverOBP18 + ; P. aeruginosa treated group was significantly higher than that of the P. aeruginosa treated group, while the mortality of dsPverOBP18 treated group and untreated group were relatively low. Unlike dsPverOBP18 treatment, no significant difference was found between the dsPverOBP15 + P. aeruginosa treated group and ; P. aeruginosa treated group. The results revealed a reduce in the resistance of P. versicolora to bacterial infection after the knockdown of PverOBP18, but not PverOBP15 ( Supplementary Figure S3B ; Figure 4B ).

To determine which immunity-related genes were involved in PverOBP18 mediated resistance to P. aeruginosa, genes in immune effectors (Attacin, Defensin, Lysosome), signal modulation (Serpin), recognition receptors (PGRP) and Toll pathway (Toll-1) were selected to analysis the expression patterns by RT-qPCR. The results revealed that PGRP gene expression in the P. aeruginosa treated group was significantly increased, compared with the dsPverOBP18 + P. aeruginosa treated group, dsPverOBP18 treated group and untreated group. Attacin gene showed a similar expression patten like PGRP gene, although there were no notable alterations detected between the P. aeruginosa treated group and the dsPverOBP18 + P. aeruginosa treated group. Furthermore, there were no notable alterations detected in the expression of other four immunity-related genes ( Figure 5 ).