The study employed a homogenous population of T. palmi derived from a single adult female. The isofemale line has been maintained on eggplant since 2018. The identity of the population was further verified by morphometric keys (Bhatti, 1980; Cluever and Smith, 2017) and mitochondrial cytochrome oxidase subunit I (mtCOI) nucleotide sequencing.

The initial GBNV inoculum was collected from a pure culture maintained at the Advanced Center for Plant Virology, Indian Agricultural Research Institute (IARI), New Delhi. Sap inoculation of healthy cowpea plants at the two-leaf stage was done with GBNV under insect-proof conditions as described by Ghosh et al. (2021). The presence of GBNV infection in cowpea was verified using RT-PCR using GBNV-specific primers AG109F-AG110R (Table 1).

T. palmi differential gene expression data in response to GBNV and CaCV (Widana Gamage et al., 2018; Mahanta et al., 2022) were analyzed for the selection of potential gene targets in response to orthotospovirus infection. The expression of the Btk29A and COL3A1 in T. palmi was highly abundant post-GBNV and CaCV acquisition. Therefore, T. palmi Btk29A and COL3A1 were targeted for silencing using the cognate dsRNA. The available sequences of Btk29A and COL3A1 were retrieved from NCBI and aligned in MEGAX (Kumar et al., 2018) to find the conserved regions. SiRNA Wizard web tool (accessed on 10-01-2022) ¹ was used to predict putative siRNAs from the conserved regions of T. palmi Btk29A and COL3A1. The region identified for dsRNA design was subjected to evaluation for off-target effects in humans, mice, birds, butterflies, bees, ants, and plants using the NCBI BLASTn tool. ² The region in both genes that showed the maximum number of putative siRNAs without having any off-target effect was selected for dsRNA synthesis. The selected regions were 1,166–1,444 for Btk29A and 213–469 for COL3A1 genes.

The dsRNA stretches were amplified using primer pairs listed in Table 1. The PCR conditions were optimized using a gradient PCR. Total DNA was isolated from T. palmi using a CTAB extraction buffer as described by Jangra and Ghosh (2022) and used for amplification in a thermal cycler. A 25 μL PCR mixture comprised of 1X PCR buffer (Thermo Fisher Scientific, United States), 0.4 μM each forward and reverse primer (GCC Biotech, India), 0.26 mM dNTP mix (Thermo Fisher Scientific), ~ 60 ng template DNA, and 2 U of DreamTaq DNA polymerase (Thermo Fisher Scientific) were used. PCR was performed in a T100 thermocycler (Bio-Rad, United States) with an initial denaturation at 94°C for 5 min, 35 cycles of denaturation at 94°C for 40 s, annealing at 55°C for 40 s, extension at 72°C for 40 s, followed by a final extension at 72°C for 10 min. The PCR products were resolved on a 2% agarose gel, stained with GoodView (BR Biochem, India), and visualized under a gel documentation system (MaestroGen Inc., Taiwan). The amplified PCR products were sequenced for confirmation purposes.

The PCR amplicons were cloned between two T7 promoters in the L4440 expression vector (Addgene, United States) and transformed into RNase III deficient E. coli HT115 cells (University of Minnesota, United States). The presence of the desired inserts was confirmed by colony PCR and restriction digestion. Production of dsRNA was induced by adding 0.8 M isopropyl-β-D-1-thiogalactopyranoside (IPTG, HiMedia, India) to the culture of recombinant E. coli HT115 cells and incubating overnight at 37°C in a shaking incubator (Gnanasekaran et al., 2019, 2021). Total RNA was extracted from HT115 cells using Trizol reagent (Invitrogen, United States) according to the manufacturer’s protocol. DsRNA from the total RNA was purified using 1 U of DNase I (RNase-free, Thermo Fisher Scientific) and 1 U of RNase A (DNase and protease-free, Thermo Fisher Scientific) for 1 h at 37°C in the presence of 500 mM sodium chloride. The enzymes were inactivated by chloroform extraction as described by Chakraborty and Ghosh (2022). The quality and concentration of the purified dsRNAs were assessed in a spectrophotometer (Nano-300, Genetix Biotech Asia, India) and visualized on a 2% agarose gel. A dsRNA targeting Bemisia tabaci transducer of erbB2.1 (BtTOB1; Thesnim et al., 2023) and not specific to T. palmi based on nucleotide homology screening was used as a negative control.

The purified Btk29A and COL3A1 dsRNAs were administered to T. palmi orally using an artificial feeding setup as reported by Priti et al. (2022). Briefly, a diet was prepared by mixing 50 mg/mL pine pollen (Lost Empire Herbs, United States) extract with 1% sucrose solution and 0.0001% methylene blue tracker dye (Bio Basic, United States). The diet was supplemented with 3 μg/μL dsRNA (total volume 300 μL). The diet with dsRNA was filled into UV-sterilized detachable caps of 2 mL microcentrifuge tubes and covered with stretched parafilm. Around 20 adults or larvae were taken in the perforated microcentrifuge tubes. The cap filled with diet was closed gently. T. palmi were allowed to feed for 24 h at 28 ± 1°C temperature, 60 ± 10% relative humidity, and 16 h light–8 h dark. In the control sets, a diet without dsRNA and a diet with BtTOB1 dsRNA (non-specific to T. palmi) were supplied. Several such replicates were maintained to get enough dsRNA-treated and untreated thrips to assess the mortality, expression of the target genes, and virus titer in thrips.

The expression of T. palmi Btk29A and COL3A1 was assessed 24 h post-dsRNA feeding by RT-qPCR. β-tubulin was taken as the endogenous control. Table 1 lists the primer pairs used in RT-qPCR. Five surviving T. palmi in three biological replicates for each gene under study were considered for estimating the mRNA expression level. Total RNA was extracted using NucleoSpin RNA XS (Macherey-Nagel, Germany) according to the manufacturer’s protocol. Total RNA was quantified in a Nano-300 Micro Spectrophotometer (Genetix Biotech Asia), and cDNA was synthesized using FIREScript RT cDNA synthesis kit (Solis BioDyne, Tartu, Estonia). One μg RNA template, 5 μM oligo dT primers, 500 μM dNTP mix, 2 μL of 1 X reaction buffer, 10 U FIREScript RT, and 1 U RiboGrip RNase inhibitor were used in 20 μL reaction mixture. cDNA was synthesized in a T100 thermocycler (BioRad, United States) by reverse transcription at 50°C for 60 min, and enzyme inactivation at 85°C for 5 min. The qPCR assay was performed in an Insta Q48M real-time PCR (Himedia, India). The 20 μL qPCR mixture included 1X GoTaq qPCR Master Mix (Promega, United States), 300 nM CXR passive reference dye, 0.25 μM of each forward and reverse primer, and 2 μL of template cDNA. The following thermal cycling was followed: initial denaturation at 94°C for 5 min, 40 cycles of 94°C for 40 s, 56°C for 40 s, and 72°C for 40 s. A melting curve analysis was carried out after each reaction to assess the specificity of the amplicons. Three biological and two technical replicates were used in the RT-qPCR. A non-template reaction is used as a negative control. The relative expression of Btk29A and COL3A1 in dsRNA-fed T. palmi was measured following the 2^−ΔΔCT method (Livak and Schmittgen, 2001). The statistical analysis was performed and graphs were generated in Microsoft Excel 2019. Expression of T. palmi Btk29A and COL3A1 in BtTOB1 dsRNA-fed thrips served as the negative control.

T. palmi individuals that fed on the artificial diet were identified by a blue tinge in their abdomen and were exclusively considered to assess the effect of dsRNA on survival. The mortality of T. palmi was recorded 24 h post-dsRNA feeding. The whole experiment was replicated nine times for each treatment. The mean mortality in dsRNA-exposed T. palmi was recorded and compared with the thrips fed on the dsRNA-free diet and diet with BtTOB1 dsRNA. Tukey’s test was performed with XLSTAT 2014.5.03 to differentiate means across categories with a 95% confidence interval. The morphological changes, if any, post-dsRNA feeding were also checked under a stereo microscope (M205 FA, Leica, Germany) and captured in a Leica DFC425 C.

The eggs of T. palmi were collected using the artificial oviposition setup reported by Jangra et al. (2020). The eggs were incubated at 28°C on wet tissue paper. The newly hatched first instar larvae (L1) were collected using a Camel hairbrush and allowed to feed on a diet mixed with dsRNA for 2 h as outlined above. T. palmi fed on the dsRNA-free diet and diet with BtTOB1 dsRNA were used as negative controls. The dsRNA-fed and non-fed L1 were allowed to acquire GBNV as reported by Ghosh et al. (2021). In brief, T. palmi L1s were released for a 12-h acquisition access period (AAP) on detached GBNV-infected cowpea leaf at 28 ± 1°C and 60 ± 10% relative humidity. Five live L1s at three biological replicates were collected in a microcentrifuge tube post-GBNV acquisition and used to quantify the virus copies. The GBNV titer acquired by dsRNA-fed and non-fed T. palmi larvae were estimated by absolute quantification by RT-qPCR using the primer pair AG335F-AG336R (Table 1). Total RNA extraction and cDNA synthesis was performed as outlined above. One μg RNA template, 5 μM random primers, 500 μM dNTP mix, 2 μL of 1 X reaction buffer, 10 U FIREScript RT, and 1 U RiboGrip RNase inhibitor were used in 20 μL reaction mixture. The cDNA was synthesized in a T100 thermal cycler with primer annealing at 25°C for 10 min, reverse transcription at 50°C for 60 min and enzyme inactivation at 85°C for 5 min. 10 μL of 1X GoTaq qPCR Master Mix (Promega), 300 nM CRX reference dye, 0.25 μM of each forward and reverse primer, and 2 μL template cDNA were used in 20 μL qPCR reaction mixture. Thermal cycling was performed as initial denaturation at 95°C for 5 min, 35 cycles at 95°C for 25 s, 54°C for 25 s, and 72°C for 30 s. A melting curve analysis was performed to ensure the reaction’s specificity. Three biological and two technical replicates were used for each treatment. A standard curve (y = −3.3251x + 28.141, efficiency = 99.86%) of GBNV using primer pair AG335F-AG336R reported by Priti et al. (2022) was used for the quantification of GBNV copies in T. palmi.

The viral titer was estimated by fitting the mean CT values in the standard curve. The following formula was used to calculate the virus copy number in Microsoft Excel 2019. N = (x × 6.022 × 10²³)/ (n × 340 × 10⁹), where N = number of viral copies, x = amount of amplicon in ng, and n = length of linearized plasmid.

One-Way ANOVA was performed to determine the mean differences in virus copies among treatments at p ≤ 0.05 in XLSTAT 2014.5.03. The effect of Btk29A and COL3A1 dsRNAs feeding on survival and mRNA expression at this stage (L1) was also recorded as mentioned above.

T. palmi adults had quadrangular heads with seven segmented antennae. At the top of the head, three brick-red ocelli were present in a triangular formation. A pair of interocellar setae originated outside the ocellar triangle. Besides, the nucleotide sequence of a 595 bp portion of mtCOI was amplified using primer pairs LCO 1490 and HCO 2198 (Folmer et al., 1994) and was 100% identical with T. palmi. The sequence can be retrieved from GenBank with the accession number OP223495.

RT-PCR with GBNV-specific primers produced ~1767 bp desired amplicon from all the sap-inoculated plants. The nucleotide sequence was found to be more than 97% identical to other GBNV isolates. The sequence can be retrieved from GenBank with the accession number MN566913.

A conserved 279 nt-long (1,166 to 1,444) fragment of the Btk29A gene (1.538 nt) and a 257 nt-long (213 to 469 nt) fragment of the COL3A1 gene (7.063 nt) of T. palmi were preferred for dsRNA designing (Figure 1). The dsRNA stretches were unique to T. palmi and showed no homology with other non-target organisms such as Homo sapiens (taxid:9606), mice (taxid:10088), Lepidoptera (taxid:7088), Aves (taxid:8782), Hymenoptera (taxid:7399), Formicidae (taxid:36668), and plants (taxid:3193) in BLASTn analysis. In siRNA Wizard, the 279 nt-long T. palmi Btk29A fragment produced 3 putative siRNAs of 21 nt-long. Similarly, the 257 nt-long COL3A1 fragment predicted 13 putative siRNAs of 21 nt-long (Figure 1). A conserved 167 nt (896–1,062) stretch of B. tabaci TOB1 was used to synthesize BtTOB1 dsRNA and considered as a negative control. The BtTOB1 dsRNA was specific to B. tabaci (Thesnim et al., 2023) and showed no homology to T. palmi.

PCR with primer pair AG507F-AG508R specific to Btk29A yielded an amplicon of 279 bp (accession no. OP346052). Whereas a 257 bp (accession no. OP345816) fragment of COL3A1 was amplified in PCR with the primer pair AG188F-AG189R. On a 2% agarose gel, the Btk29A and COL3A1 dsRNAs purified from the total RNA of recombinant E. coli HT115 cells generated ~279 and ~ 257 bp bands, respectively (Figure 2). The concentration of the purified dsRNA was 621.5 ng/ μL for Btk29A and 531.0 ng/ μL for COL31A. It was 779.9 ng/μL in the case of BtTOB1 dsRNA.

The expression of T. palmi Btk29A and COL3A1 was downregulated post-dsRNA feeding. The mean logarithmic expression of T. palmi Btk29A was estimated as −5.66 in the larval stage and 0.13 in the adult stage under controlled conditions. It was 2.53 and 0.39 in larval and adult stages, respectively for T. palmi COL3A1. A 3.62-fold downregulation of T. palmi Btk29A mRNA was recorded 24 h post-feeding on a diet containing Btk29A dsRNA in contrast to T. palmi feeding on a diet without dsRNA (Figure 3). Similarly, a 3.15-fold reduction in T. palmi COL3A1 mRNA expression level was recorded in T. palmi fed on COL3A1 dsRNA. Expression of T. palmi Btk29A and COL3A1 post-exposure to BtTOB1 dsRNA was statistically non-significant with untreated control. There was no significant difference in the expression of the endogenous control, β-tubulin, between dsRNA-fed and non-fed T. palmi, indicating that the Btk29A and COL3A1 dsRNAs had a specific action on the target mRNAs. The primer pairs for Btk29A, COL3A1, and β-tubulin produced no secondary peaks in the RT-qPCR melting curve analysis, indicating that the reactions were specific.

T. palmi fitness was significantly affected post-exposure to Btk29A and COL3A1 dsRNAs under controlled laboratory settings. Feeding on Btk29A dsRNA at a concentration of 3 μg/mL for 24 h induced 60% mortality in T. palmi adults (Figure 3). Whereas the mean mortality of COL3A1 dsRNA-fed T. palmi was 50.9% compared to 5% in T. palmi fed on the diet without dsRNA. At a 95% confidence level, the mean mortality for Btk29A and COL3A1 dsRNA treatments was significant, with p-values of 0.024 and 0.013, respectively, compared to the untreated control. There was no discernible difference in mortality between T. palmi treated with BtTOB1 dsRNA and the untreated control. No immediate morphological deformities were recorded in T. palmi adults post-Btk29A and COL3A1 dsRNA exposure (Figure 3).

The GBNV titer in early instar larvae of T. palmi was affected post-silencing Btk29A and COL3A1. Silencing of Btk29A induced a 9.53-fold decrease in the GBNV titer in T. palmi. The mean GBNV copy in T. palmi that were fed on the diet without dsRNA was 2.14 × 10⁹. In contrast, the mean GBNV copy acquired by T. palmi fed on the diet mixed with Btk29A dsRNA was 2.24 × 10⁸. Surprisingly, the silencing of T. palmi COL3A1 led to a slight increase in the mean GBNV copy acquired by T. palmi. The mean GBNV copy in COL3A1 dsRNA-fed T. palmi was 2.47 × 10⁹, i.e., 0.87-fold higher than the untreated control (Figure 4). At this stage, the expression of T. palmi Btk29A and COL3A1 was downregulated by 5.23 and 2.04-fold, respectively, in dsRNA-fed T. palmi larvae as compared to larvae fed on the diet without dsRNA. There was no significant difference in GBNV titer following BtTOB1 dsRNA exposure when compared to the untreated control.