The CLRDV host species identified in a previous study, viz., cotton, Gossypium hirsutum L. cv. PHY 339 WRF (Corteva, Indianapolis, IN, USA), hibiscus, H. acetosella Welw. Ex Hiern. (Johnny’s Selected Seeds, Winslow, ME, USA), okra, A. esculentus L. cv. ‘Clemson spineless 80’ (Clemson University, Clemson, SC, USA), and prickly sida, S. spinosa L. (Azlin Seed Service, Leland, MS, USA) were used as inoculum sources in this study (Pandey et al., 2022). Two to four seeds of each plant were sown per pot in Sunshine propagation mix (SunGro Horticulture Industries, Bellevue, WA, USA) in 10-cm-diameter plastic pots (depth 8 cm). The pots were kept in insect-proof cages of size 47.5 (l) × 47.5 (w) × 93 (h) cm³ (Megaview Science Co., Taichung, Taiwan) in the greenhouse. The greenhouse was maintained at 25°C, 60% relative humidity, and 14-h L:10-h D photoperiod. The seedlings were thinned post-germination, and only one plant per pot was used. Water-soluble Miracle-Gro (Scotts Miracle-Gro Products, Inc., Marysville, OH, USA) at 0.5 g/L was used for weekly fertilization. The aphids were originally collected from cotton fields in 2017 at Tifton, Georgia, and thereafter, the population was maintained in the greenhouse at the University of Georgia, Griffin Campus, under the same conditions indicated above.

Cotton plants infected with CLRDV were originally collected from cotton fields in September 2020 at the University of Georgia, Tifton Campus, GA, USA, and maintained in the greenhouse at the above-stated conditions. Aphid-mediated CLRDV transmission to cotton seedlings was undertaken to maintain the virus inoculum source. CLRDV-infected cotton, hibiscus, okra, and prickly sida plants were obtained following the protocols described in an earlier study (Pandey et al., 2022). A. gossypii adults were provided with a 72-h AAP on CLRDV-infected cotton plants followed by a 72-h inoculation access period (IAP) on the undersurface of leaves of young seedlings at the two-true leaf stage. The CLRDV-inoculated plants were placed in aphid-proof cages under the greenhouse conditions described above. Clip cages and aphids were removed 3 days post-inoculation. The infection status of the plants was evaluated at approximately 3 weeks post-inoculation by reverse transcription–PCR as described earlier (Pandey et al., 2022).

The non-viruliferous A. gossypii colonies were maintained on all four host species, i.e., non-infected cotton, hibiscus, okra, and prickly sida plants, in separate insect-proof cages. Approximately 1,000 A. gossypii were collected from each colony and then introduced to non-infected or CLRDV-infected cotton, hibiscus, okra, and prickly sida plants. The aphids were allowed to feed for an AAP of 72 h. After 72 h of AAP on non-infected or CLRDV-infected plants, approximately 500 adults per treatment were collected, and total RNA was extracted from the collected aphid samples (Figure 1). Each treatment was biologically replicated four times. Another set of aphids (three pools of 10 aphids/per treatment) was used to confirm the acquisition of CLRDV in different treatments. The results indicated that 100% of pools of aphids collected from CLRDV-infected plants were positive for CLRDV, whereas aphids feeding on non-infected plants tested negative for CLRDV.

The total RNA was extracted from the collected aphid samples using Qiagen RNA mini-Kit (Valencia, CA, USA) as per the manufacturer’s instructions. For each sample, 40 µL of total RNA was shipped to Novogene Corporation Inc. (Sacramento, CA, USA), and the rest of the total RNA was stored at −80°C for validation. Quality control (QC) of RNA samples was accomplished by preliminary quantitation using a NanoDrop and testing for RNA degradation and contamination via agarose gel electrophoresis. Then, RNA integrity (RIN) was assessed using Agilent 2100. Three samples failed the QC test. The samples that passed QC (RIN value >6.8 and concentration of >20 ng/μL) were used for cDNA library preparation. The library preparation began with enriching mRNA using oligo(dT) beads and removing rRNA using the Ribo-Zero kit. Then, the mRNA fragmentation was followed by first- and second-strand cDNA synthesis. Subsequently, adaptor ligation and PCR enrichment were performed for cDNA library generation. Finally, the library quality was assessed using a Qubit 2.0. NovaSeq 6000 Sequencing System (Illumina, San Diego, CA, USA) with the NovaSeq paired-end 150 sequencing platform was used for sequencing the libraries that passed QC.

FastQC v0.11.9 and multiQC v1.11 were used to assess the quality of raw reads before and after trimming (Andrews and FastQC, 2010; Ewels et al., 2016). The adapters were removed by using Trimmomatic v0.39 with the default setting (Bolger et al., 2014). Bowtie2 v2.4.1 was used with default mapping parameters to map the trimmed reads with the reference A. gossypii transcriptome (Langmead and Salzberg, 2012; Quan et al., 2019). RSEM v1.3.3 was used to obtain gene count estimates of the mapped reads (Li and Dewey, 2011). Fragments per kilobase million (FPKM) were determined using a custom R script with the following R libraries: dplyr, tidyverse, and stringr on R v4.1.0 (R Core Team, 2021). DESeq2 compared the gene counts from non-viruliferous aphids with viruliferous aphids to identify DEGs. Genes that had log2fold changes |LFC| ≥ 1 and a false discovery rate (FDR) ≤0.05 were identified as DEGs (Love et al., 2014). The DEGs were annotated and assigned to gene ontology (GO) classes (up to level 3) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways (Kanehisa and Goto, 2000) using the annotated A. gossypii genome (Quan et al., 2019). TopGO (https://www.bioconductor.org/packages/release/bioc/html/topGO.html) and visualize Gene Ontology (REVIGO) web tool were used for the processing and visualization of the GO terms (Supek et al., 2011).

The WGCNA software v1.70-3 was run to create co-expression modules and identify sets of DEGs expressed in a similar pattern in A. gossypii that acquired the CLRDV from alternate hosts using the R software v4.1.0 (Langfelder and Horvath, 2008; R Core Team, 2021). The gradient-independent method with the scale-independent condition of the signed R2 set to 0.90 was used to test the soft-thresholding power modules (1 to 40). The topological overlap matrix (TOM) was constructed using the interaction relationships across the co-expression modules by using correlation expression values. A dendrogram with the parameters mergeCutHeight = 0.15 and detectCutHeight = 0.995 was set to represent the TOM. Each module was represented using randomly assigned colors. The module eigengene was calculated from the first principal component of each module. The labeledHeatmap package in WGCNA software was used to show the topological overlap of co-expression modules based on eigengenes. The network analysis of the top 30 genes from the A. gossypii turquoise was visualized using Cytoscape v3.9.0 (Shannon et al., 2003). Also, the hub genes in the most correlated clusters (magenta, pink, brown, and gray) observed in viruliferous A. gossypii that acquired the virus from each host species (cotton, hibiscus, okra, and prickly sida) were visualized using Cytoscape v3.9.0.

To validate transcriptomic data, 10 DEGs of A. gossypii were randomly selected for each host species (n = 40). The expression levels of the DEGs were compared between viruliferous and non-viruliferous A. gossypii that acquired CLRDV from different hosts (cotton, hibiscus, okra, and prickly sida) by RT–quantitative polymerase chain reaction (qPCR). Primer pairs designed for each DEG using Primer3web version 4.1.0 are listed in Supplementary Table S1. The GoScript™ Reverse Transcription System (Promega, Madison, WI, USA) was used to reverse-transcribe the total RNA for each sample according to the manufacturer’s instructions. Then, the cDNA was diluted 20-fold for qPCR. The 2xGoTaq^® qPCR Master Mix (7.5 µL) (Promega, Madison, WI, USA), primers (0.3 µM), 1 µL of cDNA, and nuclease-free distilled water for a final volume of 15 µL were mixed. The QuantStudio™ 3 Real-Time PCR System (Applied Biosystems by Thermo Fisher Scientific, Waltham, MA, USA) was used for qPCR. The following qPCR conditions were used: an initial denaturation step at 95°C for 3 minutes followed by 40 cycles at 95°C for 15 s and 60°C for 1 minute. Three technical replicates for each sample were used, and the melting curve analysis was conducted to evaluate the specificity of the fluorescence signal. The expression level of each gene was normalized to the expression level of elongation factor 1α (EF1α)—an A. gossypii reference gene. The 2^−ΔΔCt method was used to calculate the relative expression of DEGs (Ma et al., 2016).

Three to four biological replicates were included per treatment on each host, resulting in seven cDNA libraries constructed for A. gossypii on each host (cotton, hibiscus, and prickly sida plants), whereas eight libraries were constructed for A. gossypii on okra. Hence, a total of 29 libraries were constructed. Raw read pairs for the generated libraries ranged from nearly 19 to 34 million. After trimming and removing the reads that aligned with the ribosomal RNA and the mitochondrial genome, 19 to 33 million reads were retained (Table 1). A. gossypii cleaned read pairs from different libraries (63 to 82%) were mapped to the A. gossypii transcriptome (Table 1).

The reads obtained from viruliferous and non-viruliferous A. gossypii samples from different host species were normalized and clustered using FPKM and principal component analysis (PCA) for comparison. The PCA clustered viruliferous A. gossypii samples separately from the non-viruliferous samples for all four host species (Supplementary Figures S1A–D).

Out of 14,134 annotated genes, a total of 750 (622 overexpressed and 128 underexpressed), 310 (168 overexpressed and 142 underexpressed), 1,193 (548 overexpressed and 645 underexpressed), and 689 genes (432 overexpressed and 257 underexpressed) were differentially expressed in viruliferous aphids that acquired CLRDV from infected cotton, hibiscus, okra, and prickly sida plants, respectively (Figure 2; Supplementary Figure S2).

RT-qPCR with 10 randomly selected DEGs per A. gossypii host (n = 40) was performed to validate RNA sequencing-based differential gene expression results (Supplementary Table S1). The expression trends of the randomly selected A. gossypii DEGs from RNA sequencing and RT-qPCR were highly consistent for all four host species (Supplementary Figures S3A–D).

Of the 2,942 DEGs in aphids that acquired CLRDV from different host species, only four genes were found to be differentially expressed in common (Figure 3). Two common DEGs were annotated as uncategorized proteins, whereas the other two were functionally annotated. Both the unknown DEGs were overexpressed in A. gossypii on all four hosts, whereas differences in the direction of expression were observed for the annotated common DEGs. The annotated common DEGs were juvenile hormone acid O-methyltransferase and heat shock protein. They were overexpressed in A. gossypii that acquired the virus from CLRDV-infected cotton and okra, whereas they were underexpressed in A. gossypii that acquired CLRDV from infected hibiscus and prickly sida (Table 2).

Only 320 of the 750 DEGs in A. gossypii that acquired CLRDV from infected cotton were assigned functional groups under three classification systems: biological process (313 genes), molecular function (267 genes), and cellular component (287 genes). Fifty-three GO terms were assigned under the biological process category, of which only two terms (microtubule-based process and ATP metabolic process) were significant (Figure 4A; Supplementary Table S2). Thirty-five GO terms were assigned under the molecular function category, only one (a structural constituent of the cytoskeleton) of which was significant (Figure 4B; Supplementary Table S2). In the cellular component category, 30 GO terms were identified, and only one (cAMP-dependent protein kinase complex) was significant (Figure 4C; Supplementary Table S2). Similarly, the DEGs identified in aphids that acquired CLRDV from hibiscus, okra, and prickly sida were assigned functional groups under three classification systems (Supplementary Figures S4–S6; Supplementary Tables S3-S5). The categorization of these genes was used to identify the DEGs associated with virus–vector interactions.

The co-expression of genes from A. gossypii adults that acquired CLRDV from cotton, hibiscus, okra, and prickly sida was evaluated. WGCNA, which clusters genes into modules based on weighted gene–gene interactions, was used to evaluate co-expression. For A. gossypii, 13 modules with 19 to 180 genes in each module were identified (Figure 5A; Supplementary Table S6), and Pearson’s correlation coefficient analysis showed the connections between the four CLRDV hosts. The heatmap visualized overall patterns of co-expression of (viruliferous/non-viruliferous) aphid–host relationships (Figure 5B). A. gossypii interactions for the largest module, MEturquoise, were checked for top interacting genes among the 30 identified genes (Figure 5C). The top four most highly connected genes were XM_027993669.1 (trichohyalin-like), XM_027997289.1 (uncharacterized protein), XM_027997209.1 (glucose dehydrogenase [FAD, quinone]-like), and XM_027983368.1 (neuroendocrine convertase 1-like) with interconnectivity scores of 58.99, 57.81, 53.87, and 52.58, respectively (Supplementary Table S6).

Four modules (magenta, pink, brown, and gray) were highly correlated with viruliferous A. gossypii that acquired CLRDV from infected cotton, hibiscus, okra, and prickly sida plants, respectively. The magenta module associated with viruliferous A. gossypii that acquired CLRDV from infected cotton plants contained 34 genes. Similarly, pink, brown, and gray modules associated with A. gossypii that acquired CLRDV from infected hibiscus, okra, and prickly sida plants included 40, 130, and 129 genes, respectively. The maximum connectivity in magenta, pink, and brown modules were 8.9, 8.8, and 20, respectively. However, the connectivity in the gray module was less than one. Hence, only magenta, pink, and brown modules were considered to identify candidate genes related to virus interactions and transmission (Supplementary Table S6).

Most genes in the magenta module were overexpressed only in viruliferous A. gossypii that acquired CLRDV from infected cotton plants, and 10 genes were identified as hub genes based on their high connectivity values (Figure 6A). Some of the hub genes were predicted to encode tubulin-β (XM_027988533.1 and XM_027985455.1), tubulin-α (XM_027988908.1), GPI-anchored protein (XM_027997066.1), U1 small nuclear ribonucleoprotein (XM_027985431.1), mucin-7-like (XM_027997067.1), and dynein beta chain (XM_027985730.1) (Supplementary Table S6). The tubulin-α and tubulin-β genes were associated with cellular responses following virus acquisition (Table 3).

Similarly, most genes in the pink module were underexpressed in viruliferous A. gossypii that acquired CLRDV from infected hibiscus plants, and 10 genes were identified as hub genes based on their high connectivity values (Figure 6B). Some of these hub genes were predicted to encode proteasome subunit-β (XM_027980885.1), trehalose transporter Tret1 (XM_027996416.1), and M-phase inducer phosphatase (XM_027998546.1) (Supplementary Table S6).

Most genes in the brown module were underexpressed in viruliferous A. gossypii that acquired CLRDV from infected okra plants, and 10 genes were identified as hub genes based on their high connectivity values (Figure 6C). These hub genes were predicted to encode ubiquitin carboxyl-terminal hydrolase 7 (XM_027983467.1), dynein heavy chain (XM_027987391.1), serine/threonine-protein kinase WNK1 (XM_027980933.1), and eukaryotic translation initiation factor 4 gamma (XM_027992697.1) (Supplementary Table S6). The serine/threonine-protein kinase genes were associated with signal transduction following virus acquisition (Table 4).

In this study, the acquisition of CLRDV resulted in transcriptional changes in A. gossypii, of which only four DEGs were common between the viruliferous A. gossypii adults acquiring the virus from four host species. Among four common DEGs, the direction and/or level of the expression (over or under) of common genes varied between host species. The KEGG annotation and GO enrichment analysis revealed the role of one of the common genes (XM_027981156.1, heat shock protein 70) in virus infection, signal transduction, immune responses, longevity, and endocytosis. Heat shock protein 70 was overexpressed in A. gossypii upon acquiring the virus from CLRDV-infected cotton and okra plants, whereas it was underexpressed in A. gossypii that acquired the virus from CLRDV-infected hibiscus and prickly sida plants. In this study, the expression level of heat shock protein was nearly double in A. gossypii that acquired the virus from infected okra compared with cotton. The heat shock proteins are essential chaperone proteins known to be overexpressed in response to stress conditions. One study found that heat shock protein 70 was overexpressed upon BYDV acquisition in its aphid vector (R. padi). The BYDV infection has been reported to increase the plant surface temperature and aphid heat tolerance, suggesting a protective role (Porras et al., 2020). Another study has reported the interaction of tomato yellow leaf curl virus (TYLCV) with Bemisia tabaci (Gennadius) heat shock protein 70 in the midgut using in vitro studies. The protein was suggested to play an inhibitory role in virus transmission (Götz et al., 2012). The higher expression level of heat shock protein could be one of the reasons for reduced CLRDV acquisition and inoculation from okra to cotton plants by A. gossypii reported in an earlier study (Pandey et al., 2022).

However, the other common gene (XM_027987236.1, juvenile hormone acid O-methyltransferase) was overexpressed in aphids that acquired the virus from cotton and okra but underexpressed in aphids that acquired the virus from hibiscus and prickly sida plants. In a previous study, the JHAMT (juvenile hormone-III synthase) was overexpressed in S. avenae that acquired BYDV from wheat plants (Li et al., 2019). This gene is known to play a regulatory role, as a rate-limiting enzyme in insect juvenile hormone biosynthesis, which is essential in the development, metamorphosis, and reproduction of insects (Shinoda and Itoyama, 2003; Minakuchi et al., 2008; Niwa et al., 2008). In contrast, in another study, the underexpression of juvenile hormones in aphids was linked with increased wing development and enhanced virus spread (Quan et al., 2019; Zhang et al., 2019). Variations in the expression levels of juvenile hormone acid O-methyltransferase observed in A. gossypii suggest that CLRDV acquisition may enhance or reduce the fitness of A. gossypii depending on the host species and warrants further examination.

In addition to the four DEGs in common, many unique DEGs were identified in A. gossypii depending on the plant species from which the virus was acquired. The number of DEGs uniquely expressed was the highest in aphids that acquired the virus from okra plants. For example, the ras-related protein (Rab protein) associated with signaling in the circadian clock cells in Drosophila melanogaster Meigen was uniquely identified and underexpressed in aphids that acquired the virus from okra plants (Williams et al., 2001). Rab proteins also function as transporters of vesicle cargo, responsible for trafficking among several membrane compartments (Zhang et al., 2007). Hence, the underexpression of this gene could be one of the reasons for the lower virus acquisition and/or retention ability of aphids from okra plants. The tubulin beta-1 chain gene encoding a structural constituent of the cytoskeleton was uniquely overexpressed in A. gossypii that acquired CLRDV from infected cotton plants in this study. The overexpression of this gene enhances the insects’ development and reproduction (Nielsen et al., 2010). Tubulin is also a major constituent of microtubules, which is an integral part of intracellular transport (Logan and Menko, 2019). This could be one of the reasons for better fitness and acquisition of CLRDV in adult A. gossypii that acquired the virus from cotton plants compared with the other three hosts in the previous study (Pandey et al., 2022). In contrast, this gene was underexpressed in M. persicae adults that acquired TuYV from virus-infected plants and artificial medium (Marmonier et al., 2022).

The gene coding for ubiquitin-conjugating enzyme was overexpressed sixfold in A. gossypii that acquired CLRDV from virus-infected cotton plants, whereas it was overexpressed ~1.5-fold when the virus was acquired from okra and prickly sida plants in this study. The gene was not differentially expressed in A. gossypii acquiring CLRDV from virus-infected hibiscus plants (Table S3). The overexpression of ubiquitin-conjugating enzymes was previously reported in M. persicae and B. tabaci feeding on BYDV-infected and sida golden mosaic virus (SiGMV)-infected plants, respectively (Li et al., 2020; Mugerwa et al., 2022). The conjugating enzyme can transfer the ubiquitin from E1 to the substrate and is required for Notch signaling activation during Drosophila wing development (Gonen et al., 1999; Zhang et al., 2021). Since this gene is reported in the endocytic trafficking of the Notch protein, it could potentially influence the endocytic traversion of virus particles in A. gossypii. The overexpression of this gene may partially be responsible for the efficient retention and inoculation of CLRDV upon acquisition of the virus from cotton than the other three hosts (Pandey et al., 2022). This gene also is vital for insect defense against pathogens (Lemaitre and Hoffmann, 2007).

The immune system of insects helps them defend against pathogens (Wang et al., 2016). The change in the expression level of the genes related to the immune system and different signaling pathways in A. gossypii varied between the host species from which the virus was acquired. The genes related to the MAPK signaling pathway were differentially expressed in A. gossypii that acquired the virus from all hosts. In contrast, genes related to the JAK–STAT signaling pathway were only differentially expressed in A. gossypii that acquired the virus from okra plants. The JAK–STAT signaling pathway triggers insects’ innate immunity and antiviral responses (Dostert et al., 2005; Hedges and Johnson, 2008; Kingsolver et al., 2013). One of the genes associated with the immune system in aphids is Cathepsin B (Kubo et al., 2012; Quan et al., 2019). Cathepsin B is an aphid gut cysteine protease that regulates host protein activity and plays a role in the recognition and movement of viruses at the gut level (Pinheiro et al., 2017; Heck and Brault, 2018). Cathepsin B gene transcripts were overexpressed in A. gossypii acquiring CLRDV from infected cotton and okra plants alone. It was reported previously that the cathepsin B expression in aphids depends significantly on the host species (Pinheiro et al., 2017). This may partially explain the identification of the cathepsin B gene only in two host species in this study. The overexpression of the cathepsin B gene also was reported from M. persicae that acquired TuYV from infected plants (Marmonier et al., 2022). In contrast, M. persicae that acquired PLRV from infected plants had reduced expression of cathepsin B, which was associated with enhanced PLRV transmission (Pinheiro et al., 2017). The overexpression of the cathepsin B gene in A. gossypii acquiring the virus from okra plants may be one of the reasons for reduced CLRDV retention and subsequent inoculation in the previous study (Pandey et al., 2022). However, it does not explain the CLRDV retention and inoculation results obtained in A. gossypii upon acquisition from cotton despite overexpression of cathepsin B genes (Pandey et al., 2022).

For successful aphid-mediated transmission of circulative non-propagative phytoviruses such as poleroviruses, the virus capsid protein must interact with putative receptors at the midgut and accessory salivary glands (Gray and Gildow, 2003). One of the critical gene families identified in this study was serine/threonine kinase receptors. These genes were differentially expressed in both directions (over and under) in A. gossypii upon virus acquisition from four different host species in this study. This gene also was identified as one of the hub genes in the largest module (turquoise) in WGCNA in this study. The differential expression of these genes also was reported in whiteflies that acquired another group of persistent non-propagative circulative phytoviruses (begomoviruses) compared with their non-viruliferous counterparts (Mugerwa et al., 2022). Serine/threonine kinase, in mammalian cells, also has been recorded to play a vital role in clathrin-mediated endocytosis of the rabies virus (Wang et al., 2020a). The identification of these receptors in this study highlights their potential role in the circulative movement of poleroviruses in their aphid vectors. However, the role of host plants in the differential expression of these receptors’ genes cannot be explicitly established in this study.

Another important group of DEGs is associated with xenobiotics detoxification. Genes such as cytochrome P450, ATP binding cassette transporters (ABCs), and UDP-glycosyltransferases (UGTs) were differentially expressed in A. gossypii that acquired CLRDV from different host species in this study. These detoxification genes are essential for the adaption of insects to different host plants (Quan et al., 2019). Among them, cytochrome P450 genes were mainly overexpressed in A. gossypii that acquired CLRDV from infected okra and prickly sida plants. However, they were not differentially expressed in A. gossypii that acquired CLRDV from other host species. The overexpression of cytochrome P450 genes also was reported in M. persicae upon PLRV acquisition (Patton et al., 2021). Therefore, the overexpression of these genes may assist aphids enhancing the tolerance of non-desirable host plants, which could ultimately help in virus transmission and epidemics (Casteel and Jander, 2013).

Gene co-expression networks attained through WGCNA also identified modules of highly correlated genes associated with virus transmission and vector performance. Three of the four most interacting hub genes were annotated: XM_027993669.1 (trichohyalin-like), XM_027997289.1 (uncharacterized protein), XM_027997209.1 (glucose dehydrogenase [FAD, quinone]-like), and XM_027983368.1 (neuroendocrine convertase 1-like). A previous study speculated the role of trichohyalin during immune defense via tissue remodeling and interaction with cuticular binding blocks that facilitate encapsulation (Takase and Hirai, 2012; Simons, 2015; Feng et al., 2022). Similarly, the FAD glucose dehydrogenase is a detoxification enzyme, the overexpression of which induces defense by reducing quinone in parasite-infected bumble bees (Stone and Yang, 2006; Giacomini et al., 2023). The neuroendocrine convertase 1, also called proprotein convertase (PC1/3), is a neuropeptide involved in regulating insect growth and development (Greenlee and Harrison, 2004; Callier and Nijhout, 2011). Another earlier study also has identified the essential role of PC1/3 in maintaining metabolic balance and nutrient-dependent fertility in adult beetles (Fritzsche and Hunnekuhl, 2021). These highly interacting hub genes could play a significant role in the development and defense mechanisms in A. gossypii following CLRDV acquisition and could be important targets for future investigation.