C. vestalis was reared on P. xylostella as the host at 25°C with a relative humidity of 65% under a 14:10 light:dark cycle. P. xylostella larvae were provided with cabbage, and all adult P. xylostella and C. vestalis were fed a 20% honey/water (V/V) solution (56). To obtain parasitized host larvae for the experiments, mid-third instar P. xylostella larvae were exposed to one single C. vestalis female wasp within a 10 mm (diameter) × 80 mm (height) glass vial.

L. boulardi (58) was reared on D. melanogaster (W¹¹¹⁸ strain) as the regular host at 25°C with a relative humidity of 50% under a 16:8 light:dark cycle. The newly emerged L. boulardi was provided apple juice agar medium. All Drosophila strains used in this study were maintained on standard cornmeal/molasses/agar medium at 25°C in 6-ounce square bottom plastic fly bottles.

To obtain CvBV_28-1 transgenic flies, CvBV_28-1 with the hemagglutinin (HA) epitope tag was first cloned into the pUAST-attb vector. The transgenic Drosophila line carrying the UAS-CvBV_28-1 gene was obtained by phiC31 integrase-mediated insertion into the attP2 landing-site locus on the 3^rd chromosome.

Drosophila Schneider 2 (S2) cells were maintained in 60 mm culture dishes in Schneider’s Drosophila Medium (Invitrogen) plus 10% fetal bovine serum (FBS) at 27°C under an ambient atmosphere.

P. xylostella hemocytes were sampled in TRIzol reagent at 1 h, 6 h, 12 h, 24 h, 72 h and 120 h post-parasitization with three biological replicates at each time points. RNA extraction, construction of the cDNA library and paired-end RNA-seq (Illumina) were carried out by Annoroad Gene Technology Co., Ltd. The transcriptome sequencing data statistics are listed in Supplementary Table 1 . The CvBV-related reads were taken from each Illumina library by mapping reads to the CvBV genome (14). The CvBV genome index was built using Bowtie (v2.1.0) (59), and paired-end clean reads were aligned to the CvBV genome using TopHat (v2.1.1) (60). Cuffdiff (v2.2.1) (61) was used to calculate fragments per kilobase of exon per million fragments mapped (FPKMs) for the coding genes in each group, and the FPKM was calculated based on the length of the fragments and the read count mapped to each fragment. The sum of the average FPKMs at these 6 timepoints post-parasitization was used to determine the top 50 CvBV transcriptional levels in hemocytes ( Supplementary Table 2 ). Cluster analysis of the CvBV transcription pattern was performed via the pheatmap R package (v1.0.12) (62), with FPKMs standardized by natural logarithm. Other heatmap plots were generated with GraphPad Prism (v9.0) with the standardized average FPKM from three biological replications.

The Pfam seed database of CRD (PF00059) (63) and the CTLs in Bombyx mori, Manduca sexta and Nasonia vitripennis retrieved from previous studies (64–66) were used as seeds to annotate the CTL gene family in Cotesia vestalis bracovirus. Most CTL genes in full-sequenced PDVs were primarily annotated with BLASTP (http://www.ncbi.nlm.nih.gov/) based on seed the sequences, and additional CTLs were identified as potential genomic loci by TBLASTN (http://www.ncbi.nlm.nih.gov/) and subsequently predicted using FGENESH (67). Data resources are available in the National Center for Biotechnology Information. All obtained CTL sequences were analyzed by SMART to verify the presence of CRD (68). Signal peptides were predicted with SignalP 5.1 (http://www.cbs.dtu.dk/services/SignalP/). Statistics of the annotated CTL protein sequences are listed in Supplementary Table 4 .

Annotated CTL protein sequences in the above species were aligned by MUSCLE in MEGA X with the following parameters: maximum iterations = 100 and clustering method (for iterations 1, 2) = UPGMB (69). Phylogenetic analysis was performed by IQTREE with the automated parameters (70). Phylogenetic trees were viewed using FigTree v1.4.4. Spatial structures of all bracovirus CTLs were predicted with the I-TASSER server (71).

Total RNA was isolated from homogenized parasitized P. xylostella using TRIzol and reverse transcribed into cDNA using the PrimeScript™ 1^st Strand cDNA Synthesis Kit (Takara). The entire coding region of CvBV_28-1 was cloned and sequenced. Primer sequences are listed in Supplementary Table 5 .

Total RNA was extracted and then reverse transcribed into cDNA using the ReverTra Ace qPCR RT kit (Toyobo) according to the manufacturer’s protocol. qRT–PCRs were performed in a CFX Connect real-time system (Bio–Rad) with THUNDERBIRD qPCR Mix (Toyobo). Reactions were carried out for 60 s at 95°C, followed by 40 cycles of 15 s at 95°C and 30 s at 60°C. The Px-β-Actin gene (GenBank accession number: NM_001309101) and Px-β-Tubulin gene (GenBank accession number: EU127912) were used as internal controls, and the relative concentrations were determined using the 2^−ΔΔCt method. All the primers used for qRT–PCR in this study are listed in Supplementary Table 5 .

Total protein from approximately 10 transgenic fly larvae was extracted by Minute™ Total Protein Extraction Kit for Insects (Invent) for western blot according to the manufacturer’s protocol. Samples were diluted in 5× Protein Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis Loading Buffer (Sangon), then boiled for 10 min. Proteins were separated in a denaturing polyacrylamide gel and transferred to a polyvinylidene difluoride membrane.

After blocking and washing, membranes were then incubated with primary antibodies against HA-tag (1:2500, Sangon) or primary antibodies against actin (1:2500, CWBIO) for 2 h at room temperature. Membranes were then incubated with secondary antibody horseradish peroxidase conjugated goat anti-mouse IgG (1:5000, Sangon) in Tris-buffered saline with 0.05% Tween-20 for 2 h at room temperature. After five washes, membranes were then incubated with the enhanced chemiluminescence western blotting substrate for imaging (Promega).

For RNAi, a 25-bp RNA oligo was designed based on the sequence of CvBV_28-1 and synthesized by Sangon Biotech. The sequence is listed in Supplementary Table 5 . The miRCURY LNA miRNA mimic (siNC, EXIQON 479903-001) was used as a negative control. A total of 5 pmol of siRNA was injected into each mid-third instar P. xylostella larva using an Eppendorf FemtoJet 4i Microinjector with the following parameters: injection pressure = 900 hPa and injection time = 0.15 sec (56). Parasitization was conducted 6 h post-injection, and the RNAi efficiency of CvBV_28-1 in P. xylostella hemocytes was detected 6 h post-parasitization by qRT–PCR. At least three biological replicates were performed.

Hemocyte density of P. xylostella was detected with a cell counter (CountStar). Briefly, P. xylostella larvae subjected to different treatments were carefully rinsed three times with 1× PBS and dried with filter paper before dissection. Then, 0.5 µL of hemolymph was diluted in 19.5 µL of Typan Blue Solution (Sangon), and the mixture was dropped on an exclusive slide for the cell counter. At least three nonoverlapping images of each slide were captured, and the average concentration was converted into cells per µL. As the volumes of parasitized and nonparasitized P. xylostella larvae were the same, the index of hemocyte density was considered as an indicator of hemocyte number in P. xylostella.

Hemocyte numbers of transgenic flies were counted with a hematocytometer (Watson). Briefly, transgenic fly larvae at 48 h post-parasitization were carefully rinsed three times with 1× PBS and dried with filter paper before dissection. Then, hemolymph of ten individuals was diluted in 20 µL 1× PBS, and 8 µL of mixture was dropped on the hematocytometer. Circulating hemocytes and lamellocytes were shown with GFP (Hml>GFP) and fluorescent red (MsnCherry) respectively. The circulating hemocyte numbers and lamellocyte numbers of transgenic fly larvae in four boxes of corners were counted under a Zeiss LSM 800 confocal microscope, and the average number of hemocytes per larva was converted into cells per individual.

Drosophila larvae 48 h after L. boulardi attack were used for imaging. They were carefully rinsed three times with 1× PBS and mounted in a drop of glycerol on a glass slide with their dorsal side facing upward. Mounted larvae were kept at -20°C for 20 min before imaging to ensure that the samples were completely fixed. L. boulardi eggs were dissected from the Drosophila larvae 24 h post-parasitization in 1× PBS, fixed in 4% paraformaldehyde in PBS for 30 min, and rinsed three times with 1× PBST (PBS containing 0.1% Triton X-100 and 0.05% Tween-20). Samples were mounted in ProLong Gold Antifade Mountant with DAPI (Invitrogen). Fluorescence images were captured with a Zeiss LSM 800 confocal microscope.

For EdU labeling, Click-iT EdU staining was performed on hemocytes from live larvae and S2 cells according to the manufacturer’s instructions (Invitrogen). Host larvae hemocytes were dissected in 1× PBS and allowed to adhere on the slides for 20 min. Then, they were stained with 10 mM 5-ethynyl-2-deoxyuridine (EdU) for 1 h at room temperature and fixed in 4% paraformaldehyde in PBS for 15 min after washing twice with 3% BSA in PBS and once with 1× PBST for 20 min. Fresh cocktail was prepared, and each sample was stained for 30 min, followed by washing twice with 3% BSA in PBS and once with 1× PBST for 20 min. Samples were mounted in ProLong Gold Antifade Mountant with DAPI (Invitrogen). Fluorescence images were captured with a Zeiss LSM 800 confocal microscope.

The ORF of CvBV_28-1 was subcloned into the pAc-V5/His vector (Invitrogen) with the primers shown in Supplementary Table 5 to generate the pAc-CvBV_28-1 plasmid. The Kozak sequence (GCCATGG, the G at positions –3 and +4 of translation initiation) was added to the forward primer, allowing efficient and high-level expression of the recombinant protein in S2 cells. To perform transient transfection, S2 cells were seeded on a cell culture slide in 35 mm culture dishes (to 80% confluence) and transfected with 2.5 μg of pAc-CvBV_28-1 plasmid using a Lipofectamine 3000 Kit (Invitrogen) with pAc-GFP plasmid as a control according to the manufacturer’s instructions.

The DNA fragment encoding CvBV_28-1 was subcloned into the pET32a vector and transformed into the E. coli strain BL21 (DE3). Expression of CvBV_28-1 was induced by isopropyl β-D-1-thiogalactopyranoside (IPTG) at a final concentration of 1 mM. Harvested bacterial cells were washed with 1× PBS and lysed by sonication. The CvBV_28-1 protein was expressed in its insoluble form and the sediment was resuspended in 1× PBS with 8 M urea and purified using High-Affinity Ni-NTA Resin (Roche) according to the manufacturer’s instructions. The CvBV_28-1 protein was refolded at 4°C with a stepwise decreasing gradient of urea (6 M, 4 M, 3 M, 2 M, 1 M to 0 M urea) in 1× PBS with 5% glycerol, 1% L-arginine and 2% glycine. The protein was analyzed by 12% SDS–PAGE, detected by staining with Coomassie blue, quantified by the Bradford method, and stored at -20°C for further experiments.

To assess the agglutination activity of CvBV_28-1, E. coli and S. aureus were collected at OD=0.6 by centrifugation at 3500 rpm for 5 mins, stained by incubation with acridine orange (Sangon, 30 µg/mL) for 20 min at room temperature and washed three times with Tris buffer (20 mM Tris-HCl, pH=8.0). The stained pellets were resuspended in Tris buffer at a concentration of 1× 10⁹ cells/mL for the following test.

For the agglutination response assay, CvBV_28-1 protein was diluted to 10 µg/mL in Tris buffer and mixed with the bacterial suspension in equal volumes. The mixtures were incubated for 1 h at room temperature in the presence or absence of 10 mM CaCl₂. To test the minimum concentration of bacterial agglutination, CvBV_28-1 was serially diluted in Tris buffer at the concentrations of 10, 0.1, 10^-3, and 10^-5 µg/mL. After mixing with an equal volume of bacterial suspension, the mixture was incubated for 1 h at room temperature in the presence of 10 mM CaCl₂. All samples were observed and photographed under a Zeiss LSM 800 laser confocal microscope.

E. coli and S. aureus were grown overnight at 37°C with shaking at 250 r/min. Cultures were centrifuged at 1000 g. The bacteria were resuspended in sterile PBS to achieve an OD of 0.4, and male flies 7 days post-eclosion were injected with 40 nL of the bacterial resuspension with an Eppendorf FemtoJet 4i Microinjector (Eppendorf) and a microcontroller (Narishige). 0.1 µL bacterial resuspension with an OD of 0.02 in sterile PBS was injected into parasitized P. xylostella larvae with CvBV_28-1 knockdown and control at 1 h post-parasitization. Experiments were performed in triplicate (at least 20 individuals per replicate). Injected flies and P. xylostella larvae were then kept at 25°C and recorded after 6 h. Flies and P. xylostella larvae were transferred to a fresh container of every day, and death was recorded every 12 h.

All statistical analyses were performed with GraphPad Prism (v9.0) and Fiji2. Data are expressed as the means ± SD. Log-rank tests were used to determine whether the male fly survival curves were significantly different from one another. For comparison of the hemocyte numbers at 6 and 12 h post-parasitization between the parasitized and nonparasitized host larvae, we performed two-way ANOVA followed by Šidák’s multiple comparisons test with Spearman’s test for heteroscedasticity and the D’Agostino-Pearson omnibus test for normality of the residuals. For comparison of the hemocyte numbers and circulating lamellocyte numbers at 48 h post-parasitization between transgenic flies ectopically expressing CvBV_28-1 in hemocytes (Hml>CvBV_28-1) and control, we performed two-tailed unpaired Student’s t test. One-way ANOVAs were conducted for other experiments with Tukey’s multiple comparisons test. The intensity of fluorescence signal and the number of positive cells in the images were calculated with Fiji2. Significant values are indicated as *P < 0.05, **P < 0.01 and ***P < 0.005.

To comprehensively understand the role of CvBV genes in resistance to host cellular immunity, cDNAs were generated from the hemocytes of C. vestalis-parasitized P. xylostella larvae at a series of time points, including 1 h, 6 h, 12 h, 24 h, 72 h and 120 h post-parasitization. The cDNAs were sequenced using the Illumina HiSeq 2500 platform ( Figure 1A ). The raw sequencing dataset was submitted to the SRA of the NCBI, with the accession numbers from SAMN25185395 to SAMN25185412. After discarding low-quality reads, we obtained clean reads ranging from 39,345,132 to 47,706,030 in these 18 cDNA libraries. The quality Q30 values after data filtering were all greater than 94.19% ( Supplementary Table 1 ). Then, we calculated the fragment per kilobase of exon per million (FPKM) values of the CvBV genes in hemocytes using the CvBV genome as a reference. We ranked the top 50 CvBV genes highly expressed in parasitized hemocytes of hosts, and performed the cluster analysis. The transcriptional patterns of CvBV genes in hemocytes were divided into three distinct types: early (higher expression at early time points), late (higher expression at late time points), and whole-period (high expression at all time points) ( Figure 1B ). The details of the gene names and their transcriptional levels are shown in Supplementary Table 2 . We next focused on the CvBV genes showing high transcriptional levels at early time points in the parasitized hemocytes, because they were most likely to be associated with the host immune suppression process. Among them, we observed one CTL gene, CvBV_28-1, ranked as the top highly expressed CvBV gene in the early category ( Figure 1C ). The ORF of CvBV_28-1 is 474 bp, encoding a 157 amino acid (aa) protein with a calculated molecular weight of 17.54 kDa and pI of 7.63 ( Figure 2A and Supplementary Table 3 , GenBank accession number: QZB49176.1). CvBV_28-1 has a predicted N-terminal signal peptide and one single CRD but lacks a typical transmembrane domain, suggesting that it was likely to be a secreted protein into host hemolymph. From the annotated results of the CvBV genome, two CTL genes were identified, the other being CvBV_16-8, which was placed in the whole-period category. The ORF of CvBV_16-8 is 471 bp, encoding a 156 amino acid (aa) protein with a calculated molecular weight of 17.55 kDa and pI of 6.45 ( Figure 2A and Supplementary Table 3 , GenBank accession number: QZB49081.1). CvBV_16-8 shares the same sequence feature with CvBV_28-1. To further profile a detailed dynamic pattern of the two CTLs, we sampled 8 different tissues (hemocytes, central neural system, midgut, fat body, cuticular, Malpighian tubule, silk gland and testis) of P. xylostella at different time points post-parasitization and determined the transcription levels of the two CTLs by quantitative real-time PCR (qRT–PCR). The results showed that both were highly expressed in hemocytes, with transcriptional level of CvBV_28-1 being higher at a much earlier parasitism stage and that of CvBV_16-8 being higher at 24 h post-parasitization ( Figures 1D, E ).

To analyze the sequence features of the CTLs in PDVs, we noted that these two CTLs had a single CRD and a signal peptide at the N-terminus ( Figure 2A ), which is similar to the reported CTLs from other PDVs (72, 73). We further found that CTL genes existed only in the bracovirus species of the genera Cotesia and Glyptapanteles but not in other bracovirus species or ichnovirus species. All of the CTL proteins shared the same domain architecture ( Figure 2A ), including one signal peptide and one single CRD. Interestingly, we found that the number of CTL genes was different in the bracovirus species with available genomes. Briefly, one CTL gene was identified in Cotesia sesamiae BV (CsBV), two CTL genes were present in Cotesia congregata BV (CcBV), Cotesia ruficrus BV (CrBV), Glyptapanteles indiensis BV (GiBV) and CvBV, and three CTLs were found in Glyptapanteles flavicoxis BV (GfBV) with a few different amino acids ( Figure 2A ). Multiple sequence alignment of the BV-derived CTLs separated them into two clusters with high bootstraps, and two pairs of disulfide bonds were found to stabilize the protein structures with four conserved cysteines. The Glu-Pro-Ser motif was identified as the carbohydrate ligand binding motif in the CRDs in the majority of the bracovirus CTLs, while Glu-Pro-Asp existed in only CvBV_16-8 and CcBV32_307 and Lys-Pro-Ser existed in only CrBV-CTL2 ( Figure 2A ). We next constructed the predicted spatial structures of all of these CTLs and found that except for GiBV-CTL2, which lacks the Ca²⁺-binding site, all had one Ca²⁺-binding site at site 4, which is important for salt bridge formation between α2 and the β1/β5 sheet ( Figure 2B ) (42, 74). In addition, different types of binding carbohydrates were also predicted in all of these CTLs. As a result, N-acetyl-α-D-mannosamine was predicted in CvBV_28-1 and GiBV-CTL1; α-L-fucopyranose was predicted in CvBV_16-8, two CTLs of CcBV, CrBV-CTL1 and GfBV-CTL2; β-D-mannose was predicted in one CTL of CsBV; trehalose was predicted in CrBV-CTL2; and N-acetyl-D-galactosamine was predicted in GfBV-CTL1, GfBV-CTL3 and GiBV-CTL2 ( Figure 2A ). The details of these CTLs spatial structures are shown in Supplementary Table 3 . It has been reported that CTLs are widespread in lepidopteran insects and some CTLs from bracoviruses horizontally transferred into genomes of nonparasitized hosts (50, 52), we performed the PCR experiment with specific primers of CvBV_28-1 using the DNA as template from parasitized and nonparasitized P. xylostella larvae. The results showed that CvBV_28-1 derived from C. vestalis bracovirus could be detected in the parasitized hosts, but this gene did not exist in nonparasitized host larvae ( Figure 2C ). Taken together, these results indicate that CvBV_28-1 may bind to mannose-group carbohydrates on the surface of microorganisms with the help of calcium.

Host defense against parasitoids relies on hemocytes and certain humoral components, which can recognize and respond to invading parasitoids (75–77). However, parasitoid PDVs, contribute to disrupting the resistant responses of the hosts by killing hemocytes or altering their ability to adhere to the surface of invasive foreigners (78). We measured the changes of hemocyte numbers in P. xylostella at 6 h and 12 h post-C. vestalis parasitization. Compared with the nonparasitized larvae, the parasitized hosts showed a significant reduction in the concentration of hemocytes at the early parasitization stage, indicating a decreased number of host hemocytes ( Figure 3A ). To test whether CvBV_28-1 participated in the suppression of host hemocyte numbers because of its early expression profile post parasitization ( Figure 1C ), we next performed RNA interference (RNAi) experiments to knockdown the expression of CvBV_28-1 in parasitized P. xylostella larvae. SiRNAs of CvBV_28-1 were injected into third instar host larvae, and qRT–PCR showed that the expression of the CvBV_28-1 gene decreased significantly by approximately 70% ( Figure S1 ). We also found that the reduction in hemocytes was effectively suppressed in CvBV_28-1 knockdown hosts ( Figure 3B ).

Consistent with the fact that CvBV_28-1 can reduce the number of hemocytes, it has been reported that a viral lectin gene has the ability to prevent the encapsulation response against wasp eggs (79). Because P. xylostella lacks the specific markers for hemocytes and has some difficulties in conducting some certain experiments in vivo, we tested the function of CvBV_28-1 in vivo using a Drosophila-parasitoid system. The parasitoid wasp Leptopilina boulardi–host Drosophila system is an excellent model and have been widely used to dissect the underlying mechanisms of the parasitoid-induced host immune responses. The encapsulation responses were obvious in L. boulardi-parasitized host Drosophila, but were absent in the P. xylostella and C. vestalis system. We then fabricated a GAL4/UAS binary expression system in D. melanogaster (80). A transgenic line carrying a UAS transgene encoding the CvBV_28-1 protein with hemagglutinin (HA) epitopes tagged at the C terminus was constructed. CvBV_28-1 expression was driven by the specific hemocyte GAL4 (Hml-GAL4), and verified by qRT–PCR and western blotting ( Figures 4A , S2A, B ). We then tested the anti-encapsulation response of CvBV_28-1 in Drosophila larvae ectopically expressing viral CvBV_28-1 from C. vestalis in hemocytes. We determined the degrees of the host encapsulation responses and divided them into two categories: large capsules (melanotic encapsulation that covered more than 50% of the wasp egg) and small capsules (melanotic encapsulation that covered less than 50% of the wasp egg) ( Figure 4B ). We found that the proportion of large capsules dramatically decreased, indicating a reduction in host encapsulation responses when CvBV_28-1 in host hemocytes was ectopically expressed ( Figure 4C ).

We next investigated which cell populations were reduced due to CvBV_28-1 ectopic expression in L. boulardi parasitized hosts. Whole Drosophila larvae with different genotypes were imaged, and GFP was used as an indicator of all hemocytes. Overexpressing CvBV_28-1 reduced hemocyte numbers in parasitized host larvae ( Figures 4D, E ). In addition, Drosophila larvae normally generate a special type of hemocyte, called lamellocytes, for encapsulation post wasp parasitization (58, 77, 81, 82). As such, whole larvae with different genotypes were further imaged with MSNF9MO-mCherry (msnCherry), a well-known marker for lamellocytes (83). Similar to previous reports, we confirmed that large quantities of lamellocytes were produced 48 h post-L. boulardi infection. As expected, overexpressing CvBV_28-1 led to a significant reduction in circulating lamellocyte numbers in parasitized host larvae ( Figures 4D, F ). We further dissected wasp eggs from the host larvae 24 h after L. boulardi attack to determine the encapsulation degree. In comparison with the many lamellocytes adhered to the surface of the wasp eggs, overexpression of CvBV_28-1 resulted in few lamellocytes on wasp eggs ( Figure 4G ). Collectively, these results indicate that CvBV_28-1 leads to a significant reduction in circulating hemocytes in parasitized host larvae and prevents lamellocytes from adhering to wasp eggs to initiate the encapsulation response.

To further ascertain the mechanisms of CvBV_28-1 post wasp parasitization, we proposed that excessive programmed cell death or an impaired proliferation rate may result in a reduction in circulating hemocytes (84, 85). We first examined the level of apoptosis in circulating hemocytes of CvBV_28-1-expressing Drosophila larvae post wasp infection. Notably, we did not observe a significant difference in the apoptosis level compared with the control groups ( Figure S3 ). Additional expression of DIAP (Death-associated inhibitor of apoptosis) in hemocytes (Hml>DIAP, CvBV_28-1) also had no effect on inhibiting the reduction in hemocytes ( Figure S4 ). We next tested whether the proliferation of hemocytes was impaired via 5-ethynyl-20-deoxyuridine (EdU) assays. When CvBV_28-1 was overexpressed in hemocytes (Hml>CvBV_28-1), the number of EdU-labeled circulating hemocytes was significantly lower than that in control larvae, indicating that CvBV_28-1 inhibits the proliferation of host hemocytes post parasitization ( Figures 5A, B ). Moreover, the expression of CycE (cyclin E) in hemocytes (Hml>CycE, CvBV_28-1) rescued the proliferation rate ( Figures 5A, B ). Consistent with the rescued proliferation level, the number of lamellocytes in circulation and attached to wasp eggs also increased compared to the control larvae ( Figures 5C-E ).

Finally, we performed an EdU incorporation assay for circulating hemocytes in P. xylostella, the true host of CvBV genes. We found that the ratio of EdU-positive hemocytes clearly decreased, indicating a reduction in the proliferation of circulating hemocytes post parasitization ( Figures 6A, B ). In comparison to siNC-injected parasitized host larvae, silencing CvBV_28-1 in parasitized P. xylostella partially rescued the reduction in proliferation in circulating hemocytes ( Figures 6A, B ). An EdU incorporation assay was also conducted in S2 cells with pAc-CvBV_28-1 and control (pAc-GFP) 48 h post-transfection, and overexpression of CvBV_28-1 in S2 cells downregulated the EdU-positive rate ( Figure 6C ). These data suggested that CvBV_28-1 represses hemocyte proliferation to inhibit host cellular immunity after wasp parasitization.

CTLs with CRDs can recognize pathogens by interacting with their cellular surface and promote bacterial agglutination to mediate immune defense responses (42–44). Moreover, our predicted spatial structure of CvBV_28-1 indicated that it may perform bacterial agglutination with the help of calcium. To test whether CvBV_28-1 had this ability, we performed an experiment to evaluate the function of CvBV_28-1 agglutination activity in response to a gram-positive bacterium (Staphylococcus aureus) and a gram-negative bacterium (Escherichia coli). We found that CvBV_28-1 showed a strong agglutination response to S. aureus in the presence of Ca²⁺ and no agglutination activity to E. coli with or without Ca²⁺ ( Figure 7A ). Moreover, the concentration of CvBV_28-1 resulting in an agglutination response to this gram-positive bacterium was low. In our study, the minimum concentration in a Ca²⁺-dependent manner was 10^-3 µg/mL ( Figure 7B ).

To confirm the antibacterial function of CvBV_28-1 in vivo, we performed a survival assay to assess the pathogen susceptibility of previous transgenic flies. After injection with PBS, the flies seldom died in the 72 h assay. In contrast, the survival rate of the flies decreased significantly after challenge with both S. aureus and E. coli. Interestingly, flies overexpressing CvBV_28-1 in the hemocytes exhibited decreased susceptibility to S. aureus leading to a rescued survival rate. Consistent with the in vitro results, overexpressing of CvBV_28-1 did not enhance fly survival after infection with E. coli ( Figures 8A, B ). We also conducted the survival assay in parasitized P. xylostella larvae challenged by S. aureus and E. coli, and CvBV_28-1 knockdown only led to the decreasing survival rate of hosts after S. aureus infection, compared with siNC-injected parasitized host larvae ( Figures 8C, D ).

In conclusion, CvBV_28-1 efficiently performed S. aureus agglutination in a Ca²⁺-dependent manner, which led to an antibacterial response. However, no effects were observed on E. coli infection.