Symbiotic bracovirus of a parasite modulate host ecdysis process

Abstract

Parasitoids modulate host development for the survival of their offspring, but the mechanisms underlying this phenomenon remain largely unknown. Here, we found that the endoparasitoid Cotesia vestalis disrupted the larval-larval ecdysis in its host Plutella xylostella by the 20-hydroxyecdysone (20E) synthesis pathway. After parasitization by C. vestalis, the 20E peak of host larvae disappeared before the onset of ecdysis and the expression of ecdysone synthesis genes was significantly downregulated. We further found that a Cotesia vestalis bracovirus (CvBV) gene CvBV_28 − 5 was transiently high-level expressed prior to the host’s 20E peak, enabling the precise suppression of this critical developmental signal. Consistently, the knockdown of CvBV_28 − 5 affected the expression of 20E response transcription factors in the cuticle and several ecdysis-related genes. Furthermore, we found that CvBV_28 − 5 bound directly to the Raf, a MAP3K member of the MAPK pathwaythat functions as a critical regulator of ecdysone synthesis genes in hosts. Collectively, our results provide the first evidence that parasitoids modulate host ecdysis by affecting MAPK-20E signaling during a defined developmental window and provide novel insights into the mechanism of parasitoid regulation of host development.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00018-025-05717-6.

Introduction

Parasitic wasps, also known as hymenopteran parasitoids, lay their eggs on or inside the body of their hosts [1, 2]. The successful development of parasitoids eventually leads to the death of hosts, making them effective biological control agents for insect pest management [3]. Parasitoids have evolved various virulence factors to facilitate their parasitization, including venoms [4], polydnavirus (PDVs) [5], teratocytes [6], and larval secretion [7]. Polydnavirus (PDV), recently renamed polydnaviriform, is a mutualistic viral symbiosis with Ichneumonoidea. Bracoviruses are associated with the Braconidae family (Hymenoptera: Ichneumonoidea), which comprises about 40,000 species, while Ichnoviruses are related to the Ichneumonidae family (Hymenoptera: Ichneumonoidea), with around 14,000 species [5, 8]. As a symbiotic virus, PDVs are exclusively produced in the ovary calyx of certain parasitoid wasps. Their survival depends entirely on the wasp’s reproductive success. Thus, PDVs favor parasitoid offspring survival by modulating host development [2, 9], immune response [10, 11] and metabolic homeostasis [12]. The alteration of host development by PDV has been widely reported [13]. PDVs modulate host development by prolonging host larval instar duration or impairing pupation rate, while the involvement of PDV in host ecdysis, another key developmental event of insects, has not been examined [2, 14–16].

Ecdysis is the process by which insects shed their old exoskeleton after making a new one, as the old exoskeleton can not grow with the body [17, 18]. Failure to complete ecdysis will be deleterious to the individual, making this process an excellent target for the development of new insect pest management strategies. Successful ecdysis depends on a signaling cascade giving rise to an ecdysteroid peak [18]. Being conserved among insects, ecdysteroid synthesis depends on the timely expression of Halloween genes. The gene neverland (nvd) converts cholesterol to 7-dehydrocholesterol, which is then processed by the “Black Box” enzymes: spook (spo), spookier (spok), shroud (sro). Further downstream, cytochrome P450 monooxygenases, phantom (phm), disembodied (dib), shadow (sad), shade (shd), produce the final active 20-hydroxyecdysone (20E) [19]. Recent studies have found multiple extracellular signals, including RTKs, GPCRs, Transforming growth factorβ, Hedgehog, etc., that regulate ecdysteroid synthesis in the prothoracic gland (PG) [19]. As the major endocrine organ for ecdysone production, PG integrates diverse physiological and environmental signals to regulate ecdysone synthesis via the Ras/Raf/Erk mediated mitogen-activated protein kinase (MAPK) pathway, converting extracellular signals into 20E peaks that trigger molting and metamorphosis [19]. A surge in ecdysteroid levels initiates the ecdysis prior to the initiation of the ecdysis sequence. Then the fluctuation of ecdysteroid levels induces the expression of sequential bouts of ecdysone response genes, including early response genes E74, E75, E78, BR-C, early-late response genes HR3 and HR4 and a late response gene FTZ-F1β, which translate features of the ecdysteroid peak into the different phases of the ecdysis [20]. However, whether and how parasitoids regulate ecdysis in their hosts remains unclear.

Cotesia vestalis (Hymenoptera: Braconidae) is an endoparasitoid of Plutella xylostella (Lepidoptera: Plutellidae), a destructive pest of cruciferous crops that causes severe economic losses worldwide [21–23]. During oviposition, C. vestalis introduces a symbiotic PDV named Cotesia vestalis bracovirus (CvBV), which is involved in host development arrest by expressing virulence genes after infecting host cells [2]. Our previous study has characterized the molting process of P. xylostella at the different developmental stages in detail, with the larval-larval ecdysis consisting of five stages [24]. At stage I, the larva ceases feeding and wandering for a place to ecdysis. Stage II is marked by the head capsule slip (HCS). In the following stage III, the larvae remain relatively quiescent. At stage IV, the mandibles of the pharate larva are pigmented, and pre-ecdysis behavior appears. The shedding of its old cuticle represents the final stage of ecdysis. Here, we report that C. vestalis parasitization prolongs the duration of larval-larval ecdysis of P. xylostella. We further show that a CvBV gene, CvBV_28 − 5, modulates host ecdysis duration and the expression of ecdysis-related genes by affecting the 20E peak through MAPK signaling.

Results

Cotesia vestalis parasitization affects the ecdysis of host larvae

In our previous research, the larval-larval ecdysis of P. xylostella has been divided into five stages according to the HCS and pigmentation of the new mandible [24]. Based on these two characteristics, we found that C. vestalis parasitization prolonged the duration of the 3rd instar to 4th instar (L3/L4) ecdysis of P. xylostella from 736.9 ± 5.2 to 890.5 ± 6.5 min, with a prolongation of around 20.9%. The greatest change was observed in stages I and IV, in which duration increased by 57.1% and 50.0%, respectively, while stages II and III were less affected, with a prolongation of approximately 10.9% and 6.4%, respectively. However, there were no significant changes at stage V (Fig. 1A).

Fig. 1.

Effects of parasitization of C. vestalis on larval–larval ecdysis of P. xylostella. (A) Third-fourth instar ecdysis duration of nonparasitized and parasitized P. xylostella larvae. NP: nonparasitized; P: parasitized. n = 43 in nonparasitized larvae, n = 41 in parasitized larvae. (B) The sampling time points for the 20E titer in the hemolymph and the expression of ecdysone synthesis genes in the PG of P. xylostella larvae. (C) 20E titer in the hemolymph of P. xylostella larvae. n = 3 in each treatment and each biological group consists of 15–20 P. xylostella larvae. (D) Relative mRNA levels of nvd, sro, phm and sad. n = 4 in each group. Data were analyzed by student’s t test. Values represent the means ± s.e.m. *: p < 0.05; **: p < 0.01; ***: p < 0.001; ns: no significance

To monitor the 20E peak before ecdysis, we collected hemolymph for 20E titer at five time points according to the larval developmental schedule of Bombyx mori [25] and P. xylostella [24] (Table S1). The sampling time points for non-parasitized P. xylostella larvae were 8 h before ecdysis onset (-8 h), 4 h before ecdysis (-4 h), 0 h after ecdysis onset (0 h), 6 h after ecdysis start (+ 6 h), and 12 h after ecdysis onset (+ 12 h). Their counterparts in parasitized P. xylostella larvae were 12 h before ecdysis onset (-12 h), 6 h before ecdysis (-6 h), 0 h after ecdysis onset (0 h), 7 h after ecdysis onset (+ 7 h), 14 h after ecdysis onset (+ 14 h) (Fig. 1B). After parasitization by C. vestalis, the peak of 20E titer before ecdysis disappeared and 20E titer remained at a low level (Fig. 1C).

In the PG, the ecdysteroid biosynthetic genes mediate the conversion of cholesterol to ecdysone [19]. Except for sad, the expressions of the ecdysone synthesis genes nvd, sro and phm in the PG of P. xylostella were significantly downregulated at -6 h of parasitized P. xylostella larvae compared to that at -4 h of non-parasitized P. xylostella larvae, though there were no significant differences between − 12 h of parasitized ones and − 8 h of non-parasitized ones (Fig. 1D). The results above demonstrated that C. vestalis parasitization prolongs the L3/L4 ecdysis duration by impairing ecdysone synthesis of host larvae.

CvBV is the major parasitic factor affecting the ecdysis of host larvae

C. vestalis introduces two parasitic factors, including CvBV [2] and venom [26], directly into P. xylostella larvae to modulate their development. Thus, CvBV particles and venom were injected separately into P. xylostella larvae. Both treatments prolonged the total duration of ecdysis of P. xylostella larvae from 736.9 ± 5.2 min to 925.4 ± 6.1 min and 790.7 ± 8.1 min by approximately 25.6% and 7.3%, respectively. For each stage, CvBV was able to prolong stages I, II, III, and IV. Venom could only affect stages I and III and had no significant effect on stages II, IV, and V. Briefly, compared to the non-parasitized group, CvBV prolonged stage I, II, III and IV and parasitized ones, while venom only prolonged stage I and III (Fig. 2). In conclusion, CvBV is the major parasitic factor causing the prolongation of the ecdysis process in P. xylostella larvae.

Fig. 2.

Effects of venom and CvBV on the L3/L4 ecdysis of P. xylostella larvae. NP: non-parasitized; P: parasitized. The sample numbers (n) are shown in the brackets. Data were analyzed by student’s t test. Values represent the means ± s.e.m. *: p < 0.05; **: p < 0.01; ***: p < 0.001; ns: no significance

CvBV_28 − 5 is necessary for the prolongation of ecdysis in parasitized host larvae

Ecdysteroidogenesis in the PG is regulated by multiple extracellular and intracellular signals [19]. To determine the effect of CvBV on host ecdysteroidogenesis, we analyzed the transcriptome data of the prothorax of parasitized P. xylostella larvae at 12 h and 6 h of the third instar before ecdysis. The top ten expressed CvBV genes include one EP-like gene (CvBV_09 − 5), eight BV-like genes (CvBV_9 − 2, CvBV_11 − 2, CvBV_18 − 4, CvBV_19 − 3, CvBV_19 − 5, CvBV_19 − 7, CvBV_26 − 8 and CvBV_28 − 5) (Fig. S1A, Table S2). The ten selected genes were then silenced by RNAi (Fig. S1B) and the L3/L4 ecdysis duration was observed. It was found that the L3/L4 ecdysis duration of parasitized P. xylostella larvae was reduced by approximately 7.8% from 894.1 ± 10.1 min to 816.1 ± 5.8 min after a 50% knockdown of CvBV_28 − 5. The other CvBV genes have no such effect (Fig. S1C).

As a BV-like gene, phylogenetic analysis showed that CvBV_28 − 5 was clustered with two genes from Cotesia sesamiae bracovirus, closely related to three homologs from Glyptapanteles and Euphydryas (Fig. S2A). The qPCR results showed that the mRNA level of CvBV_28 − 5 in parasitized P. xylostella remained high before ecdysis, and then sharply declined after the L3/L4 ecdysis (Fig. S2B). We also found that CvBV_28 − 5 mainly affected stage I of ecdysis, reducing its duration from 180.8 ± 5.2 min to 148.8 ± 3.8 min by approximately 12.4% (Fig. S2C). With respect to the 20E peak, we found the 50% knockdown of CvBV_28 − 5 increased the level of 20E in parasitized hosts to nearly half that in nonparasitized hosts before ecdysis at -6 h (Fig. 3A). We further determined the expression level of the ecdysone synthesis genes. Results showed that the relative mRNA levels of nvd and phm were restored at -6 h, whereas sro and sad were slightly upregulated (Fig. 3B). Therefore, the transient high expression of CvBV_28 − 5 was possibly a main factor affecting the 20E pulse before L3/L4 ecdysis.

Fig. 3.

Effect of CvBV_28 − 5 on 20E titer and the relative mRNA levels of ecdysone synthesis genes of parasitized third instar P. xylostella larvae. (A) 20E titer in the hemolymph of nonparasitized plus dsGFP treated (NP + dsGFP), parasitized host larvae treated dsGFP (P + dsGFP) and parasitized host larvae treated ds28-5 (P + ds28-5). n = 3 in each treatment and each biological group consists of 15–20 P. xylostella larvae. (B) Relative mRNA levels of nvd, sro, phm and sad. n = 4. Data were analyzed by Student’s t test. Values represent the means ± s.e.m. *: p < 0.05; **: p < 0.01; ***: p < 0.001; ns: no significance

CvBV_28 − 5 reprograms the transcriptional network of ecdysis-related genes

Ecdysis-related genes, also named “output genes” in ecdysis, include those for cuticle proteins, cuticle lipids, cuticle pigmentation and proteases [20]. The next question was to determine which ecdysis-related genes were affected by CvBV_28 − 5. Transcriptome sequencing was performed on the larval cuticles of non-parasitized (NP), parasitized (P), parasitized larvae injected with dsGFP (P + dsGFP) and parasitized larvae injected with ds28-5 (P + ds28-5) at stage I. According to the principal component analysis (PCA) results, there were few variations across the three sample repeats (Fig. S3A). Interestingly, the P + ds28-5 group showed a significant trend toward the NP group when compared to the P + dsGFP group, which suggests that dsCvBV_28-5-treated samples share some similarities with the NP sample. After parasitization, 654 differentially expressed genes (DEGs) were found, of which 417 were up-regulated and 237 down-regulated (Fig. S3B, Table S3). When CvBV_28 − 5 was silenced in the parasitized larval cuticle, 125 genes were differentially expressed, of which 67 were up-regulated and 58 down-regulated (Fig. S3C, Table S3). We further annotated and mapped the DEGs to Gene Ontology (GO) and the Kyoto Encyclopedia of Genes and Genomes (KEGG). After parasitization, a majority of the enriched genes were associated with cuticle lipid synthesis, cuticle pigmentation, protease, and cuticle protein (Fig. S4A, Table S4). KEGG analysis showed that the enrichment pathways are involved in cuticle lipid synthesis, cuticle pigmentation, and protease (Fig. S4B, Table S4). When CvBV_28 − 5 was knocked down in parasitized larvae, GO terms associated with cuticle lipid synthesis and cuticle protein were enriched (Fig. S4C, Table S4). 28 DEGs were mapped to 37 KEGG pathways, including the pathway “phenylalanine metabolism”, “tyrosine metabolism”, which associated with cuticle pigmentation; “fatty acid biosynthesis”, “fatty acid metabolism”, which associated with cuticle lipid synthesis; “neuroactive ligand-receptor interaction”, which associated with protease, although these pathways are not significantly enriched (Fig. S4D, Table S4).

We selected a subset of the above ecdysis-related DEGs and used qPCR for validation experiments. Among them, three DEGs involved in cuticle lipid synthesis, namely, fatty acid elongase LOC105382588, fatty acid desaturases LOC105394761 and fatty acid reductase LOC105392472 (Fig. S5A); three protease DEGs, namely, aminopeptidase LOC105384366, carboxypeptidase LOC105380035 and LOC105393958 (Fig. S5B); two cuticle pigmentation DEGs, tan and yellow (Fig. S5C) and 11 cuticle protein genes (Fig. S5D) were selected because they are in highly expressed in the cuticle. The above DEGs were significantly affected after parasitization and restored when CvBV_28 − 5 was silenced.

Ecdysteroid usually exerts its effects on ecdysis-related genes through ecdysone response genes, E74, E75, BR-C, HR3, HR4, FTZ-F1β, etc. The conserved transcription factor network, named “Ashburner cascade”, translates the 20E peak into the different phases of the ecdysis [20]. Since CvBV_28 − 5 affects 20E titer during ecdysis, we also performed qPCR for the ecdysone response genes after CvBV_28 − 5 was silenced. The results showed that the relative mRNA levels of E78, HR3 and HR4 (Fig. S6) at stage I showed a clear downregulation in parasitized hosts and partially recovered after CvBV_28 − 5 silencing. The findings suggested that the transcriptional network about ecdysis in the larval cuticle is mainly affected by CvBV_28 − 5.

CvBV_28 − 5 affects host larval ecdysis through MAPK signaling

Based on the above results, we predicted candidate target genes for CvBV_28 − 5 from a set of genes involved in ecdysone synthesis and the downstream of ecdysone signaling in P. xylostella using two prediction methods: (1) ESMFold and MEGADOCK [27]; (2) SpeedPPI [28, 29]. A total of 65 targets with ppiscores > 6 were obtained using ESMFold and MEGADOCK. A total of six targets with pDockQ greater than 0.23 were obtained with SpeedPPI (Table S5). All predicted results of SpeedPPI are included in the predicted results of ESMFold and MEGADOCK (Fig. S7A). The six target genes are involved in the regulation of the ecdysone synthesis genes, including serotonin receptor, insulin-dependent enzyme, epidermal growth factor receptor substrate, insulin-like growth factor 2 mRNA binding protein, Raf homologous serine/threonine protein kinase Raf isoform X2 (Raf), and FMRF acylamide receptor [19] (Table S5).

We performed yeast cotransformation on selective plates to verify protein interactions between CvBV_28 − 5 and its targets (Fig. 4A, Fig. S7B). As a result, CvBV_28 − 5 positively interacted with Raf (NCBI accession number: XP_048482022), a member of MAP3K that functions downstream of the MAPK pathway (Fig. 4A). According to the CvBV_28 − 5/Raf complex molecular modeling, four residues, Asn127, Arg129, Gln130, and Val136 in CvBV_28 − 5, formed six hydrogen bonds with five residues Glu225, Thr264, Asp623, Trp627, and Arg631 in the Raf (Fig. S7C). Since three of the five bound residues in Raf protein were located on the catalytic domain of the serine/threonine kinase, it is plausible that the CvBV_28 − 5 may influence the catalytic activity of Raf protein (Fig. S7D). The hypothesis was further corroborated by transcriptional analysis of Raf and its downstream effector ERK (Fig S7E). Specifically, transcript levels of Raf remained unchange in the PG of parasitized P. xylostella larvae compared to non-parasitized controls. In contrast, ERK expression exhibited a significant downregulation at 6 h prior to ecdysis. When Raf was knocked down in non-parasitized P. xylostella (Fig. 4B), the ecdysis duration was significantly prolonged (Fig. 4C). Consistent with the knockdown of CvBV_28 − 5 in parasitized larvae, stage I of L3/L4 ecdysis was prolonged in dsRaf-treated non-parasitized larvae (Fig. S7F). The 20E titer was also significantly reduced at -8 h and − 4 h (Fig. 4D). The relative mRNA levels of the ecdysone synthesis genes were not changed at -8 h. However, nvd and phm (Fig. 4E) transcripts were significantly downregulated at 4 h before ecdysis. The expression of sro and sad (Fig. 4E) were not changed at 4 h before ecdysis, in line with the CvBV_28 − 5 knockdown results.

Fig. 4.

Raf is the target of CvBV_28 − 5. (A) Verification of protein interactions via yeast cotransformatian. +: p53-pGBKT7×T-pGADT7, positive control; -: p53-pGBKT7×lam-pGADT7, negative control; 1: Raf (raf homolog serine/threonine-protein kinase Raf isoform X2); 2: 5-HT recepter (5-hydroxytryptamine receptor 3 A isoform X3); 3: IDE (insulin-degrading enzyme); 4: EGFR substrate 15 (epidermal growth factor receptor substrate 15-like 1); 5: insulin-like growth factor 2 mRNA-binding protein 1 isoform X6, 6: FMRFamide receptor; DDO: SD/-Trp/-Leu; QDO/X: SD/-Trp/-Leu/-His/-Ade with X-α-gal; QDO/X/A: SD/-Trp/-Leu/-His/-Ade with X-α-gal and Aureobasidin A. (B) Relative mRNA levels of Raf in parasitized P. xylostella larvae at 24 h post dsRNA injection with dsGFP treatment as a control. (C) Total ecdysis duration of nonparasitized P. xylostella larvae b with knockdown of Raf. (D) 20E titer in the hemolymph of nonparasitized plus dsGFP treated (NP + dsGFP) and (NP + dsRaf) P. xylostella larvae. n = 3 in each treatment and each biological group consists of 15–20 P. xylostella larvae. (E) Relative mRNA levels of nvd, sro, phm and sad. n = 4 in each group. Data were analyzed by Student’s t test. Values represent the means ± s.e.m. *: p < 0.05; **: p < 0.01; ***: p < 0.001; ns: no significance

Discussion

In parasitoid-host relationships, parasitoids tend to manipulate host development to promote a favorable environment for their offspring. Many studies documented parasitoid-induced changes in host development, such as the prolongation of the larval stage and the delay or arrest of pupation [2]. However, the effects of parasitoids on the host’s ecdysis have not been described previously. In this study, we investigated the influence of C. vestalis parasitization on the larval-larval ecdysis of P. xylostella and further dissected the underlying molecular mechanisms that govern the effects of C. vestalis on host larval-larval ecdysis.

Parasitism extends host larval ecdysis by impairing the 20E titer

The insect cuticle consists of three layers: the outer envelope (lipids and proteins), the middle epicuticle (crosslinked proteins and lipids), and the inner procuticle (mainly chitin covalently linked to proteins) [30]. Compared with our previous study [24] and the observation of Kiguchi and Agui [25], P. xylostella larvae ceases feeding at Stage I when the 20E titer peaks, coinciding with general apolysis and epicuticle secretion, which are related to protease and cuticle protein and lipid synthesis genes. At Stage IV, larvae initiate pigmentation of mandibles and pre-ecdysis behavior, with developmental events involving procuticle secretion and pigmentation, linked to cuticle protein and pigmentation genes.

Ecdysis is orchestrated by a rapid rise and subsequent fall in the 20E titer [20, 25]. However, our results showed that the wasp parasitism led to the disappearance of the pre-ecdysis 20E peak. This disruption in 20E dynamics likely disturbed the expression of ecdysis-related genes, leading to prolonged ecdysis. Consistently, our transcriptome analysis between NP and P larvae revealed that most DEGs were related to cuticle lipid synthesis, proteases, cuticle proteins, and pigmentation genes (Fig. S4).

CvBV modulates ecdysone synthesis to prolong ecdysis

Given the major role of PG in ecdysone biosynthesis, the effect of parasitoids on the PG of the hosts has been widely reported [13, 19]. For instance, after being parasitized by Toxoneuron nigriceps, the ecdysone synthesis genes nvd, sro, phm and sad in the PG of Heliothis virescens are also significantly downregulated [14]. Our qPCR results suggested that the ecdysone biosynthesis in the PG of P. xylostella was downregulated after parasitization. Thus, it appears that blocking the ecdysone biosynthesis is a conserved strategy of parasitoid wasps to modulate host development.

C. vestalis owns several weapons that allow it to modulate host development, including CvBV [2], venom [26] and teratocytes [31]. Teratocytes are released from the serosal membrane approximately two days after parasitization, later than the onset of ecdysis (about 36 h after parasitization) [32]. Therefore, teratocytes may not be involved in regulating the L3/L4 ecdysis. Our results showed that CvBV prolonged stages I, II, III, and IV. Conversely, the venom only affected stages I and III, with negligible effects on stages II and IV. Many previous works have shown that PDV is more effective than venom in modulating hosts’ development in PDV-carrying endoparasitoids, including the Pseudoplusia includens/Microplitis demolitor system [33] and the Heliocoverpa zea/Microplitis croceipes system [34]. Thus, we consider CvBV to be a major factor involved in ecdysone synthesis.

CvBVf_28 − 5 is a novel regulator of ecdysone biosynthesis

PDV exerts its function by expressing hundreds of virulence genes in parasitized hosts. Based on the characterized PDV genomes, the virulence genes with known domains or features in PDVs include V-ankyrin-motif genes (ank), Cys-motif genes, protein tyrosine phosphatase (PTP), BEN domain-coding genes, lectin, histone, etc [12, 35]. However, there are still plenty of virulence genes that lack known domains. As we noticed, among the ten highly PG-expressed CvBV genes, only CvBV_9 − 2 has been identified as the trigger of P. xylostella tachykinin upregulation in the parasitized host larval midgut, which lowers the host systemic lipid level [12]. However, previous studies have shown that several PDV genes are involved in host ecdysone synthesis. For instance. Falabella et al. [36]. reported that PTP encoding by Toxoneuron nigriceps bracovirus (TnBV) potentially affected host PG function by disrupting the phosphorylation balance upstream from the ribosomal S6 in the PTTH signaling cascade. Using the genetic tool of Drosophila, two polydnavirus-encoded ANK proteins, TnBVank1 [14] and TnBVank3 [16], were reported to disrupt the ecdysone biosynthesis in non-overlapping ways. Here we identified CvBV_28 − 5 as a novel regulator of insect development by directly targeting ecdysone synthesis pathways. The time-specific activity of CvBV_28 − 5 suggested a precise regulation of ecdysis in P. xylostella. The transcriptome and qPCR analysis of the ecdysone response genes and selected ecdysis-related genes after the knockdown of CvBV_28 − 5 were consistent with the developmental events of P. xylostella larvae during ecdysis, indicating a reprogramming of the ecdysis process caused by CvBV_28 − 5 after parasitization via ecdysone signaling. Our recent work has shown that a CvBV-produced miRNA (Cve-miR-novel22-5p-1) arrests host growth by modulating the expression of the host ecdysone receptor [2]. It is reasonable to assume that the functional diversification of BV-derived effectors reveals the sophisticated strategies to modulate host development.

CvBV_28 − 5 regulates ecdysone synthesis through the MAPK signaling

In the host-parasitoid interaction, few studies have shown that several ecdysone-associated signaling pathways are influenced by PDV in parasitized hosts. However, direct inhibition of MAPK signaling in host-parasitoid interactions has not been investigated. The archetypal MAPK pathway proceeds from Ras (a small GTP-binding protein) to MAP3Ks (MAP kinase kinase kinases) to MAP2Ks (MAP kinase kinases) to extracellular signal-regulated kinases (ERKs), triggering downstream responses, such as development, immunity and stress response [37, 38]. In insects, such as Lepidoptera and Diptera, the MAPK pathway is indispensable for ecdysone synthesis, acting as a hub for extracellular signaling, including PTTH, Egfr, etc [39–42]. Our results showed that Raf was the direct target of CvBV_28 − 5. As a member of MAP3K, Raf transduces signals into a multistep kinase cascade upon extracellular stimulation [42]. Since CvBV_28 − 5 bound the catalytic domain of Raf, it is plausible that CvBV_28 − 5 may influence the catalytic activity of Raf, which would possibly disrupt the signaling transduction mediated Raf and its downstream kinases. However, further research is needed to understand the role of the CvBV_28 − 5/Raf complex in vivo.

In conclusion, our findings demonstrated that CvBV_28 − 5 specifically interacted with Raf, disrupting the ecdysone synthesis pathway and diminishing the 20E pulse before L3/L4 ecdysis. Correspondingly, the expression of transcription factors, such as E78, HR3 and HR4, as well as ecdysis-related genes, was affected. As a result, the third-fourth instar larval ecdysis duration of P. xylostella was prolonged (Fig. 5). To the best of our knowledge, this is the first demonstration of the molecular mechanism by which parasitoids modulate host ecdysis via MAPK signaling. These findings provide novel insights into the mechanism of parasitoid regulation of host development and the specificity of CvBV_28 − 5 and Raf interaction offers promising avenues for pest control by disrupting insect development.

Fig. 5.

Model for a Cotesia vestalis bracovirus gene, CvBV_28 − 5, affecting host ecdysone signaling and L3/L4 ecdysis. During parasitization, CvBV particles were injected into P. xylostella larvae and then infect the different host tissues. CvBV_28 − 5 is responsible for the downregulation of ecdysone in the hemolymph of P. xylostella larvae by interacting with Raf in the prothoracic gland (PG), then downregulating ecdysone synthesis genes of P. xylostella larvae. Thus, the disappearance of the 20E peak before ecdysis leads to the change of expression of ecdysone response genes like E78, HR3, and HR4, as well as genes related to a series of developmental events of ecdysis: synthesis of cuticle lipid and cuticle proteins, degradation of cuticle proteins and cuticle pigmentation

Methods

Insects

P. xylostella and its parasitoid C. vestalis used in this study were reared as previously described [12]. In summary, P. xylostella was maintained at 25 ± 1 °C and 50 ± 10% relative humidity under a 14:10 light: dark cycle with a diet of cabbage. To obtain parasitized host larvae, early-stage third instar (3 L) P. xylostella larvae (4–6 h after ecdysis) were exposed to a single C. vestalis female wasp that had emerged for two days. Adult P. xylostella and C. vestalis were fed with a 20% honey/water (V/V) solution.

Larval ecdysis observation of P. xylostella.

The parasitized P. xylostella was divided into petri dishes of 3.5 cm diameter and a piece of cabbage leaf was placed in the centre of the dish. The ecdysis was filmed using an FDR-AX100 video camera (Sony) in a climate chamber at a temperature of 25 ± 1ºC and relative humidity of 50 ± 10%. The total duration of the L3/L4 ecdysis of P. xylostella larvae in the video was then observed by Windows Media Player (version 12.0.19041.3636) and recorded.

Gene cloning

The total RNA of P. xylostella was extracted using the Total RNA Extraction Reagent Kit (Vazyme) according to the manufacturer’s instructions. The quality and concentration of the isolated total RNA were assessed by electrophoresis and a NanoDrop 2000 (Thermo Fisher Scientific). Complementary DNA was then synthesized using a HiScript III 1st Strand cDNA Synthesis Kit (Vazyme) according to the manufacturer’s instructions. The complete coding regions of the ecdysone synthesis genes of P. xylostella and CvBV genes were cloned and inserted into the pCE2 TA/Blunt-Zero vector (Vazyme). The candidate target genes of interest were cloned and inserted into the pGADT7 linear vector for yeast two-hybrid verification. All constructs were sequenced to verify sequence identity. The primers used are listed in Table S6.

CvBV and venom collection and injection

CvBV virions and venom were collected as described in our previous work [12]. The ovaries of 3-day-old C. vestalis females were dissected in PBS on ice and the calyces were punctured with forceps. The calyx fluid was centrifuged at 20,000 × g for 10 min to remove cellular debris. The virus particle pellet was resuspended in PBS. The venom reservoirs of 3-day-old female wasps were dissected in PBS on ice and were punctured with forceps. The venom fluid was filtered through a 0.22 μm filter and centrifuged at 2,000 × g for 10 min to remove cellular debris. All samples were stored in a -80 °C refrigerator until further use. Viral particles and venom fluid collected from a single adult female were defined as one female equivalent (FE). To infect host larvae, 0.05 FE CvBV particles or venom were injected into L3 P. xylostella larvae.

RNA interference

T7 High Yield RNA Transcription Kit (Vazyme) was used to synthesize double-stranded RNA (dsRNA) of CvBV genes and the GFP gene. Primers used for dsRNA synthesis are listed in Table S6. Since CvBV_19 − 3, CvBV_19 − 5 and CvBV_19 − 7 are highly similar in their nucleotide sequence of CDS, we designed siRNAs for their RNA interference. The siRNAs were synthesized by Sangon Biotech (Shanghai, China). At approximately 4 h before parasitization, 1000 ng of dsRNA was injected into L3 P. xylostella larvae and dsGFP was used as a negative control. The efficacy of the RNA interference was determined by qPCR after 24 h. The primers used in dsRNA synthesis are listed in Table S6.

Quantitative PCR

The total RNA of P. xylostella was extracted using the Total RNA Extraction Reagent Kit (Vazyme) according to the manufacturer’s instructions. The quality and concentration of the isolated total RNA were assessed by electrophoresis and a NanoDrop 2000 (Thermo Fisher Scientific).

After RNA was extracted from different tissues, cDNA was prepared from the extracted RNAs using ReverTra Ace qPCR RT Master Mix with a gDNA Remover kit (TOYOBO) according to the manufacturer’s instructions. Quantitative polymerase chain reaction (qPCR) reactions were conducted on a CFX Connect real-time system (BIO-RAD) using THUNDERBIRD qPCR Mix (TOYOBO). Each qPCR was performed for at least three biological replicates under the following conditions: The temperature was set at 95 °C for 60 s, followed by 40 cycles of 95 °C for 15 s and 60 °C for 30 s. The β-Actin gene (GenBank accession No. NM_001309101) of P. xylostella was employed as an internal control. The relative expression levels were analyzed using the 2^−ΔΔCt method. The primers used in qPCR are listed in Table S6.

Hemolymph Preparation and 20E titer measurements

Hemolymph samples of P. xylostella larvae were prepared in three biological replications with 15–20 larvae per replicate at five time points. The feeding stage takes about 24 h in nonparasitized P. xylostella larvae and 36 h in parasitized ones. According to the developmental timetable of larval molting in the Bombyx mori, the peak of 20E titer started at about the last 1/6 of the feeding period [25]. Thus, to detect the emergence of the 20E peak, a sampling time point at last 1/3 and 1/6 of the ecdysis is necessary (-8 and − 4 h of nonparasitized P. xylostella larvae and − 12 h and − 6 h of parasitized ones). The morphological character of stage I of larval-larval ecdysis is clear in P. xylostella, which can be easily adopted in sampling. Combined with the fact that the ecdysis duration takes about 12 h in nonparasitized P. xylostella larvae, and 14 h in parasitized ones. The + 6 h in nonparasitized P. xylostella larvae and the + 7 h in parasitized ones represent the 20E titer at half of the ecdysis process. Finally, we collected the larvae just ecdysis to L4 at about 12 h in nonparasitized ones and 14 h in parasitized ones (Fig. 1B).

Larvae were individually pierced through the cuticle with forceps to collect the hemolymph. Hemolymph from each replicate was collected on an ice plate and put into an EP tube with 200 µl of methanol. Samples were homogenized and centrifuged (10 min at 18000 g). The remaining tissue was re-extracted twice in 200 µl methanol and the resulting methanol supernatants were dried using a dry bath at 70 °C. Samples were resuspended and then ELISA was performed according to the manufacturer’s instructions using a commercial ELISA kit (Bertin Bioreagents) that detects ecdysone and 20E with the same affinity. Absorbance was measured at 405 nm on a plate reader, SpectramaxPlus (Molecular Devices, Sunnyvale, CA).

Transcriptome analysis

The cuticle of the nonparasitzed (NP), parasitized (P), parasitized with dsGFP injected (P + dsGFP) and parasitized with ds28-5 injected (P + ds28-5) P. xylostella larvae during stage I of L3/L4 ecdysis were dissected under a stereo microscope in PBS.

To identify highly expressed CvBV genes in the prothoracic gland of P. xylostella during early parasitization, parasitized prothorax samples were collected and analyzed at 12 and 6 h before larval-larval ecdysis. Due to the small size of the prothoracic gland, the prothorax was collected instead [43].

Total RNA was isolated using a TRIzol reagent according to the manufacturer’s instructions. The quality and concentration of the isolated total RNA were estimated by electrophoresis and a NanoDrop 2000. The transcriptomes were produced by HaploX (Shenzhen, China). cDNA libraries for transcriptome sequencing made from total RNA were prepared using the NEBNext Ultra RNA Library Prep Kit (NEB #E753) for Illumina in conjunction with the NEBNext Poly(A) mRNA Magnetic Isolation Module (NEB #E7490). The libraries were validated and quantified before being pooled and sequenced on an Illumina NovaSeq 6000 (Illumina) sequencer with a 150 bp paired-end protocol. The raw reads were not less than 6 G and cleaned by fastp (version 0.23.2) [44]. Then, the paired-end clean reads were aligned to the P. xylostella genome (GCF_932276165.1) from the National Center for Biotechnology Information (NCBI) by STAR (version 2.7.10a) [45] and then quantified by featureCount (version 2.0.1) [46]. The following analysis was performed on R (version 4.2.2). PCA was performed by the factoextra package (version 1.0.2). The detection of differentially expressed genes was performed by the DESeq2 package (version 1.36.0) [47]. The GO and KEGG enrichment was performed by clusterProfiler (version 4.6.0) [48] and enrichplot (version 1.18.3) package.

To analyze CvBV genes expressed in P. xylostella prothoracic gland, the index of the CvBV genome (NCBI accession number MZ645181.1-MZ645210.1) was built by RSEM (version 1.3.3) [49]. The FPKM (Fragments Per Kilobase of exon model per Million mapped reads) was directly calculated by RSEM. The highly expressed genes mean the total FPKM of a CvBV gene at five time points is greater than average. The following heatmap was performed with Omicshare online (https://www.omicshare.com/tools/home/report/reportheatmap.html) and was normalized with Z-score.

Phylogenetic analysis and protein structure prediction

In the phylogenetic analysis, the homologous genes of CvBV_28 − 5 were retrieved from NCBI using BLASTP. A maximum likelihood tree was performed by MEGA 7 [50] with default parameters, 1000 bootstrap replications, and substitution with the JTT (Jones-Taylor-Thornton) model.

The conserved domain of Raf was identified by the Conserved Domain Search Service (CD Search) on NCBI. The three-dimensional structures of the interaction complexes between CvBV_28 − 5 and Raf were predicted using SpeedPPI v0.1.0 with default parameters [28]. The molecular modeling graphics of the complexes were generated using PyMOL 3 [51].

Prediction of interactions between CvBV_28 − 5 and P. xylostella proteins

The pre-trained ESMFold_v1 model in HuggingFace transformers v4.35.2 was used to predict the three-dimensional structures of CvBV_28 − 5 and a series of proteins related to developmental processes in diamondback moth. MEGADOCK 4.0 was used to dock the above three-dimensional models and evaluate their binding potential using PPIscore. According to Hwang et al. [27], the prediction results with ppiscore > 6.0 were selected for presentation.

Directly use SpeedPPI v0.1.0 [28] to predict the interactions between CvBV_28 − 5 and a series of proteins related to developmental processes in diamondback moth. According to the criteria of Bryant et al. [29], the predicted results with pDockQ > 0.23 were selected for presentation.

Yeast two-hybrid verification

The Y2H verification was performed by Yeastmaker™ Yeast Transformation System (Clontech). The CDS of CvBV_28 − 5 and six predicted targets of CvBV_28 − 5 was inserted into the pGBKT7 and pGADT7 vector via homologous recombination, respectively. The 28-5-pGBKT7 and pGADT7 vectors with predicted targets were introduced together into the Y2HGold strain of Saccharomyces cerevisiae. The empty pGBKT7 vector and pGADT7 vectors with predicted targets served as controls. T7 and 3’AD primers and T7 and 3’BD primers were used to confirm the success of transformation. Finally, a suspension of successfully transformed clones was dropped onto DDO, QDO, and QDO/X-α-gal plates. Interactions of proteins were observed and screened according to the manufacturer’s instructions. p53-pGBKT7 and T-antigen-pGADT7 were used as positive controls, while lam-pGBKT7 and T-antigen-pGADT7 were used as negative controls.

Quantification and statistical analysis

All statistical analyses were performed using GraphPad Prism 8. All data were first tested for normality. If they conformed to a normal distribution, a student’s t-test was performed; otherwise, the Mann-Whitney test was used, with a significance threshold of p < 0.05. All data are presented as mean ± s.e.m. *: p < 0.05, **: p < 0.01 and ***: p < 0.001.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

Conceptualization: Xuexin Chen, Zhizhi Wang; Methodology: Xuexin Chen, Xiqian Ye, Zhizhi Wang, Jianhua Huang; Investigation: Licheng Gu, Mujuan Guo, Pengzhan Wang, Zhiwei Wu, Zihan Wang, Sijie Zhang, Xin Yang; Funding acquisition: Xuexin Chen, Zhizhi Wang; Software: Licheng Gu, Pengzhan Wang, Jianchao Zhao; Resources: Licheng Gu, Mujuan Guo, Pengzhan Wang, Zhiwei Wu, Zihan Wang, Sijie Zhang, Xin Yang, Ruofei Ma, Lizhi Wang; Writing—original draft: Licheng Gu; Writing—review & editing: Licheng Gu, Xuexin Chen, Jianhua Huang, Zhizhi Wang; Supervision: Xuexin Chen.

Funding

This research was funded by the National Natural Science Foundation of China (U22A20485 and 32272607) and the Fundamental Research Funds for the Central Universities (226-2024-00070).

Data availability

Source data for all the graphs and tables was provided as numerical values in separate excel sheets. Each excel sheet corresponds to a graph or table, the sheets are labelled accordingly. The transcriptome dataset generated during this study has been deposited in the NCBI under accession numbers PRJNA1201736 (NP), PRJNA1204181 (P), PRJNA1201742 (P + dsGFP), and PRJNA1201751 (P + ds28-5), which correspond to the cuticle of P. xylostella larvae, while PRJNA1201864 (6 h before ecdysis, i.e. -6 h) and PRJNA1201865 (12 h before ecdysis, i.e. -12 h) correspond to the PG of P. xylostella larvae.

Declarations

Competing Interests

The authors have no relevant financial or non-financial interests to disclose.