A plant virus manipulates the long-winged morph of insect vectors

Significance

Wing dimorphism is a typical phenotypic plasticity in insects. Most cases reported previously involve indirect regulation of plant viruses to wing dimorphism of insect vectors through host plants. Here, we revealed that a plant virus directly induces a long-winged morph in male insect vectors. This regulation is mediated by a species-specific unclassified gene, which was proven a downstream factor of the insulin/insulin-like growth factor signaling pathway. The long-winged insect vectors induced by viruses in turn facilitate viral long-distant dispersal and large-scale epidemics. Our work presents not only a regulatory mechanism of insect wing plasticity but also a perspective of understanding the co-evolution of plant viruses with their insect vectors.

Abstract

Wing dimorphism of insect vectors is a determining factor for viral long-distance dispersal and large-area epidemics. Although plant viruses affect the wing plasticity of insect vectors, the potential underlying molecular mechanisms have seldom been investigated. Here, we found that a planthopper-vectored rice virus, rice stripe virus (RSV), specifically induces a long-winged morph in male insects. The analysis of field populations demonstrated that the long-winged ratios of male insects are closely associated with RSV infection regardless of viral titers. A planthopper-specific and testis-highly expressed gene, Encounter, was fortuitously found to play a key role in the RSV-induced long-winged morph. Encounter resembles malate dehydrogenase in the sequence, but it does not have corresponding enzymatic activity. Encounter is upregulated to affect male wing dimorphism at early larval stages. Encounter is closely connected with the insulin/insulin-like growth factor signaling pathway as a downstream factor of Akt, of which the transcriptional level is activated in response to RSV infection, resulting in the elevated expression of Encounter. In addition, an RSV-derived small interfering RNA directly targets Encounter to enhance its expression. Our study reveals an unreported mechanism underlying the direct regulation by a plant virus of wing dimorphism in its insect vectors, providing the potential way for interrupting viral dispersal.

Many insects have evolved the ability to adapt to changing environments by phenotypic plasticity, in which conspecifics display different phenotypes in response to different environments without genotype variation (1). Wing dimorphism, also called wing polymorphism, is a typical example of phenotypic plasticity in insects (2). Wing dimorphism is an effective and efficient strategy for insects to cope with the heterogeneous environments (3). Several environmental factors, such as population density, host plant quality, temperature, and photoperiod, are well known to affect insect wing dimorphism (4). In general, wing dimorphism influenced by environmental factors is closely associated with the tradeoff between fecundity and dispersal of insects. Moreover, winged or long-wing morphs can migrate new habitats and facilitate the transmission of the virus associated with the vector insects. Thus, different wing morphs of insects cause different damage scenarios in crops (4–9).

Wing plasticity in insects is mediated at multiple levels from external environmental cues to physiological pathways of insects. Breakthroughs in elucidating the molecular bases for wing dimorphism have been achieved in planthoppers and aphids in recent years. The insulin/insulin-like growth factor signaling (IIS) pathway and related factors play a critical role in controlling planthopper wing dimorphism. The long-winged morph is positively regulated by insulin receptor 1 (InR1), Akt and vestigial and is negatively regulated by insulin receptor 2 (InR2), transcription factors FoxO and Zfh1 in planthoppers (10–12). In addition, the c-Jun amino terminal kinase (JNK) signaling pathway and genes of wingless and juvenile hormone epoxide hydrolase (JHEH) are also involved in planthopper wing dimorphism (10, 11, 13–15). The ecdysone receptor was found to regulate transgenerational wing dimorphism in pea aphid Acyrthosiphon pisum (16). In other aphid species, microRNAs (miRNAs) are involved in the regulation of wing plasticity, potentially through their post-transcriptional action on the IIS pathway (17). Interestingly, the cucumber mosaic virus transmitted by aphid vectors has evolved to manipulate wing plasticity through changes in plant volatiles and host plant quality (18). Epidemiologically, plant virus spread is correlated with the number of winged vectors feeding on virus-infected plants, thus wing morph is an important factor driving viral transmission (19).

Planthoppers are important vectors of plant viruses, especially the small brown planthopper (SBPH, Laodelphax striatellus), brown planthopper (Nilaparvata lugens), and white-backed planthopper (Sogatella furcifera), which are the most notorious rice pest insects due to their powerful vector competence for multiple rice viruses (8, 20, 21). Rice stripe virus (RSV) is currently considered one of the most economically serious rice viruses in East Asia (22). This virus is a typical Tenuivirus, composed of four single-stranded RNA segments that encode a nucleocapsid protein (NP), an RNA-dependent RNA polymerase, and five nonstructural proteins (23). RSV is efficiently transmitted by SBPH in a persistent-circulative manner and is capable of proliferating in SBPH (23). The three species of planthoppers are all characterized by alternative wing morphs, but the migration capacity of brown planthopper and white-backed planthopper is higher than that of SBPH (10, 24). It has been reported that the corn planthopper, Peregrinus maidis, is prone to mature as long-winged morphs when reared on maize mosaic virus-infected old leaves of corn plants (25). Although plant viruses are capable of manipulating the wing plasticity of insect vectors, the underlying molecular mechanisms by which plant viruses directly regulate the wing plasticity of insect vectors remain poorly understood.

In this study, we found that RSV specifically induces a long-winged morph in male SBPHs without resorting to external environmental cues. A function-unknown gene, named “Encounter,” was found to regulate wing dimorphism in male SBPH. Encounter is closely connected with the IIS pathway as a downstream factor of Akt. RSV upregulates Encounter expression by activating the transcription of Akt and by targeting the untranslated terminal region of Encounter using a virus-derived small interfering RNA. The long-winged insect vectors induced by viruses in turn have great significance in the broad spread of viruses and viral disease epidemics.

Results

RSV Infection Induces the Long-Winged Morph in Male Planthoppers.

The SBPH strain raised in the laboratory demonstrated typical wing dimorphism in the adult stage. The long-winged male (LWM) and female (LWF) adults had fully developed forewings and hindwings, both extending from the abdomen. The forewings of the short-winged male (SWM) and female (SWF) adults were remarkably shorter than the abdomen, and the hindwings were shrunken to mere buds (Fig. 1A). Wing dimorphism was undiscernible in larvae even at the late larval stage (Fig. 1A).

Fig. 1.

RSV infection induces the long-winged morph in male planthoppers. (A) Morphological features of wing dimorphism in male and female adults and fifth-instar larvae of SBPH. Forewings at the right side were removed to better display hindwings in adults. LWM, long-winged male. SWM, short-winged male. LWF, long-winged female. SWF, short-winged female. (B) Ratio of long-winged morphs in viruliferous (V) and nonviruliferous (NV) males and females. (C) Relative RNA level of RSV NP and ratio of LWM in the next generation of SBPH. The nonviruliferous parent SPBH was injected with RSV crude extracts at the third-instar stage. The control group was injected with H₂O. (D) Relative RNA level of RSV NP and ratio of LWM in two laboratory-reared viruliferous strains with different viral titers. (E) Relative RNA level of RSV NP in viruliferous third-instar SBPH larvae with injection of double-stranded RNA for NP (dsNP) for 5 d and ratio of LWM after injection. The control group was injected with dsPsbp. Data comparison between two groups was performed using Student’s t test. N.S., no significant difference. **P < 0.01. ***P < 0.001.

We compared wing dimorphism between RSV-bearing and nonviruliferous SBPHs. To minimize the influence of environmental factors on insect wing dimorphism, 16 insects were raised on six rice seedlings under the conditions of 24 °C with 16 h of light daily. Under this condition, the ratios of LWM were 79% in the RSV-bearing strain and only 37% in the nonviruliferous strain, whereas all females developed into the short-winged morph in both strains (Fig. 1B). When the nonviruliferous planthoppers were inoculated with the crude extract of RSV, the ratio of LWM in the next generation was significantly increased from 39 to 72% (Fig. 1C). In two viruliferous strains with a 5,383-fold difference in RSV titer, the ratios of LWM were similar, i.e., 83% and 90% (Fig. 1D). When the viral titer was reduced with the injection of double-stranded RNA (dsRNA) for RSV NP in viruliferous insects, the ratio of LWM was still as high as 95% (Fig. 1E). These results from the laboratory strains indicated that RSV infection specifically induces the long-winged morph in male SBPHs in a viral dose-independent manner.

Wing Morphs Are Closely Associated with RSV Infection in Male Planthoppers from Field Populations.

The correlation between wing morph and RSV infection was also investigated in field-caught populations of SBPH. Nine field populations were collected from the main distribution regions in China during June to August, including the single-cropping northeast (Shenyang), single- and double-cropping central area (Lianyungang, Yancheng, Xuzhou, Taizhou, Nantong, and Kaifeng), and single- and double-cropping southwest plateau (Kunming and Chuxiong) (Fig. 2A). The males of all nine populations were 100% long-winged morphs, and the RSV-bearing ratios were over 70% (Fig. 2B). However, the RSV titers in the males of different field populations varied greatly, with dozens or hundreds of differences (Fig. 2C). For the females, the long-winged ratios of the nine populations fluctuated between 64% and 100% and the RSV-bearing ratios were between 36% and 100% (Fig. 2D). The two ratios were not always closely correlated, as shown by the remarkable divergence in the Kaifeng, Xuzhou, and Chuxiong populations. The RSV-bearing ratios of the three populations were below 50% while their long-winged ratios were over 90% (Fig. 2D). The results from the field populations demonstrated that the long-winged ratios of male insects are closely associated with RSV infection regardless of viral titers.

Fig. 2.

Wing morphs are closely associated with RSV infection in male planthoppers from field populations. (A) Sampling locations of nine field populations in China. Note: This image is based on a standard map obtained from the website of the National Administration of Surveying, Mapping and Geoinformation with approval number GS(2020)3184. The base map is not modified. (B) Ratio of long-winged morphs and RSV-bearing ratio for male adults from the field populations. Numbers in brackets are male individuals in each population. (C) Relative RNA level of RSV NP in single males of the field populations. (D) Ratio of long-winged morph and RSV borne ratio for females from the field populations. Numbers in brackets are female individuals in each population.

Interference of Encounter Decreases the Ratio of Long-Winged Males.

In the RSV NP knockdown experiment (Fig. 1E), in addition to injecting dsRNA for the plant gene psbp (dspsbp-RNA) as a negative control, 420 bp of dsRNA for green fluorescent protein (dsGFP-RNA) was also applied in another negative control group. Surprisingly, the ratio of LWM in the viruliferous insects decreased to 2% with dsGFP-RNA injection, and some males (less than 10%) exhibited curled wings in addition to the normal short-winged morph (Fig. 3A). The 420-bp sequence was located from 65 to 484 bp of GFP (LC337077.1 in GenBank). To narrow down the region that affected the wing morph, two fragments of dsGFP-RNA, i.e., dsGFP1-RNA containing the first 252 bp (from 65 to 316 bp) and dsGFP2-RNA containing the latter 168 bp (from 317 to 484 bp), were synthesized and injected into viruliferous third-instar larvae. Only dsGFP1-RNA decreased the LWM ratio to 6%, while dsGFP2-RNA did not affect the LWM ratio (Fig. 3B). The dsGFP1-RNA was further shortened into two fragments, i.e., dsGFP1-1-RNA containing the first 130 bp (from 65 to 194 bp) and dsGFP1-2-RNA containing the latter 122 bp (from 195 to 316 bp). Both fragments affected the LWM ratio, but dsGFP1-1-RNA dramatically decreased the LWM ratio to 5% (Fig. 3B). Insects do not have the GFP gene. The 130 bp of dsGFP1-1-RNA may have interfered with the expression of SBPH genes due to the sequence similarity in a limited region.

Fig. 3.

Interference of Encounter (Contig2490.1) decreases the ratio of long-winged males. (A) Ratio of long-winged males (LWM) in viruliferous SBPHs that were injected with double-stranded RNA for GFP (dsGFP) at the third-instar stage. Injection of dsPsbp was used as the control. Values were compared by Student’s t test. Images for normal short-winged morph and curled wings in the dsGFP treatment are displayed on the left and right panels, respectively. (B) Ratio of LWM in viruliferous SBPHs that were injected with dsRNA for different fragments of GFP at the third-instar stage. The model diagram demonstrates the localization of distinct GFP fragments for dsRNA synthesis. Different letters indicate significant differences in Tukey’s multiple comparison test. (C) Morphological features of reproductive tissues from adults and fourth-instar larvae. (D) Relative transcript levels of six SBPH genes in third-instar male larvae after injection of dsGFP1-1 for 5 d. Different letters indicate significant differences in Tukey’s multiple comparison test. (E) Fold change of Contig2490.1 transcript level in viruliferous third-instar larvae after injection of ds2490.1 for 5 d compared to that with dsPsbp injection and ratio of LWM. Values were compared by Student’s t test. **P < 0.01. ***P < 0.001.

We performed a BLAST analysis in the gene set of SBPH with 130 bp of GFP1-1 sequence, and six genes were found to be identical to GFP1-1 with 15 or 16 bp of length (Table 1). The transcript levels of the six genes were measured in male larvae after injection with dsGFP1-1-RNA or dspsbp-RNA. Larval sexes were discriminated according to similar anatomical morphological features of reproductive organs between larvae and adults. The male testis had three seminiferous tubules, and the female ovary contained multiple ovarioles (Fig. 3C). In the dsGFP1-1-RNA-injected male larvae, one gene, Contig2490.1, which was annotated as malate dehydrogenase, showed a significant transcriptional decrease compared to the dspsbp-RNA-injected insects, whereas the expression of the other five genes was not affected (Fig. 3D). When the expression of Contig2490.1 was knocked down with the injection of its dsRNA in the third-instar viruliferous SBPHs, the LWM ratio was reduced to 34% (Fig. 3E), and 9.1% males exhibited curled wings in addition to the normal short-winged morph, as seen with the dsGFP-RNA injection. Therefore, the expression of Contig2490.1 was inadvertently affected by dsGFP-RNA, leading to a decrease in the LWM ratio in viruliferous planthoppers. Due to the unusual detection process, we named Contig2490.1 the “Encounter” gene.

Table 1.

Information of six SBPH genes identical to GFP

SBPH gene ID	GFP position (bp)	Identities	Annotation
Contig394.3	84 to 98	15/15 (100%)	Chromodomain-helicase-DNA-binding protein Mi-2-like protein
Contig91.22	88 to 103	16/16 (100%)	Uncharacterized protein LOC105691925
Contig2490.1	107 to 121	15/15 (100%)	Malate dehydrogenase, mitochondrial isoform X1
Contig346.25	137 to 151	15/15 (100%)	Fat protein
Contig113.63	146 to 160	15/15 (100%)	Coiled-coil domain-containing protein AGAP005037 isoform X3
Contig852.1	150 to 164	15/15 (100%)	Uncharacterized protein LOC105691685 isoform X1

Encounter Is a Planthopper-Specific Protein without Malate Dehydrogenase Activity.

Based on the chromosome-level genomes of SBPH, white-backed planthopper Sogatella furcifera, and brown planthopper Nilaparvata lugens (26), the three planthopper species had one Encounter gene on the No. 5 chromosome. Encounter was not found in other insects with the search in GenBank. SBPH’s Encounter putatively encoded a 38.2-kDa protein with 357 amino acid residues containing the malate dehydrogenase glyoxysomal mitochondrial domain (Fig. 4A). Phylogenetic analysis showed that the planthopper Encounter cluster was divergent from the mitochondrial or cytoplasmic malate dehydrogenases of insects (Fig. 4B). When SBPH’s Encounter was expressed in human 293T cells, this protein was localized in mitochondria (Fig. 4C). However, the recombinantly expressed and purified Encounter did not exhibit malate dehydrogenase activity as the typical mitochondrial malate dehydrogenase of SBPH (Contig66.45) (Fig. 4D). Therefore, Encounter was determined to be a kind of planthopper-specific gene, and its encoded protein did not belong to classical malate dehydrogenases.

Fig. 4.

Encounter is a planthopper-specific protein without malate dehydrogenase activity. (A) Protein domain analysis of Encounter. MDH, malate dehydrogenase. (B) A maximum likelihood phylogenetic tree showing the relation of Encounter to insect mitochondrial and cytoplasmic MDH. Bootstrap values higher than 70% are shown at the nodes. GenBank registration numbers are given in parentheses. (C) Localization of the recombinantly expressed Encounter-V5 in human 293T cells. Encounter-V5 was labeled with an anti-V5 monoclonal antibody (green). Mitochondria and nuclei were labeled by mitochondria BacMam 2.0 (red) and Hoechst (blue), respectively. (Scale bar, 2 μm.) (D) Relative NAD-MDH activity of recombinantly expressed and purified Encounter-His. BSA was used as a negative control. Mitochondrial MDH of SBPH (mit MDH-His, Contig66.45) was used as a positive control. Western blotting shows the purified proteins using the anti-His monoclonal antibody.

Encounter Is Highly Expressed in the Testis and Upregulated by RSV to Affect Male Wing Dimorphism at Early Larval Stages.

To determine the expression patterns of Encounter, male and female adults, different tissues of male adults, and different larval stages of nonviruliferous and viruliferous planthoppers were investigated. Quantitative real-time PCR (qPCR) showed that the transcript level of Encounter in male adults was 557-fold that in female adults (Fig. 5A). In male adults, Encounter was predominantly expressed in the testis in contrast to the head, gut, fat body, and wing (Fig. 5B). During larval stages, the expression of Encounter increased since the second instar in the viruliferous planthoppers (Fig. 5C) while since the fourth instar in the nonviruliferous planthoppers (Fig. 5D). Compared to nonviruliferous planthoppers, the transcript levels of Encounter were upregulated at the second-, third-, and fourth-instar larval stages of the viruliferous planthoppers (Fig. 5E). When sexes were discriminated in third-instar larvae, the viruliferous male third-instar larvae had a higher transcript level of Encounter than the nonviruliferous counterpart (Fig. 5F), and Encounter was almost exclusively expressed in the testis of third-instar larvae (Fig. 5G). To clarify the key regulatory period for wing morph differentiation, the expression of Encounter was knocked down in second-, third-, or fourth-instar viruliferous planthoppers with the injection of dsEncounter-RNA. The LWM ratio was markedly reduced with interference at the second and third instars but not at the fourth instar (Fig. 5H), indicating that Encounter regulated male wing dimorphism before the fourth instar larval stage.

Fig. 5.

Encounter is highly expressed in the testis and upregulated by RSV to affect male wing dimorphism at early larval stages. (A–G) Fold change of Encounter transcript levels in nonviruliferous male and female adults (A), six tissues of nonviruliferous male adults (B), the first- to fifth-instar stages of viruliferous and nonviruliferous larvae (C–E), viruliferous (V) and nonviruliferous (NV) third-instar male larvae (F), and six tissues of viruliferous male third-instar larvae (G). The fold change of each sample is relative to the Encounter transcript level of nonviruliferous males in panel (A). (H) Ratio of long-winged males (LWM) in viruliferous SBPHs that were injected with double-stranded RNA for Encounter (dsEncounter) at the second-, third-, or fourth-instar stages. The control group was injected with dsPsbp. For (A), (E), (F), and (H), values were compared by Student’s t test. N.S., no significant difference. *P < 0.05. **P < 0.01. ***P < 0.001. For (B), (C), (D), and (G), different letters indicate significant differences in Tukey’s or Duncan’s multiple comparison test.

Encounter Regulates Male Wing Dimorphism Dependent on the IIS Pathway.

To clarify the functional mechanisms of Encounter in regulating male wing dimorphism, the transcript levels of the IIS pathway and related factors (InR1, InR2, Akt, FoxO), Vestigial, JNKs, Wingless, and JHEH were compared between third-instar male larvae of the nonviruliferous and viruliferous SBPH using qPCR. The results showed that Akt, which is a positively acting component of the IIS pathway, was upregulated in viruliferous male larvae while other eight genes did not change (Fig. 6A). When the expression of Akt was knocked down with the injection of dsAkt-RNA in the third-instar viruliferous larvae, the LWM ratio was significantly reduced to 12% and 7.2% males exhibited curled wings. The transcript level of Encounter was also reduced (Fig. 6B). Inactivation of InR1, another positively acting component of the IIS pathway and upstream of Akt, also decreased the LWM ratio in the viruliferous SBPH but did not affect the transcriptional level of Akt and Encounter (Fig. 6C). Knockdown of FoxO, an effector downstream of the IIS pathway, did not affect Encounter transcript level (Fig. 6D). Meanwhile, inactivation of Encounter did not affect the transcriptional level of InR1, InR2, Akt, and FoxO (Fig. 6E). These findings indicate that Encounter regulates male wing dimorphism dependent of the IIS pathway and Encounter is positioned downstream of Akt, of which the transcriptional level is upregulated in response to RSV infection, resulting in the elevated expression of Encounter.

Fig. 6.

Encounter regulates male wing dimorphism dependent on the IIS pathway. (A) Relative transcript levels of nine genes regulating insect wing dimorphism in third-instar male larvae of the nonviruliferous (NV) and viruliferous (V) SBPH. Different letters indicate significant differences in Tukey’s multiple comparison test. (B–E) Compared to that after injection of double-stranded RNA of Psbp (dsPsbp) in viruliferous third-instar male larvae for 5 d, fold change of Akt and Encounter transcript levels with dsAkt injection and ratio of long-winged males (LWM) (B), fold change of InR1, Akt, and Encounter transcript levels with dsInR1 injection and ratio of LWM (C), fold change of FoxO and Encounter transcript levels with dsFoxO injection (D), and fold change of InR1, InR2, Akt, and FoxO transcript levels with dsEncounter injection (E). Values were compared by Student’s t test. N.S., no significant difference. *P < 0.05. **P < 0.01. ***P < 0.001.

RSV-Derived Small Interfering RNAs Directly Upregulate Encounter Expression.

In addition to the indirect regulation of Akt on Encounter expression, we further explored the direct interaction of RSV with Encounter. RSV only encodes seven proteins, but it produces a large amount of virus-derived small interfering RNAs (vsiRNAs), which can act on target genes in a microRNA (miRNA) manner with partial sequence complementarity (27). In mammals, miRNAs enhance gene transcription or translation by binding the 5′ untranslated terminal region (5′UTR) of target genes (28, 29). Thus, we explored whether there are RSV-derived vsiRNAs targeting the 5′UTR of Encounter. Based on the full-length transcriptome of SBPH (30), a 91-bp sequence upstream of the start codon of Encounter was identified and verified through cloning and sequencing. RNAhybrid and miRanda algorithms predicted four and two candidate vsiRNAs putatively targeting the 91 bp of the Encounter 5′ UTR, respectively, from the library of RSV candidate vsiRNAs (31). qPCR showed that two candidate vsiRNAs, vsiR-7880 (5′-UUCUCUAGCGAAUUUUCUGGCU-3′) and vsiR-2675 (5′-AUUCUCCGUGGACGAUUUCUGUA-3′), were uniquely detected in viruliferous planthoppers, while other four candidate vsiRNAs were detected in both viruliferous and nonviruliferous planthoppers (Fig. 7A). Sanger sequencing confirmed the identity of vsiR-7880 and vsiR-2675, which putatively targeted the sites from -16 to -37 bp and from -16 to -36 bp of the Encounter 5′ UTR, respectively.

Fig. 7.

RSV-derived small interfering RNAs upregulate Encounter expression. (A) Relative RNA levels of six putative vsiRNAs in nonviruliferous (NV) and viruliferous (V) third-instar larvae. (B) Fold change of vsiR-2675 and vsiR-7880 RNA levels and Encounter transcript levels in viruliferous larvae after injection of vsiR-2675 or vsiR-7880 activator for 3 d compared to that with injection of negative control (activator-NC). Different letters indicate significant differences in Tukey’s multiple comparison test. (C) and (D) Fold change of vsiR-7880 RNA levels and Encounter transcript levels in viruliferous male larvae after injection with the activator or inhibitor of vsiR-7880 for 3 d. Activator-NC or inhibitor-NC was injected as a negative control. The fold change of Encounter transcript levels is relative to that of activator-NC in panel (B). (E) Dual luciferase reporter assays in Drosophila S2 cells cotransfected with recombinant psiCHECK2 plasmids containing the target sequences of vsiR-7880 and vsiR-7880 mimics. The activity of Renilla luciferase (Rluc) relative to that of firefly luciferase (Fluc) is presented. Mimic-NC was used as a negative control. Tar-WT, wild-type target. Tar-MT, mutant target. (F) Relative RNA level of vsiR-7880 in the Ago1-immunoprecipitated complexes from RIP-qPCR assays performed in viruliferous larvae. IgG was used as a negative control. (G) Relative enrichment of vsiR-7880 target (Encounter 5′UTR) in the Ago1-immunoprecipitated complexes to that of IgG-immunoprecipitated complexes after injection of vsiR-7880 activator in nonviruliferous larvae. Activator-NC was injected as a negative control. Values were compared by Student’s t test. N.S., no significant difference. *P < 0.05. **P < 0.01. ***P < 0.001.

The impact of these two vsiRNAs on Encounter transcription was first analyzed by delivery of the synthetic activator of each vsiRNA into nonviruliferous fourth-instar larvae. The transcript level of Encounter was significantly increased with injection of the vsiR-7880 activator, while it was not affected by injection of the vsiR-2675 activator (Fig. 7B). When male larvae were isolated for specific measurement, the vsiR-7880 activator also significantly upregulated Encounter expression (Fig. 7C). In viruliferous male larvae, injection of the vsiR-7880 inhibitor resulted in a considerable reduction in the transcript level of Encounter (Fig. 7D).

Direct interaction between vsiR-7880 and the Encounter 5′UTR was verified using dual luciferase assays in Drosophila S2 cells. Relative luciferase activities in S2 cells transfected with the construct containing the Encounter 5′UTR were higher than those of the control group in the presence of 20 and 100 pM vsiR-7880 mimic (Fig. 7E). Mutations of 7 bp at the target site corresponding to the seed region of vsiR-7880 abolished the effects of the vsiR-7880 mimic on luciferase activities (Fig. 7E). RNA immunoprecipitation combined with qPCR (RIP-qPCR) was performed in viruliferous larvae using an anti-Ago1 monoclonal antibody (32). Compared to the IgG negative control, vsiR-7880 was enriched in the Ago1-immunoprecipitated complexes (Fig. 7F). After injection of the vsiR-7880 activator, the 5’UTR of Encounter was enriched in the Ago1-immunoprecipitated complexes in contrast to the injection of a control activator (Fig. 7G). These results demonstrated that RSV-derived vsiR-7880 enhanced the transcription of Encounter by targeting its 5′UTR.

Discussion

In our study, we provide evidence that a plant virus directly manipulates wing dimorphism of its insect vectors when it propagates within insect vectors. This regulation is sex-biased and mediated by a planthopper-specific unclassified Encounter gene, which is closely connected with the IIS pathway as a downstream factor of Akt. In the canonical IIS pathway, InR1 promotes the phosphorylation of Akt, which phosphorylates the transcriptional factor FoxO, leading to the development of long-winged morphs (10). InR2 binds to InR1 to create a heterodimer, which deactivates the PI(3)K-Akt cascade and the following FoxO phosphorylation, resulting in the development of short-winged morphs (10). In our work, we found that RSV adopted a different approach to affect the IIS pathway (Fig. 8). The virus activated the transcription of Akt and its downstream factor Encounter. Inactivation of InR1 reversed the long-winged phenotype but did not affect the transcriptional level of Akt and Encounter. These findings indicate that the transcriptional activation and protein phosphorylation are paralleled regulatory modes for Akt to control wing dimorphism in SBPH. As for the downstream effectors FoxO and Encounter, the expression of Encounter did not affect the transcriptional level of FoxO and vice versa. It is possible that the elevated expression of Encounter promotes the accumulation of phosphorylated FoxO. As a collateral branch of the classical IIS pathway, the regulation from Akt to Encounter would have a mild effect on FoxO phosphorylation to finely tune the formation of long-winged morphs. This mild effect on the IIS pathway reduces fitness cost of plant viruses in their insect vectors considering that the IIS pathway plays vital roles in insect basic physiology such as reproduction, growth rate, and lifespan (33, 34). From the view of saving energy for viruses, it is reasonable to infer that RSV exploits the existing Encounter pathway, which could function in insect wing dimorphism induced by other biological or non-biological factors, such as population density, host plant quality, temperature, or photoperiod. Apart from RSV, other planthopper-borne rice viruses may manipulate Encounter to induce long-winged planthoppers as a consequence of viral convergent evolution.

Fig. 8.

Model of the molecular regulation of SBPH wing dimorphism by RSV. The regulation is sex-biased and mediated by a planthopper-specific Encounter gene, which is closely connected with the insulin/insulin-like growth factor signaling pathway. The virus activates the transcription of Akt and its downstream factor Encounter (red arrowhead). The transcriptional activation and protein phosphorylation are paralleled regulatory modes for Akt to control wing dimorphism in SBPH. An RSV-derived small interfering RNA (vsiRNA) directly targets Encounter to enhance its expression. Activated components are shown in red and inhibited components are shown in black. FoxO-P, phosphorylated FoxO. Akt-P, phosphorylated Akt. The SBPH images were illustrated by Miss Pan-Pan Gu.

RSV induces the development of the long-winged morph only in male SBPH, not in females. Clearly, long-winged insects are thought to be more effective vectors than short-winged insects due to their ability to migrate frequently and fly longer distances (35). Insects with long wings expend significant amounts of energy during flight as they search for food, evade predators, and disperse. For female insects, flight may deplete the resources needed for egg production, resulting in a compromise between flight capability and fecundity (36–38). That is, the acquisition of long wings in a female allows it to fly but at the cost of reduced reproductive capacity. For example, the long-winged planthopper Prokelisia dolus, which possesses greater flight capability, displayed lower levels of fecundity than their short-winged female counterparts (39). Therefore, males seem to be more suitable for long-distance migration than females. In field populations of SBPH, males may be more prone to develop long-winged morphs than females. In a study on SBPHs collected from 16 sites throughout South Korea, the percentage of short-winged males in July significantly declined compared to that in April, while that of short-winged females did not change (40). Recently, a large genomic insertion containing a duplicated follistatin gene on the X chromosome is reported to be linked to the pea aphid male wing dimorphism (41). RSV is transmitted transovarially by viruliferous females to their progeny but is not paternally transmitted (42, 43), while the ability of viral horizontal transmission from insects to plants seems comparable for females and males (44). It is likely that RSV intensifies the characteristics of male SBPHs prone to developing long-winged morphs for long-distance spread and exploits the high fecundity of short-winged females for intrafield spread.

Encounter-regulated wing plasticity is planthopper-specific and function-specific. Encounter exclusively exists in planthopper species, at least in three planthopper species after analyzing their genomes (8, 20, 21). It is not present in other insect species with known genome information. Two mysteries shroud Encounter. One is that Encounter is overwhelmingly expressed in the male reproductive tissue of SBPH. The known genes involved in the wing differentiation of planthoppers are either mainly expressed in larval wing buds, such as InR2, Vestigial and Wingless, or expressed without obvious tissue specificity, such as InR1, JNK, Zfh1, and juvenile hormone epoxide hydrolase, and all are not sexually distinct (10–15). How such a gene as Encounter almost exclusively expressed in the reproductive tissue affects wing plasticity deserves deeper exploration. The other is that Encounter is similar to mitochondrial malate dehydrogenase in sequence and localized in mitochondria but without malate dehydrogenase activity. It seems that Encounter is an outcome of gene duplication in the malate dehydrogenase family and deviates from this family to obtain new functions during planthopper evolution, probably shaping the evolution of wing plasticity across species.

We found that vsiR-7880 upregulated Encounter expression in a miRNA manner by binding to its 5′ UTR. Typically, miRNAs target the 3′ UTRs of mRNAs, leading to mRNA degradation or translational repression (45). Some miRNAs attach to the 5′ UTRs of specific mRNAs, making the mRNAs more stable or promoting the activation of transcription or translation processes (46). In contrast to the 3′UTR rich in AU sequences, the 5′UTR is usually enriched in GC sequences and is predicted to have a greater level of secondary structures for controlling gene expression (47). For example, miR-10a directly binds downstream of the 5′ terminal oligopyrimidine motif within the 5′UTR of mouse ribosomal protein mRNA to promote the translation of ribosomal protein (28). miR-483-5p binds the 5′UTR of fetal IGF2 mRNA, leading to an increase in RNA helicase DHX9 binding to the IGF2 transcript to promote IGF2 transcription (48). miR-1254 interacts with the structured elements in the 5′UTR of cell cycle and apoptosis regulator 1 (CCAR1) mRNA to enhance the stability of both molecules (49). The positive regulation of vsiR-7880 on the transcription of Encounter may be achieved by modifying the secondary structures of the 5′ UTR of Encounter or maintaining the stability of transcripts. Further studies are required to clarify the mechanisms.

The RSV-SBPH system is a unique case for the direct regulation of plant viruses to wing dimorphism of insect vectors. Most cases reported previously involve indirect regulation through host plants. Corn planthoppers were much more likely to mature as long-winged morphs when reared on maize mosaic virus (MMV)-infected old leaves than on healthy old leaves, while this difference was not obvious when they fed on healthy and MMV-infected young leaves (25). How plant developmental status affects the regulation of MMV to planthopper long-winged morphs is not clear. A mechanism-well-explored case was described in the system of cucumber mosaic virus (CMV), a nonpersistent virus, and its insect vector green peach aphid (Myzus persicae) (50). The aphids fed on yellow tobacco leaves harboring CMV and a satellite RNA called Y-sat are prone to develop wings. Wing induction is mediated by a Y-sat-derived piwi-RNA (piRNA), which targets an aphid microRNA (miRNA) called miR-9b (50). Aphid miR-9b regulates the expression of the ABC transporter ABCG4, which influences wing formation through insulin signaling (50). The presence of Y-sat-derived piRNA upregulates ABCG4 expression, leading to wing formation. As a nonpersistent virus, CMV as well as its satellite RNA Y-sat is localized only on the mouthpart of the aphid vector. Therefore, the Y-sat-derived piRNA is supposed to originate from virus-infected plants, not produced inside aphid cells, representing a kind of indirect regulation. A group of insect-specific densoviruses has been directly linked to wing formation in aphids. For example, the rosy apple aphid infected with Dysaphis plantaginea densovirus (DplDNV) produces the winged morph even at low population density (51). Interestingly, two pea aphid genes, Apns1 and Apns2, are involved in regulating wing dimorphism and supposed to be derived from the lateral transfer from DplDNV-like viral genes (52). Unfortunately, the underlying molecular mechanisms of wing induction by densoviruses are not yet known. RSV is a typical persistent-propagative virus, and vsiRNAs are synthesized with RSV proliferation inside vector cells (31). The influence of RSV on the formation of long-winged morphs in male SBPHs through RSV-derived vsiR-7880 enhancing the expression of Encounter is a unique case of direct interaction between plant viruses and wing plasticity of insect vectors. These findings not only show the diverse influence of plant viruses on the physiology and behavior of insect vectors but also reveal a molecular mechanism for the regulation of insect wing dimorphism.

Materials and Methods

Insect Rearing.

Nonviruliferous and viruliferous small brown planthopper populations were collected from the field population in Hai’an, Jiangsu Province, China, in 2008 and maintained in the laboratory for more than 10 y. The insects were reared on fresh rice seedlings, Oryza sativa Wuyujing, in glass beakers at 24 °C with 16 h of light daily. The viruliferous populations were screened every 3 mo via dot-ELISA with the homemade monoclonal anti-NP antibody as described previously to ensure that the RSV-carrying frequency was maintained above 90% (53).

Morphological Feature Identification.

Morphological feature observation of the wings of SBPH was performed using a large depth of field 3D digital microscope (Keyence). Morphological features of SBPH reproductive organs were captured with a Leica M205C microscope (Leica Microsystems). Adobe Photoshop CS5 was used for image processing.

Wing Morph Statistics.

Sixteen nonviruliferous or viruliferous third-instar SBPH nymphs, 20 third-instar viruliferous nymphs after injection with dsRNAs of Psbp, NP, GFP, GFP1, GFP2, GFP1-1, GFP1-2, Encounter (Contig2490.1), Akt, InR1, or 16 offspring from nonviruliferous third-instar nymphs after injection with RSV crude extracts were raised on six rice seedlings in a 410 cm³ glass incubator. All insects for wing morph statistics were counted at the adult stage. When the number of males reached 12 but less than 24, it was counted as a biological replicate. The proportion of long-winged males or females was reported as the mean ± SE.

RNA Extraction and cDNA Synthesis.

Total RNA was isolated from the whole bodies of viruliferous and nonviruliferous SBPHs, including nymphs at each age stage, male and female adults with short/long wings, or third-instar nymphs after injection with dsRNAs for 5 d, following the standard TRIzol reagent protocol (Invitrogen). In addition, RNA was also extracted from the head, gut, testis, fat body, wing, and carcass from nonviruliferous SBPH males and viruliferous third instars. The concentration and quality of total RNA were determined using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific). One microgram of RNA was reverse transcribed to generate cDNA using the M-MLV reverse transcription system (Promega) and random primers (Promega) following the manufacturer’s instructions. For vsiRNA evaluation, RNA was extracted from viruliferous and nonviruliferous fourth-instar SBPH nymphs or from fourth-instar nymphs after injection with vsiRNA activators for 3 d. Two micrograms of RNA were reverse transcribed using the miRcute Plus miRNA First-Strand cDNA Kit (Tiangen).

qPCR.

qPCR was performed to measure the relative RNA levels of RSV NP (GenBank accession no. DQ299151) and the relative transcript levels of SBPH genes according to the genome information of SBPH (20) on a LightCycler 480 II system (Roche) by using 2×SYBR Green PCR Master Mix (Fermentas). Due to having been widely used as a reference gene in nonviruliferous and viruliferous SBPH (30, 54, 55), translation elongation factor 2 (EF2) (Contig0.299) was quantified as an endogenous control for mRNAs and viral RNAs. Relative transcript levels of SBPH genes or RNA levels of RSV NP were reported as the mean ± SE. In addition, the relative RNA levels of vsiRNAs were determined with the miRcute miRNA qPCR Detection Kit (Tiangen). Planthopper U6 snRNA was used as an endogenous control for vsiRNAs. Six to eight replicates were prepared. In the geographical population experiment, more than 11 replicates were established. The primers used in this experiment are shown in SI Appendix, Table S1. All PCR products were sequenced for validation.

Injection of RSV Crude Extractions.

Fifty viruliferous SBPHs were ground in 100 µL of H₂O. After centrifugation at 13,000 rpm for 15 min at 4 °C five times, the supernatant containing RSV was delivered into the hemolymph of nonviruliferous third-instar nymphs by microinjection through a glass needle using a Nanoliter 2000 system (World Precision Instruments). The control group was injected with H₂O. Insects were raised on healthy rice plants, and their offspring were collected to measure the relative RNA level of RSV NP and the proportion of LWM in next-generation SBPH adults.

Geographic Population Collection of SBPH.

Field populations of SBPH were collected from nine different regions of China, including the single-cropping northeast (Shenyang), single- and double-cropping central area (Lianyungang, Yancheng, Xuzhou, Taizhou, Nantong, and Kaifeng), and single- and double-cropping southwest plateau (Kunming and Chuxiong) of China between June and August. The SBPHs were preserved in Ambion™ RNAlater™ (Thermo Fisher Scientific) before RNA extraction. Wing morph and the relative RNA level of RSV NP were further analyzed.

dsRNA Injection.

dsRNAs for Psbp, NP, GFP, GFP1, GFP2, GFP1-1, GFP1-2, Encounter, Akt, InR1, and FoxO were synthesized using the T7 RiboMAX™ Express RNAi System (Promega) following the manufacturer’s protocol. Primers are shown in SI Appendix, Table S1. A total of 18.4 nL of dsRNAs at 12 mg/mL was delivered into the hemolymph of second-, third-, or fourth-instar SBPH nymphs using a Nanoliter 2000 injector (World Precision Instruments). Planthoppers were collected for qPCR after injection with dsRNAs for 5 d. The residual insects were matured as adults to count the proportion of long-winged morphs.

Injection of vsiRNA Activator and Inhibitor.

The activators of vsiR-2675 and vsiR-7880 were chemically modified double-stranded oligonucleotides corresponding to the sequence of each vsiRNA (GenePharma, Shanghai, China). The negative control (NC) sequences for the vsiRNA activator were 5′-UUCUCCGAACGUGUCACGUTT-3′ (sense) and 5′-ACGUGACACGUUCGGAGAATT-3′ (antisense) (GenePharma). The inhibitor of vsiR-7880 was chemically modified single-strand nucleotide sequences with reverse complementarity to vsiR-7880 (GenePharma). A random sequence of 5′-CAGUACUUUUGUGUAGUACAA-3′ was synthesized as the NC sequence for the vsiRNA inhibitor.

A total of 18.4 nL of activators of vsiR-2675, vsiR-7880, or NC at 200 µM was delivered into nonviruliferous fourth-instar SBPH nymphs by microinjection using a Nanoliter 2000 system (World Precision Instruments). In addition, 18.4 nL of inhibitor of vsiR-7880 or NC at 200 µM was delivered into viruliferous fourth-instar SBPH nymphs. Male nymphs were identified and collected at 3 d after injection for qPCR analysis.

RNA Immunoprecipitation Combined with qPCR (RIP-qPCR).

RIP analysis was performed using a RIP-Kit (BersinBio) as described previously (56). In short, homemade anti-Ago1 monoclonal antibodies (32) and normal mouse IgG (Abcam) were incubated with homogenates from nonviruliferous fourth-instar SBPH nymphs after injection with 200 µM vsiR-7880 activator for 3 d for immunoprecipitation. Enriched RNAs were extracted using TRIzol reagent (Invitrogen) and reverse transcribed into cDNA by using the M-MLV reverse transcription system (Promega) or the miRcute Plus miRNA First-Strand cDNA Kit (Tiangen). qPCR was then performed to detect the transcript level of the 5′UTR of Encounter or the RNA level of vsiR-7880. The RNA level of each target RNA relative to that in the IgG control sample is reported as the mean ± SE.

Dual-Luciferase Reporter Assay in Drosophila S2 Cells.

Validation of the target of vsiRNA was performed in Drosophila S2 cells by a dual-luciferase reporter assay as described previously (27). The putative target of vsiR-7880 (220 bp sequence from the 5′UTR of Encounter) was cloned and inserted into psiCHECK2 (Promega). S2 cells from each well of a 24-well plate were cotransfected with 1 µg of recombinant psiCHECK2 plasmids and 20 pM or 100 pM vsiR-7880 mimic or NC using Lipofectamine 3000 (Invitrogen). The mimics were double-stranded RNAs of vsiR-7880. The sequences of NC for the vsiR-7880 mimic were 5′-UUCUCCGAACGUGUCACGUTT-3′ (sense) and 5′-ACGUGACACGUUCGGAGAATT-3′ (antisense). Site mutations in the sequence complementary to the seed region of vsiR-7880 were generated using a KOD-Plus mutagenesis kit (Toyobo), and the concentration of vsiR-7880 mimic was set as 100 pM. Relative luciferase activities of cells after transfection at 28 °C for 24 h were evaluated via the Dual-Luciferase Reporter Assay System (Promega). Six replicates were prepared for each group. The relative activity of Renilla luciferase normalized to firefly luciferase activity is presented as the mean ± SE. Primers are listed in SI Appendix, Table S1.

Phylogenetic Analysis.

Homologous proteins of Encounter obtained from Tribolium castaneum, Thrips palmi, Trichoplusia ni, Drosophila melanogaster, Apis mellifera, M. persicae, A. pisum, and N. lugens were downloaded from the National Center for Biotechnology Information. Encounter of L. striatellus (20) and S. furcifera (21) was downloaded from the GigaScience repository, GigaDB. A maximum likelihood phylogenetic tree was created using Mega 7 software, adopting the partial deletion and JTT model. To assess the internal support of the tree topology, a bootstrap analysis was carried out with 1,000 replicates.

Immunohistochemistry and Recombinant Protein Expression in Human 293T Cells.

Based on the transcriptome, full-length open reading frames (ORF) of Encounter with 1,071 bp were amplified and cloned in the pcDNA3.1(+) (Invitrogen) between the restriction sites Hind III-Xho I to generate Encounter-V5 plasmid. The recombinant plasmids (1 µg/well) and 30 µL of cell light mitochondria BacMam 2.0 (Invitrogen) were transfected into 500 µL of 293T cells in each well of a 24-well plate using Lipofectamine 3000 (Invitrogen). After transfection for 48 h, cells were washed two times with 1× PBS buffer (pH 7.4), fixed with 4% paraformaldehyde for 1 h at room temperature (RT), and then washed three times with 1× PBST buffer (pH 7.4). After being permeabilized at RT for an hour using Cell Penetrating Solution (Beyotime), cells were blocked at RT using 5% bovine serum albumin for an hour, incubated with the primary antibody, anti-V5 monoclonal antibody (CWBiotech), for an hour at RT and subsequently incubated overnight at 4 °C. After washing with 1× PBST buffer (pH 7.4), the secondary antibody Alexa Fluor 488 (green) AffiniPure goat anti-mouse IgG (Invitrogen) was added. The nuclei were stained with Hoechst (blue) (Invitrogen). Negative controls were cells transfected with the empty plasmid pcDNA3.1(+). Fluorescence was visualized using a Carl Zeiss AG Zeiss LSM 710 confocal microscope.

Expression, Purification, and MDH Activity Assay of the Recombinant Proteins in Escherichia coli.

Full-length ORF of Encounter and mitochondrial malate dehydrogenase (Contig66.45) were cloned in pET28a (Invitrogen) between the restriction sites BamH I-Xho I. The recombinant plasmids were introduced into E. coli BL21 strains to enable protein expression. Cells were induced with 0.5 mmol/L isopropyl-β-D-thiogalactoside (IPTG) at 16 °C for 16 h, pelleted through centrifugation, and then sonicated on ice for 30 min. The expressed recombinant proteins were subsequently purified using Ni Sepharose (GE Healthcare) and diluted to 0.2 mg/mL with 1× PBS buffer before use. To determine the MDH activity of Encounter, the NAD-MDH activity assay kit (Solarbio) was employed, as per the manufacturer’s instructions. BSA was used as the negative control group, and mit MDH was used as the positive control group. Each group’s relative NAD-MDH activity values were calculated as the means ± SE.

Bioinformatic Analyses of vsiRNA Binding Sites within the 5’UTR of Encounter.

The potential binding sites of RSV-derived vsiRNAs in the 5′UTR of Encounter were predicted using RNAhybrid (57) and miRanda (58) with cutoff values of −20 and −18 kcal mol⁻¹ for the minimum free energy (MFE) of the RNA duplex.

Statistical Analysis.

All data were analyzed using IBM SPSS statistics 19. Data comparison between two groups was performed using Student’s t test. Data comparisons among multiple groups were performed using one-way ANOVA followed by Tukey’s or Duncan’s test.

Supplementary Material

Appendix 01 (PDF)

Data, Materials, and Software Availability

All data needed to evaluate the conclusions in the paper are present in the paper and/or SI Appendix.