DDX6 Is Essential for Oocyte Development and Maturation in Locusta migratoria

Simple Summary Insect reproduction is an important and complicated process required for producing healthy individuals and maintaining their population abundance. Thus, it could become a valuable target for insect biological control. To date, many factors and pathways have been revealed to be involved in this reproductive process, but it is still far from a full understanding of the molecular network underlying this process. We herein investigated a RNA helicase, DEAD-box protein 6 (DDX6) in Locusta migratoria, a global, destructive pest, and found that knockdown LmDDX6 downregulated expression levels of juvenile hormone receptor gene methoprene-tolerant and its target genes 78-kDa glucose-regulated proteins, thus reducing vitellogenin expression and ultimately impairing the ovary development and oocyte maturation. These results demonstrate that LmDDX6 is a key player in female locust reproduction, providing, thus, a novel target for locust biological control. Abstract DEAD-box protein 6 (DDX6) is a member of the DDX RNA helicase family that exists in all eukaryotes. It has been extensively studied in yeast and mammals and has been shown to be involved in messenger ribonucleoprotein assembly, mRNA storage, and decay, as well as in miRNA-mediated gene silencing. DDX6 participates in many developmental processes but the biological function of DDX6 in insects has not yet been adequately addressed. Herein, we characterized the LmDDX6 gene that encodes the LmDDX6 protein in Locusta migratoria, a global, destructive pest. LmDDX6 possesses five motifs unique to the DDX6 subfamily. In the phylogenetic tree, LmDDX6 was closely related to its orthologs in Apis dorsata and Zootermopsis nevadensis. RT-qPCR data revealed high expression of LmDDX6 in the ovary, muscle, and fat body, with a declining trend in the ovary after adult ecdysis. LmDDX6 knockdown downregulated the expression levels of the juvenile hormone receptor Met, and genes encoding Met downstream targeted Grp78-1 and Grp78-2, reduced LmVg expression, and impaired ovary development and oocyte maturation. These results demonstrate that LmDDX6 plays an essential role in locust female reproduction and, thus, could be a novel target for locust biological control.


Introduction
Reproduction is an essential and complicated process required for producing healthy individuals and maintaining the population abundance of organisms. Insect female reproductive system is composed of two ovaries, containing a number of ovarioles, connected directly to the oviduct. The ovariole is the functional unit for egg production, and the number of ovarioles in each ovary varies widely, depending on the particular insect species [1,2]. In brief, the ovariole consists of the apical terminal filament, the germarium region linked to the terminal filament, and the vitellarium in the basal part [1]. Insects of different orders adopt distinct reproductive strategies. Thus far, three types of oogenesis have been described based on the presence and position of the nurse cells: (1) panoistic, (2) telotrophic meroistic, and (3) polytrophic meroistic [3,4]. Proper ovary development and oocyte maturation in insects are prerequisites for successful reproduction. Many intrinsic and extrinsic factors are involved in this process, including hormones, nutrition, and growth conditions [5][6][7][8].
Juvenile hormone (JH), one of the classical endocrine hormones produced by the corpora allata, a pair of endocrine glands in the retrocerebral complex behind the brain, has been long known to regulate female reproduction in many insects, exerting a central role in vitellogenesis, an indispensable process of ovary development and oocyte maturation [8]. As a gonadotropin, JH promotes female reproduction mainly by inducing the expression of Vg [8][9][10]. Vg is a key gene expressed largely in the fat body during vitellogenesis that encodes vitellogenin, a major precursor of the yolk protein, which is secreted into the hemolymph and taken up by maturing oocytes [11]. Besides the induction of Vg expression, JH enhances Vg uptake and promotes oocyte maturation [12]. In Locusta migratoria, several downstream genes of JH and its receptor complex have been found to regulate fat body polyploidization, including mini-chromosome maintenance 4/7 (Mcm4/7), cell division cycle 6 (cdc6), cyclin-dependent kinase 6 (Cdk6), and adenovirus E2 factor 1 (E2f1), reducing LmVg expression level and ultimately impairing oocyte growth and maturation [13][14][15]. Very recently, it has been found that JH promotes the expression of Cdc2 and origin recognition complex subunit 5 (Orc5) via the LCMT1-PP2A-FoxO pathway, mediating fat body ploidy, and reduces the expression of LmVg [16]. In Tribolium castaneum, JH induces the expression level of the gene encoding insulin-like peptides in the fat body, phosphorylating fork head transcription factor FOXO, and promoting Vg expression [10]. In the American cockroach, it has been recently reported that the insulin/IGF signaling and targeting of rapamycin induce the expression of Jhamt and Cyp15a, the enzymes of the last two steps of JH biosynthesis, thus activating JH biosynthesis, eventually affecting vitellogenesis and oocyte maturation [17]. In D. melanogaster, JH enhances Vg uptake and promotes oocyte maturation [18].
Ecdysteroid is another classical hormone that controls oocyte development and maturation in insects [19]. The active form of ecdysteroid, 20-hydroxyecdysone, has been reported to contribute to Drosophila oogenesis, where it tightly controls the developmental checkpoint at stage 8, which allows the onset of vitellogenesis and egg maturation. Females with mutation in ecdysone receptor contained abnormal egg chambers [20]. In mosquitoes, inhibition of the target of rapamycin protein synthesis impeded Vg expression and reduced fecundity, suggesting that the nutritional signaling pathway contributes to mosquito ovary development and oocyte maturation [6]. In addition, miRNAs have been shown to be involved in female reproduction [21]. These data indicate that although there has been some progress in the elucidation of the molecular mechanism of ovary development and oocyte maturation in insects, the current body of knowledge is just the tip of the iceberg. To fully understand the molecular network underlying complicated reproductive processes in insects, more factors and pathways need to be taken into account.
DEAD-box proteins (DDXs) comprise a large family of RNA helicases that are conserved from bacteria to eukaryotes. They share nine common, conserved motifs in the helicase core [22]. Motif II that contains four amino acids (Asp-Glu-Ala-Asp, i.e., DEAD) is required for ATPase activity. Other motifs are involved in ATP binding and hydrolysis as well as in RNA binding [23]. Multiple studies have demonstrated that DDXs are involved in almost every aspect of RNA metabolism, from transcription, splicing, transport, ribosome biosynthesis, and translation to RNA decay, so they have multifaceted biological functions in cells. For example, in yeast, 15 of 25 DDXs have been demonstrated to regulate ribosome biogenesis [24]. DDX20/DP103 in mammalian cells has been shown to repress transcription [25,26]. Ded1/DDX3 is required for translation initiation [23,27], and Belle, the DDX3 ortholog in Drosophila, is required for male and female fertility [28,29].  [30]. Vasa/DDX4 is a germ cell marker required for fertility [31]. Many of these DDXs have been also shown to participate in tumorigenesis, antiviral reactions, and immune responses [32][33][34].
In our previous study, we isolated 32 DDX genes from L. migratoria and identified seven of these LmDDXs that were indispensable for nymph survival [35]. However, their detailed biological functions remained largely unknown. DDX6, an important translational repressor [30], has not yet been extensively studied in insects. Herein, we focused on the LmDDX6 gene and characterized its function in oocyte development and maturation. Knockdown of LmDDX6 reduced LmVg expression and downregulated expression levels of the methoprene-tolerant gene (Met) encoding the JH receptor and its downstream target gene Grp78, which ultimately resulted in oocyte abortion. These results confirmed that LmDDX6 is a key player in locust female reproduction and, as such, it may be a new target for locust biological control.

Experimental Insects
Nymph locusts were purchased from a locust breeding center (Cangzhou, China) and reared in a cage with 50% relative humidity at 30 ± 2 • C under the 14 h:10 h (light:dark) photoperiod. Fresh wheat leaves and bran were fed to the locusts twice per day. Adult locusts after eclosion were used for the following experiments.

Motif Pattern Analysis
Motif pattern analysis was conducted by using online program MEME (http://memesuite.org/tools/meme). The ortholog sequences of DDX6 used in this analysis are listed in Supplemental File S1. The parameters were as follows: minimum width = 10, maximum width = 10, and maximum number of motifs to find = 18.

Phylogenetic Analysis of DDX6
DDX6 protein sequences from different species were obtained from NCBI (National Center for Biotechnology Information) and a multiple-sequence alignment was performed by using ClustalW software. The phylogenetic tree was generated by MEGA 6 by using the neighbor-joining method with 1000 repetitions. The protein accession numbers are shown in Table 1.

RNA Extraction and RT-qPCR
Integument, fat body, ovary, foregut, midgut, hindgut, malpighian tubule, and muscle tissues of female adults 2 days post-adult eclosion (PAE) were first sampled. Ovaries from 0, 2, 4, 6, and 8 days PAE locusts were also collected. Three individuals were sampled for one replicate, and three replicates were repeated. Total RNA was extracted by using RNAiso Plus reagent (Takara, Japan). First-strand cDNA was synthesized with 1 µg of total RNA by using an RNA HiScript ® III RT SuperMix for qPCR (+ gDNA wiper) Kit (Vazyme, Nanjing, China) according to the manufacturer s instructions. Real-time quantitative PCR (RT-qPCR) was performed to measure the relative transcript level by using a LightCycler ® 480 Instrument II (Roche, Basel, Switzerland) with 2 × ChamQ TM Universal SYBR ® qPCR MasterMix. The RT-qPCR program was conducted at 94 • C for 2 min, followed by 40 cycles of 94 • C for 15 s and 60 • C for 31 s. The specific primer sequences are summarized in Table  S1. Relative gene expression was calculated by the 2 −∆∆CT method. The level of β-actin mRNA expression was used as internal control.

RNA Interference (RNAi)
The synthesis of the LmDDX6 double-stranded RNA (dsLmDDX6) was described previously [35]. In brief, the region (484 bp) for dsLmDDX6 from LmDDX6 gene was amplified by PCR using the specific dsRNA primers (Table S1), which contain the T7 RNA polymerase promoter sequence. DsLmDDX6 was synthesized by using T7 RiboMAX ™ Express RNAi System (Promega, USA) and dissolved in nuclease-free water. The dsGFP was synthesized in parallel and served as mock control. Female adult locusts within 12 h after eclosion (0 PAE) were injected with 10 µg of dsRNA at the second to third segments of the abdomen. The silencing efficiency of LmDDX6 in the ovary and fat body from female locusts at 4, 6, and 8 days PAE was analyzed by RT-qPCR.

Tissue Imaging
Epson Perfection V600 Photo was used for imaging ovary morphology. The morphology and length of the ovarioles were analyzed by using a Leica M205C microscope.

Data Analysis
The relative expression level of LmDDX6 in various tissues was calculated using oneway analysis of variance (ANOVA), as appropriate, by using SPSS 16.0 software. The post hoc Tukey's test was used if F value in one-way ANOVA reported a significant effect. The different letters indicate a significant difference. The comparison of the gene expression and the size of primary oocytes between the dsLmDDDX6-and the dsGFP-treated locusts were analyzed by the two-sample and two-tail t-test. All statistical analyses were conducted at the significance level of α = 0.05 (p < 0.05).

Motif Patterns of LmDDX6 and Its Orthologs
LmDDX6 encodes a protein of 449 amino acids (aa) that form the conserved DEXDc and HELICc domains with the N-and C-terminal regions of 78 aa and 58 aa, respectively ( [35]; Table 1). Both the N-and C-terminal sequences are more variable than the conserved domains among the members of the DDX family or even of the same DDX subfamily. To search for some motifs that might be present only in the DDX6 subfamily, we first selected 13 sequences from different phyla, including yeast (Saccharomyces cerevisiae, DHH1), cnidarian (Hydra vulgaris, HvDDX6), worm (Caenorhabditis elegans, Cgh-1), insects (L. migratoria, LmDDX6; D. melanogaster, Me31B; Zootermopsis nevadensis, ZnDDX6), vertebrates (Homo sapiens, HsDDX6; Mus musculus, MmDDX6; Danio rerio, DrDDX6), and green plants (Chlamydomonas reinhardtii, CrDDX6; Arabidopsis thaliana, AtRH8), as well as sequences of two outgroup members, HsDDX39A from H. sapiens and DBP2 from S. cerevisiae. These sequences were analyzed as one group using MEME program (www.meme-suite.org). We found six motifs (motifs 14, 13, 12, 7, 1, and 17) in the DDX6 subfamily but not in Hs-DDX39A or DBP2 among the 18 motifs examined (Figure 1a). We then explored more sequences (48 as one group) and confirmed five out of the six motifs (motif 17 was removed this time, Figure S1). To obtain further confirmation, 27 sequences from algae, 191 sequences from insects, and 500 sequences from vertebrates, plants, and fungi, respectively, were downloaded and examined manually for the five motifs, one by one (Supplemental Files S2-S6). Without exception, all five motifs were maintained after this examination. Therefore, we discovered five new motifs highly conserved in the DDX6 subfamily, including two motifs (13,KRELLMGIFE and 14,TKGNEFEDYC) in the N-terminal region, one motif (12, PYEINLMEEL) in the DEXDc domain, one motif (1, YSCYYIHAKM) in the HELICc domain, and one motif (7,KVHCLNTLFS) in the linker region between the DEXDc and HELICc domains (Figure 1b,c). The specificities and biological roles of these motifs are currently unknown.

Phylogeny of LmDDX6 and Its Orthologs
To understand the relationship between LmDDX6 and other DDX6 subfamily members, we generated a phylogenetic tree ( Figure 2) by using DDX6 sequences from 32 species selected from diverse phyla (Table 1). DDX6 sequences from green plants, vertebrates, and insects were classified into distinct clades. LmDDX6 in the insect clade was closely related to the orthologs from Apis dorsata and Z. nevadensis. Cgh-1 from worm and a group of DDX6 orthologs from Mollusca (Crassostrea gigas), Annelida (Capitella teleta), and Rotifera (Brachionus plicatilis) were more closely related to the insect clades. The flatworm (Dugesia japonica) and porifera (Amphimedon queenslandica) sequences were closely related to the vertebrate clade. DHH1, a yeast DDX6 ortholog, was strikingly related to the sequences from green plants. These data demonstrate that DDX6 appears in unicellular eukaryotes, such as C. reinhardtii and S. cerevisiae, and is retained by multicellular eukaryotes.  of DDX6 orthologs from Mollusca (Crassostrea gigas), Annelida (Capitella teleta), and Rotif-era (Brachionus plicatilis) were more closely related to the insect clades. The flatworm (Dugesia japonica) and porifera (Amphimedon queenslandica) sequences were closely related to the vertebrate clade. DHH1, a yeast DDX6 ortholog, was strikingly related to the sequences from green plants. These data demonstrate that DDX6 appears in unicellular eukaryotes, such as C. reinhardtii and S. cerevisiae, and is retained by multicellular eukaryotes.

Figure 2.
A phylogenetic tree of DDX6 orthologs. DDX6 protein sequences from different species were obtained from NCBI (National Center for Biotechnology Information), and a multiple-sequence alignment was performed using Clus-talW software. The phylogenetic tree was generated by MEGA 6 using neighbor-joining method with 1000 repetitions. The filled, red circle indicates DDX6 from L. migratoria. Protein accession numbers are shown in Table 1.
Intriguingly, when we searched for LmDDX6-like sequences in the BLAST bacterial database in NCBI, we found sequence WP_162815294.1 from Microbacterium arborescen, which showed an overall 90% identity with LmDDX6 and also contained the five motifs Figure 2. A phylogenetic tree of DDX6 orthologs. DDX6 protein sequences from different species were obtained from NCBI (National Center for Biotechnology Information), and a multiple-sequence alignment was performed using ClustalW software. The phylogenetic tree was generated by MEGA 6 using neighbor-joining method with 1000 repetitions. The filled, red circle indicates DDX6 from L. migratoria. Protein accession numbers are shown in Table 1.
Intriguingly, when we searched for LmDDX6-like sequences in the BLAST bacterial database in NCBI, we found sequence WP_162815294.1 from Microbacterium arborescen, which showed an overall 90% identity with LmDDX6 and also contained the five motifs unique to the DDX6 subfamily. Interestingly, when a BLAST search of WP_162815294.1 was performed in NCBI, we found another sequence, from Papilio polytes, with 100% identity with WP_162815294.1. That sequence was a DDX6 ortholog from P. polytes and in the reconstructed phylogenetic tree ( Figure S2) it was closely related to the sequence of BmDDX6, a DDX6 ortholog in Bombyx mori. Based on these data, we assume that sequence WP_162815294.1 from M. arborescen was probably a contamination from the sequence of P. polytes.

Expression Profile of LmDDX6 in Female Adults
In our previous study, we detected high expression of LmDDX6 in the testis and ovary and intermediate expression in the fat body and Malpighian tubules of five-instar nymphs [35]. To determine LmDDX6 expression in adult females, we collected various tissues from female locusts at 2 days PAE and analyzed them by RT-qPCR. A high expression level of LmDDX6 was detected in the ovary and an intermediate level of expression was detected in the fat body and muscle. Other tissues, including integument, foregut, midgut, hindgut, and Malpighian tubules showed lower LmDDX6 expression levels (Figure 3a). To determine the relationship between the expression of LmDDX6 and ovary development, we sampled the ovaries on different days PAE and conducted RT-qPCR. We observed a high level of LmDDX6 expression in the first two days PAE, with a declining trend afterwards (Figure 3b). These expression data suggest that LmDDX6 plays a role in ovary development and oocyte maturation. P. polytes.

Expression Profile of LmDDX6 in Female Adults
In our previous study, we detected high expression of LmDDX6 in the testis and ovary and intermediate expression in the fat body and Malpighian tubules of five-instar nymphs [35]. To determine LmDDX6 expression in adult females, we collected various tissues from female locusts at 2 days PAE and analyzed them by RT-qPCR. A high expression level of LmDDX6 was detected in the ovary and an intermediate level of expression was detected in the fat body and muscle. Other tissues, including integument, foregut, midgut, hindgut, and Malpighian tubules showed lower LmDDX6 expression levels (Figure 3a). To determine the relationship between the expression of LmDDX6 and ovary development, we sampled the ovaries on different days PAE and conducted RT-qPCR. We observed a high level of LmDDX6 expression in the first two days PAE, with a declining trend afterwards (Figure 3b). These expression data suggest that LmDDX6 plays a role in ovary development and oocyte maturation.

Knockdown of LmDDX6 Leads to Oocyte Abortion
To elucidate the function of LmDDX6, we injected dsLmDDX6 and dsGFP into female locusts within several hours PAE. Then, we dissected the injected locusts on different days PAE and observed the ovaries carefully. An obvious difference between dsLmDDX6treated and control dsGFP-treated locusts first appeared at 4 days PAE. The size of the ovary was slightly smaller in the dsLmDDX6-injected locusts than in the control, dsGFPinjected, locusts. Strikingly, primary oocytes were smaller in dsLmDDX6-treated locusts (Figure 4a,b). Furthermore, this difference dramatically increased at 6 and 8 days PAE. The sizes of the ovary and primary oocytes in dsGFP-treated locusts increased greatly, with the latter reaching the maturation size of ~7.6 mm 2 at 8 days PAE under our experimental conditions. In contrast, in the dsLmDDX6-treated locusts, the ovary and primary oocytes did not change much from 4 to 8 days PAE, with only a slight increase from 0.5 to 0.9 mm 2 for the primary oocyte size (Figure 4a,b).

Knockdown of LmDDX6 Leads to Oocyte Abortion
To elucidate the function of LmDDX6, we injected dsLmDDX6 and dsGFP into female locusts within several hours PAE. Then, we dissected the injected locusts on different days PAE and observed the ovaries carefully. An obvious difference between dsLmDDX6treated and control dsGFP-treated locusts first appeared at 4 days PAE. The size of the ovary was slightly smaller in the dsLmDDX6-injected locusts than in the control, dsGFPinjected, locusts. Strikingly, primary oocytes were smaller in dsLmDDX6-treated locusts (Figure 4a,b). Furthermore, this difference dramatically increased at 6 and 8 days PAE. The sizes of the ovary and primary oocytes in dsGFP-treated locusts increased greatly, with the latter reaching the maturation size of~7.6 mm 2 at 8 days PAE under our experimental conditions. In contrast, in the dsLmDDX6-treated locusts, the ovary and primary oocytes did not change much from 4 to 8 days PAE, with only a slight increase from 0.5 to 0.9 mm 2 for the primary oocyte size (Figure 4a,b).

Downregulation of Vg Expression by LmDDX6 Knockdown
Vitellogenin is synthesized in the fat body and is essential for oocyte development. The phenotype of the ovaries and oocytes in the dsLmDDX6-treated locusts might be ascribed to the abnormal expression of LmVg. To investigate the expression profile of LmVg in the locusts, we first examined the silencing efficiency of LmDDX6 in the fat body and found that it was high: 93.3%, 97.7%, and 96.7% at 4, 6, and 8 days PAE, respectively (Figure 5a). In locusts, two vitellogenin genes, A and B, have been described [36]. Therefore, we performed RT-qPCR for both LmVgA and LmVgB in the dsGFP-and dsLmDDX6-treated locusts at 4, 6, and 8 days PAE. As expected, the expression level of LmVgA significantly decreased by 64.5%, 91.9%, and 62.4%, respectively (Figure 5b). Similarly, the expression level of LmVgB at 6 and 8 days PAE was downregulated by 93.5% and 65%, whereas at 4 days PAE, no significant changes were detected between the dsGFP-and dsLmDDX6-treated locusts (Figure 5c).

Downregulation of Vg Expression by LmDDX6 Knockdown
Vitellogenin is synthesized in the fat body and is essential for oocyte development. The phenotype of the ovaries and oocytes in the dsLmDDX6-treated locusts might be ascribed to the abnormal expression of LmVg. To investigate the expression profile of LmVg in the locusts, we first examined the silencing efficiency of LmDDX6 in the fat body and found that it was high: 93.3%, 97.7%, and 96.7% at 4, 6, and 8 days PAE, respectively (Figure 5a). In locusts, two vitellogenin genes, A and B, have been described [36]. Therefore, we performed RT-qPCR for both LmVgA and LmVgB in the dsGFP-and dsLmDDX6treated locusts at 4, 6, and 8 days PAE. As expected, the expression level of LmVgA significantly decreased by 64.5%, 91.9%, and 62.4%, respectively (Figure 5b). Similarly, the expression level of LmVgB at 6 and 8 days PAE was downregulated by 93.5% and 65%, whereas at 4 days PAE, no significant changes were detected between the dsGFP-and dsLmDDX6-treated locusts (Figure 5c). The Vg receptor (VgR) is highly expressed in oocytes and is indispensable for the entry of Vg from the hemolymph into the oocyte. To this end, we examined LmDDX6 silencing efficiency in the ovary and found a slight reduction (4.6%) at 4 days PAE and decreases by 39.4% and 35.7% at 6 days and 8 days PAE, respectively. We then checked   . (b). The length × width index of primary oocytes from locusts injected with dsLmDDX6 or dsGFP at 4-8 days PAE. Data are presented as the mean ± standard error of the mean (n = 8-13). Statistical significance of differences is indicated as follows: *** p < 0.001.

Downregulation of Vg Expression by LmDDX6 Knockdown
Vitellogenin is synthesized in the fat body and is essential for oocyte development. The phenotype of the ovaries and oocytes in the dsLmDDX6-treated locusts might be ascribed to the abnormal expression of LmVg. To investigate the expression profile of LmVg in the locusts, we first examined the silencing efficiency of LmDDX6 in the fat body and found that it was high: 93.3%, 97.7%, and 96.7% at 4, 6, and 8 days PAE, respectively (Figure 5a). In locusts, two vitellogenin genes, A and B, have been described [36]. Therefore, we performed RT-qPCR for both LmVgA and LmVgB in the dsGFP-and dsLmDDX6treated locusts at 4, 6, and 8 days PAE. As expected, the expression level of LmVgA significantly decreased by 64.5%, 91.9%, and 62.4%, respectively (Figure 5b). Similarly, the expression level of LmVgB at 6 and 8 days PAE was downregulated by 93.5% and 65%, whereas at 4 days PAE, no significant changes were detected between the dsGFP-and dsLmDDX6-treated locusts (Figure 5c). knockdown compared with the corresponding levels in the dsGFP-injected group. Data in all panels are presented as the mean ± standard error of the mean (n = 8-12). Statistical significance of differences is indicated as follows: * p < 0.05; ** p < 0.01; *** p < 0.001.
The Vg receptor (VgR) is highly expressed in oocytes and is indispensable for the entry of Vg from the hemolymph into the oocyte. To this end, we examined LmDDX6 silencing efficiency in the ovary and found a slight reduction (4.6%) at 4 days PAE and decreases by 39.4% and 35.7% at 6 days and 8 days PAE, respectively. We then checked knockdown compared with the corresponding levels in the dsGFP-injected group. Data in all panels are presented as the mean ± standard error of the mean (n = 8-12). Statistical significance of differences is indicated as follows: * p < 0.05; ** p < 0.01; *** p < 0.001.
The Vg receptor (VgR) is highly expressed in oocytes and is indispensable for the entry of Vg from the hemolymph into the oocyte. To this end, we examined LmDDX6 silencing efficiency in the ovary and found a slight reduction (4.6%) at 4 days PAE and decreases by 39.4% and 35.7% at 6 days and 8 days PAE, respectively. We then checked LmVgR expression in the dsGFP-and dsLmDDX6-treated locusts. RT-qPCR results indicated that there was no significant difference in LmVgR expression between dsGFP-and dsLmDDX6-treated locusts at 4, 6, and 8 days PAE, respectively ( Figure S3). This finding could have two explanations. First, due to the weak reduction of LmDDX6 expression in the ovaries of dsLmDDX6-treated locusts, the remaining amount of LmDDX6 expression could be sufficient to maintain the normal expression of LmVgR. Secondly, it is possible that LmDDX6 indeed has no effect on the expression of LmVgR.

Knockdown of LmDDX6 Affects JH Receptor Met Expression and Its Downstream Target Genes
JH is a well-known regulator of vitellogenin synthesis during oocyte development and maturation [18]. To understand the relationship between LmDDX6 expression and JH signaling pathway activity in vitellogenin synthesis, we chose to examine mRNA levels of the JH receptor Met and its downstream target genes Grp78-1 and Grp78-2 [37]. Indeed, the expression levels of these three genes were strongly downregulated at 6 d PAE in the dsLmDDX6-treated locusts with reductions by 91.4%, 90.1%, and 79.2%, respectively. However, these reductions in expression levels slightly recovered by 8 days PAE. Furthermore, no significant differences in expression levels of these genes were detected at 4 days PAE between the dsGFP-and dsLmDDX6-treated locusts (Figure 6a-c). These data clearly indicate a role for LmDDX6 in the regulation of the JH signaling pathway.

Knockdown of LmDDX6 Affects JH Receptor Met Expression and Its Downstream Target Genes
JH is a well-known regulator of vitellogenin synthesis during oocyte development and maturation [18]. To understand the relationship between LmDDX6 expression and JH signaling pathway activity in vitellogenin synthesis, we chose to examine mRNA levels of the JH receptor Met and its downstream target genes Grp78-1 and Grp78-2 [37]. Indeed, the expression levels of these three genes were strongly downregulated at 6 d PAE in the dsLmDDX6-treated locusts with reductions by 91.4%, 90.1%, and 79.2%, respectively. However, these reductions in expression levels slightly recovered by 8 days PAE. Furthermore, no significant differences in expression levels of these genes were detected at 4 days PAE between the dsGFP-and dsLmDDX6-treated locusts (Figure 6a-c). These data clearly indicate a role for LmDDX6 in the regulation of the JH signaling pathway.  Based on these data, we propose a model for the function of LmDDX6 in ovary development and oocyte maturation (Figure 6d). LmDDX6 is expressed in the fat body and its protein product may directly regulate LmVg expression there. Alternatively or in parallel, LmDDX6 may alter the expression of the JH receptor gene LmMet and its downstream genes LmGrp78-1 and LmGrp78-2 and thereby indirectly affect the expression of LmVg. In either case, downregulation of LmVg expression blocks oocyte development and maturation.

Discussion
DDX6 and its orthologs comprise one of the evolutionarily earliest families of DDX proteins. Most members of this subfamily are composed of 400-600 amino acids. In addition to the helicase core of 350-400 amino acids, some have elongated N-terminal regions, for example, orthologs in the vertebrates, whereas others have long C-terminal regions, e.g., yeast DHH1 and Ste13 [30]. The sequences of the DDX6 orthologs show high similarity even in their N-and C-terminal regions. To clarify the motifs that might be unique to the DDX6 subfamily, we conducted a MEME motif analysis among various orthologs from the yeast to the mammals and plants. Intriguingly, five such motifs, each containing 10 residues, were identified in this study (Figure 1). Two motifs (consensus TKGNEFEDYC and KRELLMGIFE) are located at the end of the N-terminal region and are closely connected to the Q-motif, which is essential for the ATPase activity of the DDX family members [22]. The other three motifs are in the helicase core, namely, in the DEXDc domain (PYEINLMEEL), in the HELICc domain (YSCYYIHAKM), and in the linker region between the two domains (KVHCLNTLFS). These motifs may determine ATP binding affinity or specificity of the mRNA and other factors associated with DDX6 and its orthologs. To this end, information on the detailed structure and residue mutation analysis as well as related biochemical data are needed.
DDX6 sequence is closely related to that of the translational initiation factor eIF4A, which contains just the helicase core without the longer N-and C-terminal sequences [30,38]. In S. cerevisiae, two genes, TIF1 and TIF2, encode the same product, eIF4A [39]. Orthologs of eIF4A are also found in bacteria, e.g., in Acidimicrobiaceae and Magnetococcales (Supplemental File S7). Curiously, to see whether we could find a direct ortholog of DDX6 in bacteria, we performed a BLAST search using the LmDDX6 sequence in the bacterial database in NCBI. Interestingly, we found only one ortholog in M. arborescens, but not in any other bacterial species. However, in the constructed phylogenetic tree, this ortholog was found to be closely related to that of B. mori ( Figure S2), raising the question of the origin of this sequence. When we reblasted this sequence in NCBI, we found that it had 100% identity to the DDX6 ortholog in P. polytes, a species from Lepidoptera. It was reported that Lepidopteran species have large gut bacterial community and M. arborescens had been found in the midgut of Spodoptera litura [40]. Therefore, it is possible that this sequence was a contamination from P. polytes during sequencing of the M. arborescens genome. In this scenario, DDX6 appears in unicellular eukaryotes and is retained by all multicellular eukaryotes.
Many studies have demonstrated that DDX6 and its orthologs are involved in the assembly of messenger ribonucleoprotein (mRNP), RNA storage, translational repression, and mRNA decay [30]. Xp54, an DDX6 ortholog of Xenopus, is an integral component of mRNP particles in oocytes: It changes the conformation of the mRNP complex by displacing one subset of proteins to enable recruitment of the next one and thereby is involved in mRNP remodeling [41,42]. Me31B, Cgh-1, and DHH1, the DDX6 orthologs in Drosophila, Caenorhabditis, and Saccharomyces, respectively, are the core components of the processing body (PB) involved in mRNA decay, translational repression, and miRNA-mediated gene silencing [43][44][45][46][47]. The mRNA decay in the PB involves the central complex CCR4-CAF1-NOT, the decapping factor DCP1/2, and exonuclease Xrn1 [48]. DDX6 orthologs interact with the factors in the CCR4-CAF1-NOT complex, and this interaction seems to be conserved in some species [49]. Translational repression relies on other factors, such as the repressor protein RAP55 in the vertebrates and CAR-1 in C. elegans [50,51]. DDX6 orthologs bind to non-translational mRNA during oogenesis and early embryo development, and, thus, temporarily mask these mRNAs. Later, some of these mRNAs may be reused in later development [42]. In addition, DDX6 interacts directly with AGO1 and AGO2, which are involved in miRNA-mediated gene silencing [52].
Yeast Ste13 is necessary for sexual reproduction, whereas DHH1 regulates cell cycle [53,54]. Me31B, as a component of Drosophila germ granules, plays an essential role in germline development [55]. Moreover, DDX6 in mice has been shown to function in gametogenesis and early embryogenesis [56]. Similarly, the DDX6 ortholog Cgh-1 in C. elegans is required for gametogenesis and protection from physiological germline apoptosis [57]. However, the function of DDX6 in locusts has not yet been investigated.
We found that knockdown of LmDDX6 arrested locust oocyte development and maturation, indicating an essential role for LmDDX6 in female locust reproduction. Furthermore, mRNA expression levels of LmVg and JH receptor Met, as well as of Grp78-1 and Grp78-2, two downstream genes of the JH receptor complex, were significantly reduced in the fat body of the dsLmDDX6-injected locusts ( Figure 5). Therefore, LmDDX6 likely regulates ovary development and oocyte maturation by affecting vitellogenesis, at least partially, via the maintenance of the JH signaling pathway activity (Figure 6d). To determine whether LmDDX6 is also involved in mRNA decay, translational repression, and miRNA-mediated gene silencing, as are other DDX6 orthologs [30], we plan to isolate various components of the CCR4-NOT complex and the decapping factors Dcp1 and Dcp2 from L. migratoria. In order to understand better the mechanism whereby LmDDX6 controls the key processes of female locust reproduction, it will be necessary to establish whether LmDDX6 directly interacts with those factors, including LmAGO1.

Conclusions
In this study, we characterized LmDDX6, the DDX6 ortholog in L. migratoria, and found that it possesses five unique motifs to the DDX6 subfamily. LmDDX6 is closely related to its orthologs in Apis dorsata and Zootermopsis nevadensis. In adult female, LmDDX6 is highly expressed in ovary, muscle, and fat body. Knockdown of LmDDX6 elicits reduced expression levels of JH receptor Met and its downstream targets Grp78-1 and Grp78-2, downregulates LmVg expression, and impairs ovary development and oocyte maturation. As such, LmDDX6 is a key player in female locust reproduction and thus could be a novel target for locust biological control.
Supplementary Materials: The following are available online at https://www.mdpi.com/2075-445 0/12/1/70/s1. Figure S1: Motif pattern analysis of multiple DDX6 sequences from diverse species. Figure S2: Effect of LmDDX6 RNAi on the expression of locust vitellogenin receptor gene. Figure S3: Phylogenetic tree of DDX6 orthologs. Table S1: Primers used in this study. Supplemental File S1: The 48 sequences used for Figure 1 and Figure   Institutional Review Board Statement: Not applicable.

Informed Consent Statement: Not applicable.
Data Availability Statement: All data produced from this study are included in this published paper.