Cold tolerance strategies of the fall armyworm, Spodoptera frugiperda (Smith) (Lepidoptera: Noctuidae)

The fall armyworm (FAW), Spodoptera frugiperda, is native to the tropical and subtropical areas of the American continent and is one of the world's most destructive insect pests and invaded Africa and spread to most of Asia in two years. Glycerol is generally used as a cryoprotectant for overwintering insects in cold areas. In many studies, the increase in glycerol as a main rapid cold hardening (RCH) factor and enhancing the supercooling point was revealed at low temperatures. There are two genes, including glycerol-3-phosphate dehydrogenase (GPDH) and glycerol kinase (GK), that were identified as being associated with the glycerol synthesis pathway. In this study, one GPDH and two GK sequences (GK1 and GK2) were extracted from FAW transcriptome analysis. RNA interference (RNAi) specific to GPDH or GK1 and GK2 exhibited a significant down-regulation at the mRNA level as well as a reduction in survival rate when the RNAi-treated of FAW larvae post a RCH treatment. Following a cold period, an increase in glycerol accumulation was detected utilizing high-pressure liquid chromatography and colorimetric analysis of glycerol quantity in RCH treated hemolymph of FAW larvae. This research suggests that GPDH and GK isozymes are linked to the production of a high quantity of glycerol as an RCH factor, and glycerol as main cryoprotectant plays an important role in survival throughout the cold period in this quarantine pest studied.


Results
Glycerol content in plasma in response to low temperatures and exposure time. When fifth instar larvae of FAW were incubated at low temperatures (5 and 10 °C) compared to higher temperatures, their glycerol content increased more than threefold (15 and 20 °C) (Fig. 1A). The exposure period was also linked to an increase in the amount of glycerol in plasma, with the most glycerol found after 24 h of incubation (Fig. 1B). The high glycerol level indicated that it is a major component of plasma after cold stress.

Molecular architecture of glycerol biosynthesis genes.
To explain the increase in glycerol content, we attempted to identify the enzymes involved in glycerol biosynthesis (Fig. 1C). Based on a previous study 30 we chose dihydroxyacetone-3-phosphate (DHAP) as a precursor of glycerol biosynthesis from glycolysis intermediates. The catalytic activity of GPDH and GK converts DHAP to glycerol (Fig. 1D). As a result, the GPDH and GK genes were predicted to be involved in the synthesis of glycerol, which is an important cryoprotectant in insects when temperatures are extremely low. The transcriptome of FAW (NCBI accession number: GSE175545) were used to determine full open reading frames (ORFs) of GPDH (Sf-GPDH) and GK (Sf-GK1 and Sf-GK2) of FAW. Sf-GPDH, Sf-GK1 and Sf-GK2 ORFs encode for 353, 343, and 332 amino acid residues, respectively. Sf-GPDH protein contains a bi-domain protein structure, as illustrated in Fig. 2A that it encoded NAD + -dependent GPDHs with an N-terminal NAD + -binding domain and a C-terminal NAD + -dependent GPDH domain. Both identified FAW glycerol kinases shared N-terminal (FGGY-N) and C-terminal (FGGY-C) domains, which are colored blue and red, respectively, as shown in Fig. 2B, confirming that the targeted proteins are members of the FGGY carbohydrate kinase family. Based on comparisons with other well-known insect proteins, the threedimensional structures of Sf-GPDH, Sf-GK1, and Sf-GK2 proteins were predicted using the homology modeling method (Fig. 3). These findings indicated that the sequences of Sf-GPDH and two Sf-GKs closely matched the homologous templates on the server, indicating that these protein models were reliable. The GPDH domain structure of Sf-NAD + -binding revealed two key components: a spatially symmetric β-sheet core and multiple helices (α1-α17) wrapping on both sides of the β-sheet core. The bioinformatics analysis indicated that four functional amino acids including Arg99, Glu100, Phe155, and Asn266 in Sf-GK1 and Arg92, Glu93, Phe148, and Asp258 in Sf-GK2 which are as glycerol binding residues (Fig. 3C, E). Three-dimensional analysis indicates 66% homology of Sf-GPDH with Tribolium castaneum (Herbert) GPDH under 73% coverage. When the Sf-GK1 and Sf-GK2 were compared by Spodoptera litura (Fabricus) glycerol kinase, the homology was 45   Each treatment was replicated three times with 10 larvae per replication. Different letters indicate significant differences among means at (Type I error = 0.05, LSD test). (C) Chromatograms of hemolymph extracted from larvae exposed to 5 °C for 24 h. (D) A putative glycerol production pathway. Glycerol is formed by catabolizing glucose to dihydroxyacetone-3phosphate (DHAP), which is then reduced to glycerol-3-phosphate (G3P). TRE, PGM, PGI, and GPP represent for trehalase, phosphoglucomutase, phosphoglucoisomerase, and glycerol-3-phosphate phosphatase, respectively. www.nature.com/scientificreports/ and 59% coverage (Fig. 3D, F). In these two glycerol kinases, several amino acids were conserved including ATPbinding motif and FGGY signature motives (Figs. 3,4). A phylogenetic analysis indicated that the Sf-GPDH and Sf-GK1 were clustered with lepidopteran insects quite distinct from other insect orders. However, interestingly Sf-GK2 was clustered with Homopteran insect (Fig. 5).

Scientific Reports
Expression profile of glycerol biosynthesis genes and inducible expression in response to low temperature in FAW. Three glycerol biosynthesis genes were expressed in FAW (Fig. 6). They were expressed from egg to adult in whole stages of development ( Fig. 6A, C, E). In larval stage, they were expressed in different tissues such as hemocytes, fat body, midgut, and epidermis ( Fig. 6B, D, F). However, their expression levels were varied among treatments during the developmental stages. All the three genes showed high expression level at adult stages of female insects. The highest expression level of all three genes was detected at midgut tissue. The expression levels of all three glycerol biosynthesis genes were inducible in response to low temperature (5 °C), and they showed a positive correlation with increasing incubation time. (Fig. 6G).
Glycerol content is reduced by RNAi targeting glycerol biosynthesis genes following RCH. RNAi was done on each glycerol biosynthesis gene (Sf-GPDH, Sf-GK1, and Sf-GK2) by injecting gene specific double-stranded RNAs (dsRNAs) into L5 larvae (Fig. 7). All three genes showed significant decreases (P < 0.05) with incubation time when one µg of dsRNA for each gene was injected into each larva. In all three genes, the strongest RNAi effect was observed at 48 h post injection, with a ⁓ 40-80 percent drop in mRNA expression levels (Fig. 7A). RNAi downregulation of glycerol biosynthesis gene expression significantly suppressed glycerol amount (P < 0.05) in plasma at 48 h post-dsRNA injection after RCH treatment (Fig. 7B). The larvae treated with dsRNA for three genes had a basal amount of glycerol (29-35 mmol/mL), but control larvae (injected with dsRNA to enhanced green fluorescent protein (EGFP) gene) had approximately 73 mmol/mL glycerol (Fig. 7C). After RCH treatment and RNAi, the cryoprotectant(s) was monitored in hemolymph of fifth instar larvae using HPLC (Figs. 1C, 7B). Glycerol content significantly increased from 17.1 to 44.0 mM (Table 1) when the larvae were incubated at 5 °C. RNAi treatment larvae also showed a reduction in glycerol level when compare with control treatment (EGFP). Injection of dsGK2 resulted in a significant reduction in glycerol levels of more than seven times (6.08 mM) ( Table 1, Fig. 7B).

RNAi of glycerol biosynthesis genes increases the mortality of treated larvae of FAW.
Larvae at 48 h post-dsRNA injection did increase their mortality after RCH treatment (Fig. 7C, D). There was no significant difference in mortality between RCH and control (no RCH) treatment after RNAi of either Sf-GPDH or www.nature.com/scientificreports/ Sf-GK2, However the mortality decreased significantly when the larvae injected with dsRNA specific to Sf-GK1 after RCH treatment than to control (Fig. 7C).
Following RCH treatment, the SCP increased. The effect of RCH on SCP was evaluated in all developmental stages including both sexes in pupal and adult stages (Table 1). Egg, first instar and pupal stages exhibited SCP at lowest temperature than other developmental stages. The data showed that supercooling capacity was unaffected by RCH treatment in egg, first and second instar, male pupae, and female adult whereas SCP temperature in the others was significantly reduced ( Table 2). From these data, we found that RCH treatment is often accompanied by elevated SCPs. To investigate the involvement of glycerol biosynthesis genes in SCP, RNAi-treated larvae (L3 to L6) were incubated at RCH conditions and their SCP was assessed ( Table 3). The SCPs of larvae injected with dsRNA specific to glycerol biosynthesis genes, specifically dsGPDH and dsGK2, were significantly lower than those of dsEGFP-injected larvae, suggesting that glycerol biosynthesis genes elevate The functional domains of Sf-GPDH, Sf-GK1, and Sf-GK2 were demonstrated, respectively. (B, D, and F) the Sf-GPDH, Sf-GK1, and Sf-GK2 proteins respectively, were compared with same protein from another wellknown insect, including Tribolium castaneum and Spodoptera litura. Blue and pink region in (A) indicate beta sheet and alpha helices, respectively. In (C and E), the glycerol binding residues were indicated with blue atoms as well as yellow part that showing ATP-binding domains. N and C are an abbreviation for N-terminus and C-terminus of amino acid sequences. These models were made using SWISS-model web database. Three dimensional constructs were made using Chimera, version 1.13.1. www.nature.com/scientificreports/  Table S2. www.nature.com/scientificreports/    www.nature.com/scientificreports/ their SCP by accumulating extracellular cryoprotectant including glycerol in their bodies. In compared to the dsEGFP treatment group, the SCPs of larvae injected with dsGK1 were significantly lower than to dsGPDH and dsGK2 injected larvae (Table 3).

Discussion
Many insect species can develop cold-hardiness well below freezing temperatures, and various features of insect cold-hardiness have been studied 23,35 . The most significant part of acclimatization for cold resistance is low temperature exposure 22,36 . Low-weight molecular molecules, often known as cryoprotectant, such as polyols and sugars, are produced during this procedure 21 . The most prevalent cryoprotectants include polyols (glycerol, sorbitol, and manitol), sugars (glucose, trehalose, and fructose), and amino acids [37][38][39][40] . High polyol concentrations not only lower the temperature at which an insect's body fluids crystallize but also stabilize the state of proteins, even when collected in relatively low concentrations 41 . Polyols regulate the amount of water accessible for freezing, which reduces the amount of cell dehydration caused by extracellular freezing. They protect biological membranes and proteins from freezing-induced dehydration by preserving their structures 41,42 . In the present work, the tolerance of FAW was analyzed by rapid cold hardening (RCH). In insects without diapause, RCH is especially important for overcoming lethal cold shock by rapidly increasing cold tolerance 20 . RCH has been induced in a variety of insects at temperatures ranging from 0 to 5 °C 30,43-47 . Glycerol production is divided into two distinct pathways, depending on the insect. In Epiblema scudderiana (Clemens), a moth belongs to Tortricidae family, polyol dehydrogenase catalyzes the reaction of glyceraldehyde with NADPH + H + in one route 48 . The other pathway converts dihydroxyacetone-3-phosphate to glycerol via GPDH/GK (S. exigua) 30 . Identification of key genes associated with overwintering in Anoplophora glabripennis (Motschulsky) larva, a coleopteran species, using gene co-expression network analysis, was demonstrated that, fatty acid desaturase, glycerol phosphate dehydrogenase, glycerol kinase, and trehalose phosphate synthase were among the 15 genes implicated in the control of antifreeze protectants 49 . We studied on GPDH and GK genes expression to investigate the glycerol production pathway. In the FAW transcriptome, we discovered two GK isoforms and one GPDH isoform. It was discovered that both genes expressed and associated with glycerol biosynthesis pathway. The whole Plutella xylostella (Linnaeus) genome was used to predict four GKs and one GPDH 50 . The genome of FAW contains only one type of GPDH, indicating that it is a unigene with a conserved biological function in metabolism. Because we obtained these sequences from transcriptome data and there are likely no other endogenous genes of GPDH and GK, we believe our expression and functional analysis are associated with these isozymes. GPDH and both GK isoforms were discovered to be widely expressed in different studied tissues. As we know at low temperatures, most gene expression decreases 51 . However, in 5 °C, real-time PCR of cold-exposed larvae revealed that GPDH, GK1, and GK2 were expressed at relatively high levels (Fig. 6G). This suggests that these proteins are important for cold tolerance to the low temperature by RCH. As found in other insects 52,53 , cold tolerance rose as acclimatization time increased, which could be in line with our findings, that mRNA expression levels of analyzed genes increased as incubation time increased (Fig. 6G).
RNAi is a non-invasive way of delivering dsRNA into insects to knockdown specific gene expression [54][55][56][57] . We have shown that injecting RNAi is feasible and can suppress the transcription level of target genes in FAW larvae. Our system confirmed the effective knockdown of three genes at the mRNA expression level.
Their expression was knocked-down by specific dsRNAs associated with glycerol biosynthesis genes. In response to pre-exposure to a low temperature, this RNAi treatment reduced RCH and prevented glycerol accumulation. According to the RNAi experiments, sorbitol dehydrogenase, trehalose-6-phosphate synthase, and glycerol kinase are all involved in the overwintering stage of Chinese white pine larvae (Dendroctonus armandi (Tsai and Li)) 58 . Glycerol phosphorylation, which is essential for glycerol consumption, is catalyzed by GK 59,60 . GK has a function in overwintering termination in Hyalophora cecropia (Linnaeus) eggs that accumulate glycerol by converting glycerol to glycerol-3-phosphate for other intermediary metabolism 61 . In A. glabripennis larvae, the gene expression level of glycerol kinase increased sharply at the midpoint of the overwintering stage, and then declined at the latter, which corresponded to the change in glycerol content. The findings suggest that glycerol kinase is involved in the synthesis of glycerol, which could help this insect adapt to low temperatures 49 .
Because RNAi targeting GK1 and GK2 significantly reduced glycerol accumulation in a 5 °C pretreatment, it was shown that both GKs catalyze the dephosphorylation of glycerol-3-phosphate to generate glycerol, as reported earlier in S. exigua, a near Noctuidae species to FAW 30,62 . However, it was discovered that GK2 has a greater effect on glycerol production based on RNAi data for mortality, glycerol accumulation, and HPLC results. The significant increase in Sf-GK2 expression vs Sf-GK1 shows that it has physiological significance in RCH, as evidenced by the RNAi functional study. P. xylostella GK1 showed a significant increase in expression in response to 5 °C exposure vs other three isozyme of GKs 50 . In Bombyx mori (Linnaeus), at least three GK isozymes have been discovered, but only one, GK3, appears to be connected to glycerol utilization 63 . The knockdown of the target genes Sf-GPDH, Sf-GK1, and Sf-GK2 not only reduces their transcription levels but also affects larval cold-tolerance capacity, leading to an increase in low-temperature mortality. The most obvious explanation for these findings is that these genes are necessary for overwintering larvae's cold tolerance. As there is no existing evidence of systemic spread in Lepidoptera 64 , we were unable to totally silence these three genes, however, the partial knockdown had a clear effect on low-temperature mortality.
The hemolymph polyol analysis revealed that trehalose was the primary blood sugar, with a concentration of 4.42 mmol −1 in hemolymph and a slight increase with low temperature exposure (5.99 mmol −1 after 6 h at 4 °C). Trehalose titers in insect hemolymph are relatively high in general, but very considerably between insects (ranging from 0.1 to 133 mmol −1 ) 30 . We detected 5.7 mM of trehalose following dsGK1-RCH treatment, whereas the titer increased considerably to 21.53 mM following dsGK2-RCH treatment. This may be a compensating effect of the glycerol depletion. However, we believe that in order to obtain more precise results, we need incorporate trehalose(s) (which catalyzes the conversion of trehalose to two glucose monomers) into our future studies.
In conclusion, due to a lack of a diapause mechanism, FAW cannot overwinter in area with a cold winter, despite the fact that they can disperse thousands of kilometers north during the growth season 65,66 . However, in this study, RNAi investigation of two types of important genes linked to glycerol production and their effects www.nature.com/scientificreports/ on glycerol accumulation and insect mortality in response to low temperature pre-exposure, revealed that glycerol is a substantial cryoprotectant in RCH in FAW. Increased glycerol concentrations may contribute to whole animal freeze tolerance by enhancing cell survival by freeze-tolerant. However, each cryoprotectant may have a distinct non-overlapping function and contribute to freeze tolerance through memchanisms distinct from those of others with different potency. In addition, the permeability of different tissues to cryoprotectants can be vary, affecting their ability to protect cells and this constituents. Supercooling data clearly demonstrated that FAW can endure very low temperatures, and as a key agricultural pest, it may be able to become one of most important migratory insect pests in Korea. To limit the impact of this pest, it is critical to create pest management strategies and detecting systems. In addition, more research on migration behavior is needed to predict source areas and migration times.

Materials and methods
Insect rearing, exposure temperatures and sample preparation. to detect the release of the latent heat of fusion as body water froze, as described previously 32,71 . In SCP measurement, all developmental stages of FAW were examined after RCH treatment (exposed to 5 °C for 6 h prior to − 10 °C for 1 h). The thermocouples were kept in contact with the cuticle by putting the insect in a 1.5 mL tube and filling it with cotton wool to keep the insect and thermocouple together ( Figure S1). They were then put in a styrofoam box (30 × 30 × 20 cm), and the box was placed into a freezer at − 80 °C. The cooling rate was measured as 1 °C min −1 .
Glycerol analysis in FAW hemolymph. The glycerol content of the samples was determined using the Glycerol Assay Kit (BioVision, Milpitas, CA, USA). We followed the manufacturer's instruction for fluorometric measurements. In summary, a hemolymph from 10 fifth instar larvae (Day 1) was collected by cutting prologs of the treated larvae and mixed with a 100 µL volume of anticoagulant buffer (ACB). ACB was prepared with 186 mM NaCl, 17 mM Na 2 EDTA, and 41 mM citric acid 72 . The ACB was adjusted to pH 8.0 by the addition of NaOH. The resultant hemolymph was centrifuged at 13,500 rpm for 10 min at 4 °C. The supernatant (100 µL) were mixed with 100 µL of glycerol assay buffer (GAB) (provided by the kit) for 10 min on ice. The amount of 10 µL resulted supernatant (12,000 rpm, 5 min) was mixed with 86 µL GAB followed by 2 µL Probe (provided by www.nature.com/scientificreports/ the kit) and 2 µL glycerol enzyme mix (GEM) (provided by the kit) in a 96 well plate. The background control mixture was prepared as described above without GEM. In this assay, glycerol in the presence of glycerol enzyme mix is converted to an intermediate after incubation at 37 °C for 60 min, which reduces a colorless probe to a colored product with strong absorbance at 450 nm.

Down-regulation of associated glycerol biosynthesis genes by RNA interference (RNAi).
RNAi was performed using dsRNA prepared with Megascript RNAi Kit (Ambion, Austin, TX, USA) according to the manufacturer's instruction and a previous method 73 . Partial segments were amplified with gene-specific primers containing T7 promoter sequence at 5′ end (Table S1). dsRNAs (dsGPDH, dsGK1 and dsGK2) were synthesized at 37 °C for 4 h and then left at 70 °C for 5 min to inactivate T7 RNA polymerase. As control dsRNA ('dsCON'), 300 bp fragment of enhanced green fluorescent protein (EGFP) was synthesized 74 . Three µg of dsRNA (1 µg/ µL) was injected into each fifth instar larva with a Hamilton micro syringe. RNAi efficiency was determined by qPCR described above at 24, 48, 72, and 96 h post injection. For each treatment, at least 10 larvae were used. Each treatment was replicated three times.
Sample preparation and HPLC condition. Hemolymph of 10 fifth instar (Day 1) was collected by cutting prologs of the treated larvae and mixed with a 100 µL volume of ACB. The resultant hemolymph was centrifuged at 13,500 rpm for 10 min at 4 °C. The supernatant (500 µL) was transferred to a new 1.5 mL tube, and then the same volume of acetonitrile (ACN) was added into the tubes and were shaken for 15 s. The tubes were incubated at room temperature for 10 min and then centrifuged as described above. The upper phase was collected in a new 1.5 mL tube. The previous step was repeated with the addition of 250 µL ACN to increase the purification. The final supernatant was filtered out by 0.22 µM syringe filters. The purified samples were directly used for HPLC in Metabolomics Research Center for Functional Materials, Kyungsung Univeristy (Busan, Korea). A reversed-phase HPLC connected to an evaporative light scattering detector (ELSD) (ELSD-LT II, Shimadzu, Japan) was optimized for simultaneous determination of cryoprotectant. HPLC separation was achieved using a Unison UK-Amino column (250 × 4.6 mm). Water and acetonitrile were used as the mobile phase. The ideal flow rate of 0.7 mL/min and a proportion of acetonitrile of 90% over 30 min were used to optimize the separation of cryoprotectant using isocratic elution conditions. The temperature of the ELSD detector was set at 30 °C with a temperature of column oven at 64 °C. Calibration curves were generated in GraphPad Prism by plotting the area against cryoprotectant concentration. Averages of the areas for each standard were calculated and plotted against the known concentrations.

Data analysis.
All studies were performed in three independent biological replicates. Results were plotted using Sigma plot 10.0. Means were compared by least squared difference (LSD) test of one-way analysis of variance (ANOVA) using PROC GLM of SAS program 75,76 and discriminated at Type I error = 0.05.