Serine protease inhibitors LmSPN2 and LmSPN3 co-regulate embryonic diapause in Locusta migratoria manilensis (Meyen) via the Toll pathway

FENG Shi-qian, ZHANG Neng, CHEN Jun, ZHANG Dao-gang, ZHU Kai-hui, CAl Ni, TU Xiong-bing,ZHANG Ze-hua

State Key Laboratory for Biology of Plant Diseases and Insect Pests, Institute of Plant Protection, Chinese Academy of Agricultural Sciences, Beijing 100193, P.R.China

Abstract Female adults of the migratory locust, Locusta migratoria manilensis (Meyen), can sense seasonal photoperiod changes,which induces embryonic diapause as a key strategy to overwinter.Serine protease inhibitor genes (SPNs) were thought to play key roles during diapause, while few SPNs were functionally characterized.LmSPN2 was one of those genes differentially expressed between diapause and non-diapause eggs; however, its biological function remained to be explored.So, we conducted RNAi knockdown of LmSPN2, resulting in a significant decrease of the egg diapause rate by 29.7%.Using yeast two-hybrid assays, co-immunoprecipitation, and pull-down methods, we found an interaction between LmSPN2 and LmSPN3, which was proved to be mediated by a glutamate (E331) binding site of LmSPN2.RNAi knockdown of LmSPN3 resulted in a significant increase in diapause rate by 14.6%, indicating an inverse function of LmSPN2 and LmSPN3 on diapause regulation.Double knockdown of two SPN genes resulted in a 26.4% reduction in diapause rate, indicating that LmSPN2 was the dominant regulatory signal.Moreover, we found four Toll pathway genes(easter, sp?tzle, pelle, and dorsal) upregulated significantly after the knockdown of LmSPN2 while downregulated after the knockdown of LmSPN3.Therefore, we speculate that two SPNs regulate diapause through the Toll pathway.Our results indicated that LmSPN2 positively regulates locust egg entry into diapause, while LmSPN3 is a negative regulator of embryonic commitment to diapause.Their interaction is mediated by the binding site of E331 and influences egg diapause through the Toll pathway.This mechanistic understanding of diapause regulation expands our understanding of insect developmental regulation and provides functional targets for developing locust management strategies.

Keywords: Locusta migratoria, insect diapause regulation, Toll pathway, protein interaction, serine protease inhibitor

1.lntroduction

Diapause is an important strategy for insect survival under the cold conditions of winter, which entails slow or stagnated growth, respiratory rate and free water content depression, and increased reserve energy sources such as fats, proteins, and saccharides (MacRae 2010; Chenet al.2020b).Insect diapause can occur at several different life stages, including the egg, larval,pupal, and adult stages.For example, some species of the Chrysomelidae family undergo egg diapause(Krysanet al.1984),Aedesatropalpusundergoes embryonic diapause (Gregg 1985), Saturniidae insects undergo pupal diapause (Sahooet al.2018), andLeptinotarsadecemlineatadiapauses at the adult stage (Yocum 2001).In addition, diapause can be categorized as either facultative or obligate.That is,obligate diapause insects such asOsmialignariaenter diapause at a determined growth stage regardless of the environmental conditions (Sgolastraet al.2010).By contrast, facultative diapause insects are greatly affected by environmental factors.For example, a short day length stimulatesOstrinianubilalisto store energy and enter diapause, while suitable environmental stimulation helpO.nubilalisend the diapause process and maintain normal development (Irvinget al.2000).

Locustamigratoria, an insect pest notorious for its massive damage to crops (Zhanget al.2019),undergoes facultative embryonic diapause, mainly triggered by the maternal perception of temperature,photoperiod, and egg incubation temperature (Tuet al.2015; Wanget al.2021).Before diapause, the embryo undergoes anatrepsis, and development stagnates(Tanaka 1992, 1994).It takes approximately 25 days for the normal eggs to hatch at 25°C while the diapause embryo stops at the revolution stage in about 15 days(Wanget al.2021).The embryo in diapause starts its development after a low-temperature treatment.However, the maternal mechanism for inducing the diapause phenotype in response to environmental signals, or the regulatory signaling used by adult females to stop development in the offspring, remains unknown.This lack of knowledge is exemplified by the situation inBombyxmori(Tsuchiyaet al.2021).It is known that serine protease inhibitors (serpins or SPNs) contribute several essential regulatory functions to embryonic development in insects (Irvinget al.2000; Chuet al.2015).This superfamily of serine and tyrosine protease inhibitors is widely distributed across most branches of the tree of life (Robertset al.2003;Viswanathanet al.2012; Eappenet al.2013).Serpins have been shown to perform a wide range of biological functions by inhibiting or modifying target proteins (Jianget al.2003; Kanost and Gorman 2008; Kanost and Jiang 2015).For example, inDrosophila, Serpin27A can regulate embryonic development through the Toll protein cascade (Ligoxygakiset al.2003; Valanneet al.2011).In this cascade, Sp?tzle controls development by activating the Toll signaling pathwayviaa proteolytic regulatory cascade that includes Gd, Snake, and Easter(Gangloffet al.2008; Meekinset al.2017).Serpin27A can thus inhibit Easter activity and consequently inhibit the degradation of the Sp?tzle precursor, thereby blocking the transmission of the activation signal to the downstream target MyD88 and ultimately terminating the Toll nuclear development signal (Ligoxygakiset al.2003;Lemaitre 2004; Qinet al.2015).

Previously, we conducted transcriptomics analyses to compare gene regulation in diapausing and nondiapausing eggs inL.migratoria.The results revealed thatLmSPN2expression in diapausing eggs was significantly higher than that in non-diapausing eggs(Haoet al.2017, 2019; Cuiet al.2019), which suggests a potential involvement ofLmSPN2in the regulation of insect embryonic development (Denlinger 2002).SPN genes were proven to regulate the Toll pathway and then influence the development, providing a useful clue on detecting SPN functions (Etebari and Sassan 2013; Katsukawaet al.2018).In the present study,we used RNAi suppression of LmSPN2 to determine whether and how LmSPN2 is involved in the regulation of embryonic diapause in the locust.To screen for candidate interacting proteins, we used yeast two-hybrid(Y2H) assays, followed by co-immunoprecipitation and pull downs of LmSPN2 bothinvivoandinvitro.Ultimately, we confirmed that LmSPN2 interacts with a second serpin, LmSPN3, to coordinate the regulation of the Toll pathway genes, includingeaster,myD88,pelle,anddorsal, as a mechanism for controlling the initiation or suppression of diapause in locust (Meekinset al.2017).

2.Materials and methods

2.1.lnsect rearing

TheL.migratoriawas previously obtained from fields in Tianjin, China (38°49′N, 117°18′E) in November 2007.All insects were maintained in our lab at the State Key Laboratory for Biology of Plant Diseases and Insect Pests, Institute of Plant Protection, Chinese Academy of Agricultural Sciences.New first instars were kept in 40 cm×40 cm×40 cm rearing cages and transferred to 20 cm×20 cm×28 cm mesh cages until the fourth instar.Then, the cages were placed in artificial climate chambers(PRX-250B-30; Haishu Saifu Experimental Instrument Factory, Ningbo, China).The conditions were set at 27°C and 60% RH under a short photoperiod of 10 h L:14 h D to produce diapause eggs, while non-diapausing eggs were collected under long photoperiod (27°C, 60% RH,16 h L:8 h D).

2.2.cDNA synthesis, gene cloning, and vector construction

Locust eggs were used for the extraction of total RNA by using TRIcom reagent (Tianmo, Huailai, China).Total RNA was isolated according to the manufacturer’s protocol.The quality was also checked.cDNA was synthesized according to the Prime-ScriptTM II 1st Strand cDNA Synthesis Kit (TaKaRa, China).By analyzing the transcriptome of the migratory locust, we obtained the sequence of SPN genes, and primers were then designed by DNAMAN (version 7.212; Appendix A).Using the cDNA of migratory locusts as a template, we amplified the SPN genes.The obtained PCR product was purified by TIANgel Midi Purification Kit (Tiangen, Beijing, China) and was connected to the pGADT7, pGBKT7, Pet28a, pEGX-6P-2, pAC5.1/5V-HisB vectors.Later, the recombinant was transformed intoEscherichiacoliTrans1-T1 strain,and 500 μL lysogeny broth (LB) liquid medium was added.Notably, no restriction enzyme was used.The obtained product was allowed to shake at 200 r min-1at 37°C for 2 h.A total of 100 μL bacterial solution was applied to LB solid medium, including 1% of ampicillin.The medium was incubated at 37°C for 12 h.The recombinant colonies were transferred into a liquid LB culture medium containing 1% ampicillin and were shaken for 3-6 h at 37°C.Finally, the medium for the PCR template was prepared.

2.3.Synthesis and injection of dsRNA

The recombinant plasmid containing SPN gene fragment was extracted using EZgeneTMPlasmid Miniprep Kit(Biomiga, San Diego, CA, USA).The SPN genes were amplified by primers using the recombinant plasmid as a template (Appendix A).The amplified PCR products were then purified with TIANgel Midi Purification Kit(Tiangen, Beijing, China), followed by quantification through NanoPhotometerTM(Implen GmbH, Munchen,Germany).SPN genes’ double-stranded RNA (dsRNA)was synthesized using the T7 RiboMAXTMExpress RNAi System Kit (Promega, Madison, WI, USA).The dsRNA concentration of SPN genes was detected by a NanoPhotometerTM(Implen GmbH, Munchen, Germany),and the final concentration was adjusted to 1 μg μL-1for further analysis.

A total of 5 μL of dsRNA (2 μg μL-1) targetingLmSNP2,or dsGFPas a control, were injected into the ventral section between the 2nd and 3rd abdominal segments of female adults within 72 h after molting under a short photoperiod.After dissecting the whole bodies of dsRNAinjected and control group’s adult locusts after 48 h, we obtained ovary, and the efficiency of RNAi-mediated knockdown was then determined.

2.4.Quantitative real-time polymerase chain reaction (qRT-PCR)

cDNA was synthesized from the RNA samples above using M-MLV reverse transcriptase and recombinant RNase inhibitor (TaKaRa, China) (Shinet al.1998).The expression levels of the genes were determined by qRT-PCR using the SYBR PremixExTaqKit (TaKaRa,China) following the manufacturer’s instructions in an ABI 7500 Real-time PCR System (Applied Biosystems,Foster City, CA, USA).qRT-PCR was performed as per the following conditions: 95°C for 10 min; 40 cycles of 95°C for 15 s, 60°C for 45 s.Gene expression was quantified using the 2-??Ctmethod (Cuiet al.2019; Chenet al.2020b) withβ-actinas the internal control for data normalization.The specific primers used for qRT-PCR are listed in Appendix A.

2.5.Diapause rate test

Female locusts of each treatment and replication were placed in new mesh cages (25 cm×25 cm×35 cm)and provided with wheat grown in a greenhouse.Subsequently, 30 adult males were presented to each replicate to mate.The bottom of the cages was covered by a 5-cm layer of sieved sterile sand, which was changed every two days.Mating occurred for about 10 days until oviposition was observed.Eggs were collected at a 48 h-interval for 10 days using a camel paint brush and transferred into paper cups(10 mm×5 mm), where the eggs were incubated on vermiculite before shifting to 27°C and 60% RH to slow down the development.Around 150 eggs were obtained from 3-4 pods, which were then used in each experimental replication.Eggs were kept under 27°C for 20 days until the eclosion of the first instar nymphs ceased (D1).To account for non-viable eggs, all remaining uneclosed eggs were kept at 4°C for 60 days to receive ample time to break the diapause and were then incubated at 27°C for 20 days (D2).The diapause rate (DR) was calculated as DR (%)=D2/(D1+D2)×100.

2.6.Yeast two-hybrid (Y2H) analysis

We used the Matchmaker? Gold Yeast Two-Hybrid System of Clontech (Takara Bio USA, CA, USA).FulllengthLmSPN2cDNA cloned into pGBKT7 vector served as bait.Then, we clonedeasterinto the pGADT7 vector as prey.Next, we co-transformed the bait vector and prey vector into Y2HGold yeast and observed the growth conditions on the quadruple dropout medium (QDO), i.e., SD/-Leu/-Trp/, as well as on the QDO supplied with AbA and X-a-Gal, to confirm their interactions.pGBKT7-53 (encodes Gal4 DNA-BD fused with murine p53) mated with pGADT7-T (encodes Gal4 AD fused with SV40 large T-antigen) served as positive control.pGBKT7-Lam (encodes Gal4 BD fused with lamin) mated with pGADT7-T served as negative control.Other genes used for verifying the interaction were carried out by the same procedure.

2.7.Pull-down assay

Protein purification, magnetic bead incubation, and Western blot analysis was conducted following the method of Loucheet al.(2017).The following antibodies were used:anti-GST-Tag antibody (HRP conjugated) (Abmart M20025, 1:5 000) and anti-His-Tag antibody (HRP conjugated) (Abmart M20020, 1:5 000).

2.8.Co-immunoprecipitation assay

DrosophilaS2 cells were cultured following the method of Longet al.(2019).Then, lysates were incubated in a mixture containing 3 μL mouse anti-Flag (Abmart M20008)antibody and Protein A+G Agarose beads (Beyotime P2012) overnight at 4°C.After washing in lysis buffer five times, immunoprecipitants were boiled in 5× SDS loading buffer to dissociate proteins from beads.The following antibodies were used for Western blots to detect coimmunoprecipitated proteins: mouse monoclonal anti-Flag(Abmart M20008, 1:2 000) and Myc-Tag (19C2) Mouse Antibody (Abmart M20002, 1:5 000).

2.9.Protein 3D structure prediction

The protein data bank (PDB) profile of LmSPN2 was obtained from the website (http://www.sbg.bio.ic.ac.uk/phyre2/html/page.cgi?id=index), and the document was visualized by Swiss-PDB Viewer developed within the Swiss Institute of Bioinformatics (SIB).

2.10.Statistical analysis

We compared the differences between treatments by Student’st-test.We usedP<0.05 as the threshold level to determine the variability and reported values as mean±SE.We analyzed data using the SPSS Software version 20.0 (SPSS Inc., Chicago, IL, USA)and GraphPad Prism Software version 6.01 (GraphPad Software Inc., San Diego, CA, USA).

3.Results

3.1.LmSPN2 expression is upregulated during embryo diapause

To better understand how changes in the expression ofLmSPN2can participate in the regulation of diapause inL.migratoriaembryos, we first quantified the differences in its transcription between diapausing and nondiapausing eggs.The results showed thatLmSPN2FPKM values were 34.9124 and 9.9304 in diapausing and non-diapausing eggs, respectively, indicating strong up-regulation of mRNA levels in the diapausing eggs(Log2FC=1.8138) (Fig.1-A).This shift inLmSPN2relative expression between diapausing and non-diapausing eggs was verified by qPCR-based assays (Fig.1-B), which confirmed the results of the transcriptomic analysis.We thus speculated thatLmSPN2was likely involved in the regulation of diapause inL.migratoria.

Fig.1 LmSPN2 plays a positive role in embryonic diapause in Locusta migratoria.A, transcriptome analysis of LmSPN2 in diapausing and non-diapausing eggs.B, relative mRNA level of LmSPN2 in diapausing and non-diapausing eggs by qPCR.C,knockdown efficiency (RNAi) for LmSPN2 in ovaries.D, the egg’s diapause rate after LmSPN2 knockdown.E, the expression level of the indicated Toll pathway genes after LmSPN2 knockdown.Data are the mean±SEM from five independent replicates.＊＊, P<0.01; ＊＊＊, P<0.001.

3.2.LmSPN2 knockdown decreases diapause rate in adult L.migratoria

To explore whether and howLmSPN2regulates the embryonic shift to diapause, we next tested the effects ofLmSPN2knockdown by RNAi on the locust diapause.To this end, we cloned the 1 164 bpLmSPN2gene,encoding 387 amino acids (molecular weight=42.78 kDa)and synthesized double-stranded RNA sequences ofLmSPN2(dsLmSPN2).Injection of adult females with the dsLmSPN2construct resulted in a significant reduction of 97.3% (P<0.001) inLmSPN2transcription in eggs under diapause conditions (27°C, 10 h L:14 h D), compared with that of the dsGFP-injected controls (Fig.1-C).Furthermore, the diapause rate in the dsLmSPN2group(43.0±5.6%) was significantly lower than that of the control insects (72.7±4.0%) (P=0.0017) (Fig.1-D).These results showed that interference withLmSPN2could inhibit the diapause of locust eggs (29.7% decrease in diapause rate afterLmSPN2knockdown), indicating thatLmSPN2contributed to the regulation of locust diapause.

3.3.LmSPN2 knockdown leads to suppression of Toll pathway signaling

We next tested whether the Toll pathway was required for the regulatory effects ofLmSPN2in diapause by investigating changes in the expression of its downstream signal genes due to major role ofLMSPN2inL.migratoriaembryonic development.We found that the expression of theeaster(P<0.0001),myD88(P=0.0004),pelle(P<0.0001),anddorsal(P=0.0002) were significantly increased after thedsLmSPN2treatment (Fig.1-E).These results indicated that interference withLmSPN2could promote developmental signal transmission and subsequently inhibit diapause in locust eggsviathe Toll pathway.

3.4.LmSPN2 potentially interacts with LmSPN3 in diapausing eggs

To further identify potential interaction partners of LmSPN2, we next used LmSPN2 as the bait protein in a Y2H system, which yielded 43 candidate target genes from a cDNA library of diapausing eggs.After rescreening these candidates on SD/-Trp/-Leu/-His/-Ade/X/A medium, we obtained 12 single clones.The genetic comparison showed that five of these had frame-shift mutations.Interestingly, the NCBI blast comparison showed that six of the remaining seven proteins were the same, identified as LmSPN3 inL.migratoria(Appendix B).Further analysis ofLmSPN3showed that its nucleotide sequence included 1 155 bp and encoded a 384 amino acid gene product with a molecular weight of 43.06 kDa.

Since Serpin27A can reportedly inhibit Easter activity as part of its function in the Toll protein cascade regulation of embryonic development (Meekinset al.2017), we next sought to determine whether LmSPN2 and LmEaster directly interact in the migratory locust.We thus constructed LmSPN2 pGBKT7 expression vectors for Y2H assays and assessed the auto-activation and toxicity of the bait protein in yeast.The results showed that LmSPN2 was safe for yeast and did not undergo autoactivation, suggesting that LmSPN2 and LmEaster did not directly interact (Appendix C).

3.5.LmSPN3 is a negative regulator of diapause via Toll signaling

To better understand howLmSPN3participates in diapause, we first examined its patterns of expression in RNA-seq data, which revealed that its FPKM values were 3.4455 and 22.5285 in diapausing and non-diapausing eggs, respectively, indicating a Log2FC=-2.7090 downregulation in its expression during diapause(Fig.2-A).qPCR-based validation of these expression levels in diapausing and non-diapausing eggs also indicated that its expression was decreased under diapause, confirming the results of RNA-seq analysis(Fig.2-B).In light of these findings, it was reasonable to speculate that LmSPN3 was involved in the regulation of diapause in the migratory locust.

Fig.2 LmSPN3 plays a negative role in embryonic diapause in Locusta migratoria.A, transcriptome analysis of LmSPN3 in diapausing and non-diapausing eggs.B, relative mRNA level of LmSPN3 in diapausing and non-diapausing eggs by qPCR.C,knockdown efficiency (RNAi) for LmSPN3 in ovaries.D, the egg’s diapause rate after LmSPN3 knockdown.E, the expression level of the indicated Toll pathway genes after LmSPN3 knockdown.Data are the mean±SEM from five independent replicates.＊＊, P<0.01; ＊＊＊, P<0.001.

To illuminateLmSPN3gene function, we generated dsLmSPN3RNAi constructs and injected them in 72-h-old adult females under diapause conditions.Quantification ofLmSPN3in locust eggs indicated that its expression was significantly lower than that in the control eggs by 92.1% (P<0.001) (Fig.2-C).Notably, the diapause rate(87.3±2.1%) was significantly higher than the diapause rate in control eggs (72.7±4.0%) (P=0.0050) (Fig.2-D).These results (14.6% increase in diapause rate afterLmSPN3knockdown) showed thatLmSPN3was likely a negative regulator of locust diapause under short photoperiod.In addition, we also assessed changes in the transcription of downstream genes in the Toll pathway and found that the expression of theeaster(P<0.001),myD88(P<0.001),pelle(P<0.001), anddorsal(P<0.001) were significantly decreased after the knockdown ofLmSPN3(Fig.2-E).Our results suggested thatLmSPN3knockdown could inhibit developmental signal transmission and subsequently induce diapause in locust eggs.

3.6.LmSPN2 interacts with LmSPN3 in vitro and in vivo

To verify the interaction between LmSPN2 and LmSPN3,we inserted the LmSPN2-encoding sequence into the pGADT7 vector to generate the AD-LmSPN2 vector construct.Similarly, we inserted LmSPN3 into the pGBKT7 vector to create the BD-LmSPN3 construct.The Y2H assays indicated that LmSPN2 and LmSPN3 could apparently interact with each other (Fig.3-A).To further verify this interaction, we then fused LmSPN2 with a His tag and fused LmSPN3 with a GST tag and respectively transfected the two constructs intoE.colifor expression and purification.Western blot analysis of lysates purified by GST affinity chromatography (Fig.3-B)showed that LmSPN2 and LmSPN3 could indeed interactin vitro.In addition, we usedD.melanogasterS2 cells to verify the interaction between these proteins, with purified recombinant Easter used as a negative control for binding and β-actin as an internal reference.Coimmunoprecipitation assays showed that LmSPN2 could also interact with LmSPN3in vivoof theD.melanogasterS2 cells (Fig.3-C).

Fig.3 Exploration of the interaction site between LmSPN2 and LmSPN3.A, reverse verification of interaction between ADLmSPN2 and BD-LmSPN3 in Y2HGold yeast, T-53 was used as the positive control while T-Lam was used as the negative control.B,confirmation of interaction between His-LmSPN2 and GST-LmSPN3 through GST pull-down.C, demonstration of the interaction between Flag-LmSPN2 and Myc-LmSPN3 using co-immunoprecipitation in Drosophila S2 cells, Myc-Easter was used as a negative control, and β-actin was served as reference protein.D, the diapause rate after co-injection with dsLmSPN2 and dsLmSPN3.E,confirmation of interaction between truncated proteins and LmSPN3 respectively in Y2HGold yeast, T-53 was used as the positive control while T-Lam was used as the negative control.F, the site of LmSPN2-E331 on the predicted 3D structure of LmSPN2 using Swiss-Pdb Viewer.G, confirmation of interaction between the point mutants of LmSPN2 and LmSPN3 in Y2HGold yeast, T-53 was used as the positive control while T-Lam was used as a negative control.The shallow grey represents no interaction, while the bright blue is an indication of molecular interaction.

In addition, we also checked the combined effects of both LmSPN2 and LmSPN3 on locust diapause by coinjection of dsLmSPN2and dsLmSPN3into 72-h-old adult females under diapause conditions.We found thatLmSPN2-LmSPN3double knockdown resulted in a diapause rate of 46.3±5.7%, which was significantly lower than that of control (72.7±4%) (P=0.0028) (Fig.3-D).However, this diapause rate was not significantly different from that of insects with a single knockdown ofLmSPN2(43.0±5.6%).This result indicated that although LmSPN2 and LmSPN3 both participated in regulating diapause in locust eggs (double knockdown resulted in a 26.7%reduction in diapause rate), LmSPN2 function was dominant over that of LmSPN3 under diapause conditions.

3.7.The interaction site between LmSPN2 and LmSPN3

To identify the interaction site on LmSPN2 required for its binding with LmSPN3, we first divided its full peptide sequence into five fragments (i.e., 1-188, 189-288,289-331, 332-378, 379-387).We then generated sequential deletions beginning at the N-terminus, which resulted in truncation variants with decreased size but intact carboxyl terminus (Appendix D).These variants were individually inserted into the pGBKT7 vector to check for the interaction with AD-LmSPN3 by Y2H assay(all primers are listed in Appendix A).The results showed that the deletion of the 378th N-terminal residues did not affect LmSPN2 interaction with LmSPN3, but the deletion of the 331st N-terminal amino acids resulted in abolished interaction between LmSPN2 and LmSPN3(Fig.3-E).We also deleted the 288th N-terminal amino acids and the 188th N-terminal amino acid residues;both of them did not affect the interaction (Fig.3-E).We then generated four deletions (at breakpoints 331-334,335-347, 348-364, and 365-377, beginning at the N-terminus) to sequentially shorten the region containing residues 331-377.We found that only the deletion of SPN2331-334resulted in lost interaction.We subsequently removed individual residues SPN2334, SPN2333, SPN2332,and SPN2331and found that the deletion of glutamate residue 331 blocked binding, thus demonstrating that it was specifically essential for interaction with LmSPN3(Fig.3-E).Based on structural analysis, we found that LmSPN2-E331was located at the beginning of a loop in the reaction center (Fig.3-F; blue).Based on these results,we then used the PCR method to generate point mutation conversions, which verified that the loss of E331(glutamic acid) resulted in abolished interaction of LmSPN2 and LmSPN3 (Fig.3-G).Collectively, these results led us to conclude that LmSPN2-E331was required for the interaction between LmSPN2 and LmSPN3.

4.Discussion

Numerous signal pathways contribute to activating diapause during insect development, especially the insulin-like pathway (Das and Arur 2017), the MAPK pathway (Wartenberget al.2001), and the Toll pathway(Eappenet al.2013).In this study, we found that both LmSPN2 and LmSPN3 participate in the regulation of locust diapauseviaToll pathway signaling (Figs.1-E and 2-E).Diapause is a special developmental process with increased immunity (Kubraket al.2014).It is predictable that the Toll pathway participated in the regulation of diapause as it played important roles in insect development and immunity (Medzhitov 2001; El-Zayatet al.2019).Serpin genes were proved to be activated or concomitantly changed with the Toll pathway (Bianet al.2005; Anet al.2010).In our study, Y2H assays indicated that neither LmSPN2 nor LmSPN3 directly interacted with Easter prior to diapause (Appendix C).This finding was unexpected in light of previous work that showed the closely related Serpin27A interacts with Easter at the embryo development stage inDrosophila(Ligoxygakiset al.2003).This difference may be due to functional divergence between insect species despite the common participation of LmSPN2, LmSPN3, and Serpin27A in insect development (Huntington 2011; Meekinset al.2017).Alternatively, LmSPN2/LmSPN3 may interact with Easter at different developmental stages in locust than those examined here (Sugdenet al.2010; Raglandet al.2019).Although we found no direct interaction between LmSPN2/LmSPN3 and Easter (Appendix C),the knockdown of either LmSPN2 or LmSPN3 positively or negatively impacted the expression of Easter protein(Figs.1-E and 2-E), leading us to speculate that LmSPN2/LmSPN3 may instead target another serine protease upstream of Easter in the Toll pathway (Baglinet al.2002;Lawet al.2006; Veillardet al.2016).Therefore, we will focus on finding the intermediate gene/protein between LmSPN2/LmSPN3 and the Toll pathway using more flexible methods (i.e., protein phosphorylation test).

Structural analysis of the serpin protein revealed the‘P1’ target recognition site on the reactive center loop(RCL), with a weak ‘P1-P1’ bond that allows serpin proteins to change conformation to form a stable covalent complex (Belvin and Anderson 1996; Silvermanet al.2001).Three types of interactions have been described between serpins and substrates.In the first type, serpins undergo stable and irreversible covalent bonding with their target substrate, such as in the interaction between SERPINA1 and trypsin inManducasexta(Yeet al.2001; Lawet al.2006).In the second type, serpins can inhibit the activity of substrate proteins, and binding with serpin leads to degradation of the target protein, such as between Serpin18 and Papain in silkworms (Guoet al.2015).In the third type of interaction, serpins confer no inhibitory effects through interaction with their target, as shown in the interaction between Spn76 A and Hsp47 inDrosophilamelanogaster, wherein the serpins function as a transporter or modify the substrate (Garrettet al.2009;Ito and Nagata 2017).Here, our Western blot analysis of diapausing eggs indicated that although LmSPN2/LmSPN3 may directly interact, they do not covalently bind, which suggests that these two proteins likely undergo partly the second or the third type of regulatory interaction (Silvermanet al.2001; Lawet al.2006).In addition, unlike the interaction between Serpin18 and Papain, neither LmSPN2 nor LmSPN3 is targeted for degradation after the interaction.This non-covalent interaction between these two proteins suggests that LmSPN2 may conformationally modify LmSPN3, which inhibits its function as a negative regulator (Garrettet al.2009; Ito and Nagata 2017).Overall, further research is needed to explore the interactions between LmSPN2 and LmSPN3 during embryogenesis, testing their interactioninvivo, and supplement biological experiments leveraging specific antibodies.

Migratory locusts spawn eggs in autumn and then die,leaving the eggs to overwinter.Under natural conditions,a certain proportion of eggs undergo diapause, i.e., the proportion is low in a warm winter but high during a cold winter (Tuet al.2015; Wanget al.2021).Compared to a previous study onLmSPN7(Chenet al.2020a), our study not only detected the interaction between LmSPN2 and LmSPN3 but also explored their potential downstream pathway, which helped understand insect diapause greatly.In this paper, we showed that LmSPN2 promotes diapause by inhibiting the Toll pathway.Conversely,we found that its interaction partner, LmSPN3, initiates development inL.migratoriaby activating the Toll pathway.Moreover, these two LmSPNs likely undergo weak bonding interactions to regulate embryo diapause.These results together show that LmSPN2 drives the initiation of diapause in locust eggs, while LmSPN3 blocks locust eggs from entering diapause.We thus speculate that environmental stimuli of a cold winter result in LmSPN2 inhibition of LmSPN3 to promote diapause,whereas a warm winter results in high expression of LmSPN3 to overcome suppression by LmSPN2 and thus activates Toll-mediated embryonic development(Tuet al.2015).Further exploration of the interactions between LmSPN3 and a truncation mutant of LmSPN2 showed that a glutamate residue at position 331 (E331)in LmSPN2 served as the binding site for interaction between these two proteins.Collectively, these findings provide a mechanistic explanation for the regulation of embryonic diapause inL.migratoria, and the discovery of this interaction site can facilitate the development of pesticides targeting diapause in this notorious insect pest(Denlinger and Armbruster 2016).

5.Conclusion

Our study provided new clues on how SPNs regulate insect diapause.LmSPN2positively regulates egg diapause, whileLmSPN3is a negative regulator.LmSPN2 interacts with LmSPN3 through E331(in LmSPN2), and LmSPN2 plays a relatively dominant role in diapause regulation.They are very likely to function by regulating genes from the Toll pathway.Our results prove the importance of SPNs in insect development and suggest a new target for locust management.

Acknowledgements

We thank members of the Grassland Pest Group from the Institute of Plant Protection, Chinese Academy of Agricultural Sciences, for their discussion.This work was supported by the National Key R&D Program of China(2022YFD1400500), the China Agriculture Research System of MOF and MARA (CARS-34-07), the Publicinterest Scientific Institution Basal Research Fund, China(Y2022GH12) and the Central Public-interest Scientific Institution Basal Research Fund, China (S2021XM22 and S2022XM21).

Declaration of competing interest

The authors declare that they have no conflict of interest.

Appendicesassociated with this paper are available on https://doi.org/10.1016/j.jia.2023.05.019