Overexpression of PGE₂ synthase by in vivo transient expression enhances immunocompetency along with fitness cost in a lepidopteran insect

Prostaglandins (PGs) along with other eicosanoids are oxygenated metabolites of C20 polyunsaturated fatty acids (PUFAs) derived from cellular phospholipids (PLs) by catalytic activity of phospholipase A₂ (PLA₂) (Dennis et al., 2011). Arachidonic acid (AA) is one of the PUFAs and is oxygenated by cyclooxygenase (COX) into PGs (Stanley, 2006). PGs act as dynamic lipid messengers mediating numerous homeostatic biological functions in mammals (Funk, 2001). In insects, they also act as autocrine or paracrine mediators in various physiological processes such as reproduction, secretion and immune responses (Stanley and Kim, 2014).

Among various PGs, prostaglandin E₂ (PGE₂) acts as immune mediator and performs other physiological activities such as smooth muscle dilation and contraction (Smith et al., 1991), body temperature regulation (Milton and Wendlandt, 1971) and the sleep–wake cycle in mammals (Hayaishi, 1991). In invertebrates, PGE₂ signaling mediates immune responses and influences several aspects of oogenesis (Park et al., 2005; Spracklen et al., 2014). In insects, PGE₂ stimulates egg-laying behavior of the cricket Acheta domesticus and the moth Spodoptera exigua (Stanley-Samuelson et al., 1986; Ahmed et al., 2018). Egg development in the bug Rhodnius prolixus is mediated by PGE₂ (Medeiros et al., 2009). In Drosophila melanogaster, PGE₂ mediates oocyte development by stimulating nurse cell-dumping at late oogenesis and coordinating expression of various chorion genes (Tootle et al., 2011).

PGE₂ synthase (PGES) catalyzes isomerization of prostaglandin H₂ (PGH₂) to PGE₂. Recently, we identified a membrane-bound PGES2 in S. exigua (Se-PGES2), and suppression of its expression led to physiological alterations in immunity and reproduction (Ahmed et al., 2018). In mammals, uncontrolled and excessive production of PGs leads to serious abnormalities of physiological processes (Umamaheswaran et al., 2018). In insects, immune defense consists of energy-consuming processes and requires resource allocation via competition for one or more limiting resources, leading to trade-offs between reproductive and immune processes (Schwenke et al., 2016). This led us to pose a hypothesis that uncontrolled production of PGE₂ may result in serious fitness cost in insects.

This study analyzed the change of immune responses by excessive PGE₂ production induced by overexpression of Se-PGES2 in S. exigua. To overexpress Se-PGES2, in vivo transient expression (IVTE) was applied. IVTE technology has been approved to express a foreign gene in S. exigua (Hepat and Kim, 2012). After confirmation of the overexpression of Se-PGES2, changes in both cellular and humoral responses were monitored. In addition, the effect of the overexpression of Se-PGES2 was analyzed in immature development and adult reproduction in S. exigua.

Larvae of Spodoptera exigua (Hübner 1808) were reared on an artificial diet (Goh et al., 1990) at 25±1°C. Under the rearing conditions, larvae underwent five instars (L1–L5). Adults were provided with 10% sucrose solution.

Infective juveniles (IJs) of an entomopathogenic nematode, Steinernema monticolum, were multiplied and harvested according to the method described by Park et al. (1998). Xenorhabdus hominickii ANU101 was isolated from the nematode in a previous study (Park et al., 2017) and cultured in tryptic soy broth (TSB; BD Korea, Seoul, Korea) at 28°C for 24 h with shaking at 150 rpm (Sadekuzzaman et al., 2017). Escherichia coli Top10 (Invitrogen, Carlsbad, CA, USA) was cultured in Luria–Bertani (LB) medium (BD, Franklin Lakes, NJ, USA) overnight at 37°C with shaking at 180 rpm. Heat-killed bacteria were prepared by treatment at 95°C for 10 min.

Arachidonic acid (AA; 5,8,11,14-eicosatetraenoic acid) and prostaglandin E₂ [PGE₂: (5Z,11α,13E,15S)-11,15-dihydroxy-9-oxoprosta-5,13-dienoic acid] were purchased from Sigma-Aldrich (Seoul, Korea) and dissolved in dimethyl sulfoxide. Anticoagulant buffer (ACB) was prepared with 186 mmol l⁻¹ NaCl, 17 mmol l⁻¹ Na₂EDTA and 41 mmol l⁻¹ citric acid, and then adjusted to pH 4.5 with HCl.

A full open reading frame (ORF; 1167 bp) of Se-PGES2 was inserted to a eukaryotic expression vector pIB/V5-His (Invitrogen) by TA cloning. After confirming the insert orientation and DNA sequence, the recombinant vector was transiently expressed in the Sf9 cell line by cationic lipid-mediated transfection using X-treme GENE 9 DNA transfection reagent (Roche, Mannheim, Germany). Sf9 cells (∼5×10⁵ cells ml⁻¹) were seeded into a 75 cm² tissue culture flask (Nunc, Jiangsu, China) in TC100 insect culture medium (Welgene, Daegu, Korea) containing 5% fetal bovine serum. Following the manufacturer's protocol, recombinant vector DNA (3 µg) was mixed with 15 µl of transfection reagent and incubated for 30 min at 25°C. After 30 min of incubation, the DNA lipid complex was added into the flask containing Sf9 cells. After 24 h of cell culture, the medium was replenished with 15 ml of fresh medium without serum. Cells were further incubated for 72 h at 28°C. To extract the proteins from the cultured cells, Sf9 cells were centrifuged at 3000 g for 5 min at 4°C. The pellet was washed with 5 ml of 100 mmol l⁻¹ phosphate-buffered saline (PBS; pH 7.4) and lysed with a lysis buffer containing 50 mmol l⁻¹ Tris HCl, 150 mmol l⁻¹ NaCl, 0.1% Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS, 5 mmol l⁻¹ EDTA and protease inhibitor cocktail (Spruce, St Louis, MO, USA). After 5 min lysis on ice with intermittent mixing, cells were centrifuged at 12,500 g for 5 min. Supernatant containing proteins was quantified using the Bradford (1976) method.

A recombinant vector described above was mixed with a transfection reagent Metafectene PRO (Biontex, Plannegg, Germany) in 1:1 (v/v) ratio and then incubated at 25°C for 30 min to form liposomes to increase transfection efficiency. One microgram of vector was injected into 1-day-old L5 larvae (L5D1) using a microsyringe (Hamilton, Reno, NV, USA) equipped with a 26 gauge needle. To extract the proteins from the larvae, total hemolymph (∼500 µl) from ten L5 individuals was collected from larval proleg, mixed with 500 µl of ACB. Hemolymph was then centrifuged at 3000 g for 5 min at 4°C. Pellet was washed with 5 ml of PBS and lysed with a lysis buffer as described above. After 5 min lysis on ice with intermittent mixing, cells were centrifuged at 12,500 g for 5 min to obtain supernatant proteins.

Extracted proteins (∼100 µg sample⁻¹) were separated on 10% SDS-PAGE. The separated proteins in the gel were transferred onto a 0.2 µm pore nitrocellulose membrane (Bio-Rad Laboratories, Hercules, CA, USA) for 45 min at 100 V in a chilled transfer buffer (25 mmol l⁻¹ Tris, 190 mmol l⁻¹ glycine and 20% methanol, pH 8.5). The membrane was briefly rinsed in Tris-buffered saline with Tween 20 (TBST) (20 mmol l⁻¹ Tris, 150 mmol l⁻¹ NaCl and 0.1% Tween 20, pH 7.5) and then blocked for 1 h at room temperature (RT) in TBST containing 3% bovine serum albumin (BSA). Membranes were then incubated for 2 h at 4°C in Anti-V5 antibody (Invitrogen) as a primary antibody at a dilution of 1:5000 with TBST containing 3% BSA. Membranes were then washed three times in TBST for 5 min per washing and then incubated with anti-mouse IgG-alkaline phosphatase secondary antibody (Sigma-Aldrich, Korea) at a dilution of 1:30,000 in TBST containing 3% BSA for 1 h at RT. Blot was rinsed three times with TBST. To detect alkaline phosphatase activity, nitrocellulose membrane was incubated with a substrate (BICP/NBT, Sigma-Aldrich).

One microliter of recombinant pIB vector containing Se-PGES2 was injected into L5D1 larvae at a dose of 1 µg larva⁻¹. After 24 h post-injection (hpi), IFA was performed. Briefly, total hemolymph (∼150 µl) from five L5 individuals was collected from larval proleg, mixed with 850 µl of ACB and incubated on ice for 30 min. After centrifugation at 180 g for 2 min at 4°C, 700 µl of supernatant was discarded. The cell suspension was gently mixed with 700 µl of TC100 insect tissue culture medium (Welgene). Ten microliters of this hemocyte suspension was placed onto a glass coverslip and incubated in a wet chamber under darkness. Cells were then fixed with 4% paraformaldehyde for 10 min at RT. After washing three times with PBS, cells were permeabilized with 0.2% Triton X-100 in PBS for 2 min at RT. Cells were washed once in PBS and blocked with 5% skim milk in PBS for 10 min at RT. After washing once with PBS, cells were incubated with 5% of Alexa Fluor 555 phalloidin (Invitrogen) and 0.2% fluorescein isothiocyanate (FITC)-tagged Anti-V5 antibody (Invitrogen) individually in PBS for 1 h at RT. After washing three times, cells were incubated with 0.1% of 4′,6-diamidino-2-phenylindole (DAPI) (Thermo Fisher Scientific Korea) in PBS for 5 min. Finally, after washing twice in PBS, cells were observed under a fluorescence microscope (DM2500, Leica, Wetzlar, Germany) at 400× magnification.

In Sf9 cells, cells were collected at 24 h after transfection as described above and centrifuged at 180 g for 2 min at 4°C. Pellet cells were gently mixed with 500 µl of TC100 insect tissue culture medium (Welgene) and the IFA procedure was followed described above.

Total RNA was extracted from hemocytes of ten L5 larvae of S. exigua as an experimental unit. By cutting a proleg, hemolymph was collected in cold ACB and then centrifuged at 800 g for 5 min at 4°C. The resulting hemocyte pellet was used to extract total RNA with Trizol reagent (Invitrogen) according to the manufacturer's instructions. After DNase treatment, 1 µg of total RNA was used to prepare first-strand cDNA synthesized by RT-Premix oligo-dT [5^′-CCAGTGAGCAGAGTGACGAGGACTCGAGCTCAAGCT(16)-3′, Intron Biotechnology, Seoul, Korea] in a reaction volume of 20 µl. The synthesized single-stranded cDNA was used as a template for PCR amplification with 35 rounds of a temperature cycle (95°C for 1 min, 52°C for 1 min and 72°C for 1 min) after an initial heat treatment step at 95°C for 5 min with gene-specific primers (Table S1). The PCR products were separated on 1% agarose gel under 100 V and subsequently stained with ethidium bromide. RT-qPCRs were performed with a qPCR instrument (CFX Connect Real-Time PCR Detection System, Bio-Rad Laboratories) using SYBR Green Realtime PCR Master Mix (Toyobo, Osaka, Japan) according to the general guideline suggested by Bustin et al. (2009). A ribosomal protein, RL32, gene was used as a stably expressed reference gene in S. exigua (Park et al., 2015) for qPCR with gene-specific primers (Table S1). Each cycle was scanned by measuring fluorescence intensity to quantify the PCR products. After the PCR reactions, melting curve analyses were performed from 60 to 95°C to ensure consistency and specificity of the amplified products. Each treatment was replicated three times using independent RNA collections. Quantitative analysis of gene expression was performed using the comparative C_T (2^−ΔΔCT) method (Livak and Schmittgen, 2001).

A nodulation assay was performed at 24 h after IVTE (1 µg larva⁻¹) treatment. Briefly, heat-killed E. coli (∼4.2×10⁴ cells larva⁻¹) was injected in larvae through an abdominal proleg. The treated larvae were then incubated for 8 h at 25°C. After they were kept in 4°C to prevent intentional movement, they were fixed on paraffin plate using insect pins and dissected. The melanized nodules in the hemocoel were counted on gut and fat body under a stereoscopic microscope (Stemi SV11, Zeiss, Jena, Germany) at 50× magnification. Each treatment used 10 larvae. Each treatment was independently replicated three times. All assessments were performed by a blind test.

Recombinant vector (1 µg larva⁻¹) for IVTE or PGE₂ (1 µg larva⁻¹) was injected into the hemocoel of L5 larvae as described above. At 24 hpi for IVTE or 8 hpi for PGE₂, three tissues (hemocytes, fat body and midgut) were isolated and used for RNA extraction. For each replicate, 10 larvae were used for collecting tissues and subsequent RNA extraction. Each treatment used three independent sample collections (n=30 larvae). Eleven AMPs were assessed for their expression levels by RT-qPCR using gene-specific primers (Table S1).

Recombinant vector (1 µg larva⁻¹) for IVTE was injected into the hemocoel of L5 larvae. At 24 hpi, nematode or bacteria was sprayed onto or injected into the larvae, respectively. For the nematode pathogenicity test, 1000 IJ larva⁻¹ of S. monticolum were sprayed onto L5 larva in a Petri dish (5 cm diameter, 3 cm height). For the bacterial pathogenicity test, X. hominickii (∼3.5×10⁴ CFU larva⁻¹) was injected into L5 larvae. Larvae injected with empty vector were used for controls. For the rescue experiment, AA (10 µg larva⁻¹) or PGE₂ (1 µg larva⁻¹) was injected along with the nematode or bacterial treatment. Each treatment used 10 larvae and was independently replicated three times.

One microgram of the recombinant vector for IVTE was injected into larvae (within an hour after emerging into L5). Larvae injected with empty vector were used for controls. Pupal mass was measured on the first day after pupation. Each treatment was replicated three times, and each replicate consisted of 10 insects. A day after adult emergence, treated males and females were kept together for mating. Total numbers of eggs laid were counted for 3 days. Each treatment used five pairs and was replicated three times.

Data from all assays were analyzed by one-way ANOVA using PROC GLM for continuous variables in SAS (SAS Institute, Cary, NC, USA). Percent data were treated by arcsine transformation. All variables were tested against a normal distribution based on the Shapiro–Wilk test using PROC UNIVARIATE. The means were compared using the least squared difference (LSD) test at a Type I error of 0.05.

The cellular locality of Se-PGES2 was assessed by its heterologous expression in Sf9 cells transfected with a eukaryotic expression vector containing Se-PGES2 (Fig. 1). The recombinant Se-PGES2 (rSe-PGES2) was expressed in Sf9 cells and the recombinant protein was confirmed by western blot at the expected size (∼47 kDa) (Fig. 1A). This protein spanned the membrane two times (Fig. 1B). LocTree 3 (https://rostlab.org/services/loctree3/) and LocSigDB (http://genome.unmc.edu/LocSigDB/) predicted cellular locality of Se-PGES2 in the endoplasmic reticulum, Golgi body, mitochondria or lysosome. IFA showed that the rSe-PGES2 protein was localized in the cytosol of Sf9 cells (Fig. 1C).

After 24 hpi of pIB vector containing Se-PGES2 in L5 larvae, its expression was analyzed in hemocytes by IFA (Fig. 2). This IVTE induced rSe-PGES2 expression because rSe-PGES2 proteins were observed in the cytosol of S. exigua hemocytes (Fig. 2A). IVTE was further confirmed by western blot analysis (Fig. 2B). To monitor rSe-PGES2 expression in L5 larvae, RT-PCR primers were designed from its V5 tag to discriminate it from endogenous Se-PGES2 transcripts (Table S1). Total (endogenous+recombinant) Se-PGES2 expression levels significantly (F_1,40=203.90, P<0.05) increased from 6 to 48 hpi whereas the endogenous Se-PGES2 expression levels were not significantly different (Fig. 2C). The enhanced Se-PGES2 expression levels by IVTE treatment were not detected after 72 hpi.

When the IVTE-treated larvae were challenged with bacteria, the total Se-PGES2 expression level was further elevated compared with that of control larvae (pIB) challenged with the bacteria (Fig. 3A). Under these conditions of overexpression of Se-PGES2, a cellular immune response was assessed by counting nodules in response to bacterial challenge (Fig. 3B). Larvae treated with IVTE of rSe-PGES2 formed significantly (F_3,36=189.14, P<0.05) higher numbers of nodules (∼88 nodules larva⁻¹) than control larvae (∼62 nodules larva⁻¹). It was interesting that larvae treated with IVTE of rSe-PGES2 formed nodules (∼21 nodules larva⁻¹) without bacterial challenge.

The influence of PGE₂ injection on expression levels of AMP genes of S. exigua was monitored without immune challenge (Fig. 4A). All 11 AMPs were inducible in their expression in response to PGE₂. However, the inducible AMP genes varied among different tissues. Cecropin expression was significantly (F_1,16=43.88, P<0.05) enhanced in all three tissues. Except gallerimycin and gloverin, nine other AMP genes were induced in their expression levels in fat body. Hemocytes induced seven AMP genes including gallerimycin and gloverin in response to PGE₂. Compared with hemocytes or fat body, the gut exhibited overall lower expression levels (0.01–0.16 in relative mRNA levels).

Similar inducible expression profiles of AMP genes were observed in larvae treated with IVTE of rSE-PGES2 (Fig. 4B). Cecropin expression was also significantly (F_1,16=9.52, P<0.05) enhanced in all three tissues in response to IVTE treatment. Six AMP genes significantly (F_1,34=23.04, P<0.05) increased their expression levels in fat body and five of them were inducible in PGE₂ injection. Five AMP genes (apolipophorin III, attacin 1, attacin 2, gallerimycin and cecropin), which were inducible to PGE₂ injection, were significantly (F_1,28=18.54, P<0.05) enhanced in their expression levels in hemocytes. Three immune-associated genes (defensin, hemolin and cecropin) that were inducible to PGE₂ injection were also induced by IVTE treatment in gut.

Enhanced cellular and humoral immune responses by IVTE of rSe-PGES2 were likely to be advantageous for S. exigua larvae to defend against entomopathogens. To test this hypothesis, two entomopathogens of S. monticolum (Fig. 3C) and X. hominickii (Fig. 3D) were used to infect larvae of S. exigua treated along with IVTE. At nematode or bacterial doses to kill 100% of larvae, an addition of PGE₂ or its biosynthetic precursor (=arachidonic acid) significantly (F_2,27=595.50, P<0.05) rescued the larvae. IVTE treatment also significantly (F_1,18=596.50, P<0.05) rescued the larvae from the nematode or bacterial infection.

Without any immune challenge, overexpressed Se-PGES2 levels resulted in nodule formation and increased AMP gene expression levels, suggesting uncontrolled immune responses. These unnecessary immune responses might lead to adverse effects on larval development and adult reproduction. Indeed, larvae treated with IVTE of rSe-PGES2 exhibited delayed development and led to significantly reduced size of pupae (t=4.11, d.f.=9, P<0.05; Fig. 5A). Adults emerged from the pupae exhibited a significantly lower number of laid eggs (t=7.04, d.f.=9, P<0.05; Fig. 5B).

Various physiological processes are mediated by PGE₂ in insects (Stanley and Kim, 2019). Thus, biosynthesis of PGE₂ by Se-PGES2 plays a crucial role in immunity and reproduction of S. exigua (Ahmed et al., 2018). The results of the present study support its significance in mediating both cellular and humoral immune responses, and also suggest that a tight regulation of Se-PGES2 expression is necessary to avoid excess immune responses.

Heterologous expression of rSe-PGES2 in Sf9 cells and its IVTE in S. exigua hemocytes indicated that Se-PGES2 is localized in cytoplasm. In addition, bioinformatics analysis suggests that it is localized in the endoplasmic reticulum. To synthesize PGE₂ in mammals, PLA₂ catalyzes PLs to release AA, which is then oxygenated by COX to produce PGH₂ (Burke and Dennis, 2009; Stanley, 2006; Park et al., 2014). In mammals, cytosolic or membrane-bound PGES isomerizes PGH₂ to PGE₂. Insects also have been considered to have all biosynthetic machinery to synthesize PGE₂, as predicted in previous bioinformatics analysis of Drosophila melanogaster (Scarpati et al., 2019). Indeed, three different PLA₂s have been identified in S. exigua and include one secretory PLA₂ and two cellular PLA₂s. Also, COX-like peroxynectins have been identified and play a role in oxygenating AA (Park et al., 2014). Thus, along with identification of Se-PGES2 (Ahmed et al., 2018), a biosynthetic pathway of PGE₂ has been identified in S. exigua. In that study, Se-PGES2 was predicted to be a membrane-bound type among three different types of PGESs. The present study, using the heterologous expression in Sf9 cells, supports the cytosolic locality, at which Se-PGES2 performs its PGE₂ biosynthesis.

IVTE of rSe-PGES2 upregulated total Se-PGES2 expression levels for at least 48 hpi in S. exigua. Transient expression of a foreign gene has been usually performed by transfection of recombinant genes in an expression vector to cell lines via liposome formation to enhance DNA entry to target cells (Felgner et al., 1987). IVTE, which has been done by direct transfection to post-embryonic tissues, was also successfully used to deliver DNA into larval tissues of Bombyx mori (Xiao et al., 2002). Indeed, Hepat and Kim (2012) tested the applicability of the IVTE technique in S. exigua using a recombinant pIB vector containing enhanced green fluorescence protein (EGFP) and showed that EGFP expression occurred mostly in hemocytes and fat body and persisted with the dependency of expression levels on injection dose of the vector.

Overexpression of Se-PGES2 induced by IVTE enhanced both cellular and humoral immune responses. The enhanced immunocompetency via IVTE of Se-PGES2 was supported by the effective defense against infection of entomopathogens. Our previous study showed that Se-PGES2 is associated with cellular immune response presumably by enhancing PGE₂ biosynthesis (Ahmed et al., 2018). The additional PGE₂ might facilitate the cellular and humoral immune responses, which would explain the enhanced survivorship against pathogen infection. Miller (2005) showed that hemocyte-spreading behavior increases in a dose-dependent manner of the amount of treated PGE₂ in Manduca sexta larvae. The role of PGE₂ in hemocyte spreading may be induced by activating actin remodeling by the growth of F-actin via activation of the actin-bundling protein Fascin (Groen et al., 2012). PGE₂ also increased AMP production in a dose-dependent manner in S. exigua (Shrestha and Kim, 2009). Toll and IMD immune signaling pathways are identified in S. exigua (Hwang et al., 2013). A Toll signaling component, Pelle, activates PLA₂ activity, presumably to increase PGE₂ production (Shafeeq et al., 2018). Yajima et al. (2003) showed that the PLA₂ activity is associated with IMD immune signaling to produce AMPs in Drosophila. Furthermore, Zhang et al. (2018) showed that PGs rescued Spodoptera frugiperda larvae infected with an entomopathogen, Serratia marcescens.

Overexpression of Se-PGES2 induced by IVTE had an adverse effect on development and reproduction of S. exigua. The uncontrolled expression of Se-PGES2 may lead to untimely production of PGE₂. The excess PGE₂ may divert the nutrient usage from development and reproduction to energy-consuming immune responses even without pathogen infection, which resulted in fitness costs such as reduction in pupal body mass and adult fecundity. In the honey bee, immune challenge reduces expression and accumulation of stored plasma proteins such as vitellogenin and hexamerin, whereas it highly induces immune-associated genes, defensin and prophenoloxidase, suggesting an adaptive strategy to redirect resources to combat injury or infection (Lourenço et al., 2009). In S. frugiperda, in which PGs played a crucial role in mediating cellular immune response, the PG-treated larvae suffered reduced survivorship compared with controls (Zhang et al., 2018), indicating a fitness cost. Thus, PGE₂ is crucial for S. exigua survivability, but the excessive amount may impose to reduced pupal mass and adult fecundity owing to a trade-off effect.

We thank Youngim Song for supplying materials and Al Baki Abdullah for technical guidance on the micro-injection of plasmids into young insect larvae.