Larvae of two lepidopteran species (S. exigua and Plutella xylostella) were collected from Welsh onion (Allium fistulosum L.) and Chinese cabbage (Vigna angularis) fields in Andong, Korea, respectively. Coleopteran (Tenebrio molitor) larvae were provided by Bio Utility (Andong, Korea). Insects were reared in a laboratory under our standard conditions of 25 ± 2°C temperature, 16:8 h (L:D) photoperiod, and 60 ± 5% relative humidity (RH). S. exigua larvae were reared on an artificial diet (28) and underwent five larval instars (L1-L5). Larvae of P. xylostella were reared on fresh cabbage leaves and underwent four larval instars (L1-L4). Approximately 3.5 cm body length T. molitor larvae were used for pathogenicity testing.

Arachidonic acid (AA: 5,8,11,14-eicosatetraenoic acid), dexamethasone [DEX: (11β, 16α)-9-fluoro-11,17,21-trihydroxy-16-methylpregna-1,4-diene-3], naproxen [NAP: (S)-(+)-2-(6-methoxy-2-naphthyl)propionic acid], and esculetin (ESC: 6-hydroxy-7-methoxycoumarin) were purchased from Sigma-Aldrich Korea (Seoul, Korea) and dissolved in dimethyl sulfoxide (DMSO). Fura-8AM was purchased from AAT Bioquest (Sunnyvale, CA, USA) and dissolved in DMSO. Thromboxane B₂ (TXB₂: 9α,11,15S-trihydroxythromba-5Z,13E-dien-1-oic acid) and thromboxane A₂ (TXA₂: 9α,11α-methylene-15S-hydroxy-11α-deoxy-11α-methylene-thromba-5Z,13E-dien-1-oic acid) were purchased from Cayman Chemicals (Ann Arbor, MI, USA). DAN (dantrolene sodium: 1-[(5-(p-nitrophenyl) furfurylidene)amino] hydantoin sodium salt) were purchased from Sigma-Aldrich Seoul, Seoul, Korea. Prostaglandin E₂ (PGE₂: (5Z,11α,13E,15S)-11,15-dihydroxy-9-oxoprosta-5,13-dienoic acid), prostaglandin D₂ (PGD₂: 9α,15S-dihydroxy-11-oxoprosta-5Z,13E-dien-1-oic acid), prostaglandin I₂ (PGI₂: 6,9α-epoxy-11α,15S-dihydroxy-prosta-5Z,13E-dien-1-oic acid), 2-aminoethoxydiphenylborate (2-APB), thapsigargin (TPG), U-73122 (1-[6-[((17β)-3-methoxyestra-1,3,5[10]-trien-17-yl)amino]hexyl]-1H-pyrrole-2,5-dione), terutroban (TTB), and dazoxiben (DAZ) were purchased from Sigma-Aldrich Korea and dissolved in DMSO. Phosphate-buffered saline (PBS) was prepared with 100 mM phosphoric acid and adjusted to pH 7.4. Anticoagulant buffer (ACB) was prepared with 186 mM NaCl, 17 mM Na₂EDTA, and 41 mM citric acid.

Based on GPCR sequences of Bombyx mori (GenBank accession number: NP_001037033.1), S. litura (XP_022813850, XP_022814039, XP_022816105, XP_022817498, XP_022817499, XP_022821197, XP_022822728, XP_022822897, XP_022823006, XP_022826743, XP_022827237, XP_022827354, XP_022827490, XP_022827519, XP_022827626, XP_022834024, XP_022834058, XP_022834098, XP_022838045), Heliothis virescens (CDF56917.1), and Trichoplusia ni (XP_026733990), their orthologs of S. exigua were inferred from Transcriptome Shotgun Assembly (GGRZ01048721.1, GGRZ01057920.1, GGRZ01138956.1, GGRZ01102221.1, GGRZ01084708.1, GGRZ01092489.1, GGRZ01087211.1, GARL01043383.1, GGRZ01038487.1, GGRZ01037013.1, GARL01065745.1, GARL01040814.1, GGRZ01082913.1, GGRZ01068498.1, GGRZ01117535.1, GGRZ01069701.1, GGRZ01224538.1, GGRZ01247645.1, GGRZ01215736.1, GGRZ01095571.1, GGRZ01137765.1, GGRZ01106723.1, GGRZ01113408.1, GGRZ01069261.1) using the BlastN program (http://www.ncbi.nlm.nih.gov). The resulting sequences were subjected to open reading frame (ORF) analysis by using ORFfinder (https://www.ncbi.nlm.nih.gov/orffinder/). The ORF sequences were deposited to GenBank with accession numbers (AEO27700.1, AXY04240.1, AXY04245.1, AXY04246.1, AXY04251.1, AXY04254.1, AXY04275.1, AXY04297.1, AXY04299.1, AZA07970.1). Alternatively, the protein coding sequences of the GPCRs were inferred by mapping the protein sequences of the GPCR sequences at NCBI and newly identified ORF sequences against a reference whole genome assembly in S. exigua (GCA_011316535) using exonerate-2.2.0 with protein2genome model. Phylogenetic analysis was performed with the Neighbor-Joining method and a Poisson correction model using MEGA6.06 software (www.megasoftware.net). Bootstrapping values were obtained with 1,000 replications to test supports on each node in the resulting phylogenetic tree.

Larvae of S. exigua infesting Welsh onion in Andong, Korea suffered from green muscardine disease symptoms and were collected for diagnosis. Fungal spores were collected in tubes containing sterile water by scraping from infected insect larvae. After stirring, 250 µL of the fungal suspension was spread on potato dextrose agar (PDA) and incubated at 25 ± 1°C and 70 ± 5% RH. The cultured spores were sub-cultured on PDA.

The cultured fungal colonies were transferred onto slides with PVA mounting medium (PVA MTNG) (BioQuip Products, Gladwick Street, CA, USA) and incubated at 50°C for 48 h. The slides were observed under an optical microscope (DM500, Leica, Wetzlar, Germany) with 400× magnification. Fungal samples were collected from PDA cultured for 14 days and subjected to Au‐coating using a Sputter Coating machine (Jeol Korea, Seoul, Korea). The treated samples were observed under a scanning electron microscope (SEM; JSM‐6300, Jeol Korea) at 200 × magnification. The resulting morphological characters were used for the identification of the entomopathogenic fungi according to a taxonomy of key characters such as colony, conidiophore, and conidial shapes (29).

Genomic DNA (gDNA) was extracted from a mixture of isolated spores and hyphae with phenol extraction and ethanol precipitation, as described by Park and Kim (17). Internal transcribed spacer (ITS) region was amplified using ITS forward primer (5’-TTGATTACGTCCCTGCCCTTT-3’) and ITS reverse primer (5’-TTTCACTCGCCGTTACTAAGG-3’) as reported by Vrain et al. (30). For PCR amplification of the ITS region, the extracted gDNA was used as a template with 35 cycles under the following conditions: 1 min at 94°C for denaturation, 1 min at 46°C for annealing, and 1 min at 72°C for extension. A second PCR was conducted using M13 universal primer-conjugated primers and sequenced by Macrogen (Seoul, Korea). The obtained nucleotide sequence was analyzed using the BlastN program of the National Center for Biotechnology Information (NCBI, www.ncbi.nlm.nih.gov). The evolutionary relationship was inferred using a Neighbor-Joining phylogenetic tree with MEGA6.06 (31). Bootstrap values on the branches were estimated with 1,000 replications.

A conidial suspension of M. rileyi was prepared by scraping the fungal culture into 1.5 mL tubes containing autoclaved Triton X-100 (0.1%) solution (Duksan Pure Chemicals, Ansan, Korea). After disentangling the conidial clumps by vortexing for 3 min, conidia in the suspension were counted using a Neubauer hemocytometer (Marienfeld-Superior, Lauda-Königshofen, Germany) under 40 × magnification.

The isolated fungi were fed to L3 larvae of S. exigua or P. xylostella, or 3.5 cm-length T. molitor larvae. A piece (1 × 1 cm) of cabbage containing 1,000 conidia was completely consumed by each larva. Each treatment was replicated 3 times and each replication consisted of 10 insects, in which each Petri dish held a single insect. The Petri dishes were kept in desiccators (4202-0000, Bel-Art Products, Pequannock, NJ, USA) maintaining a constant temperature of 25 ± 1°C and 75 ± 5% RH, following Winston and Bates (32) using a saturated solution of NaCl. Dead insects were counted every 24 h up to 7 days. For topical or injection application, 1,000 conidia per larva were used. Newly (< 30 min) molted fourth instar larvae were considered ‘unsclerotized’ while the larvae at 1 day post molting were considered ‘sclerotized’ cuticles. To remove the epicuticle layer of the integument, the abdominal tergum was rubbed with a brush soaked in acetone for 2 min.

To observe the Ca²⁺ signal and hemocyte aggregation behavior in response to fungal challenges (33), L5 larvae of S. exigua were injected with 2 µL Fura-8 (1 mM) and incubated for 30 min. 2 µL of fungal conidia and hyphae were injected. At selected time points post-injection (PI), hemolymph was collected and fixed on a slide glass by using 2.5% paraformaldehyde. The Fura-positive cells were observed under a fluorescence microscope (DM500, Leica, Wetzlar, Germany) at 400 × magnifications. Fluorescence change over time was analyzed using ImageJ software (https://imagej.nih.gov/ij). Each time point was replicated three times.

To inhibit Ca²⁺ flux, L5 larvae were co-injected with 1 µL of DAN (1 µg/µL), 2-APB (1 µg/µL), U-73122 (1 µg/µL), TPG (1 µg/µL), TTB (1 µg/µL), or DAZ (1 µg/µL) along with Fura-8 (1 mM). After 30 min, 2 µL of fungal conidia and hyphae were injected with TXB₂ (1 µg). After another 10 min, hemolymph was bled and the hemocytes were fixed on a slide glass by using 2.5% paraformaldehyde. Aggregated hemocyte percentage and Fura-8-positive cells were determined as described above.

M. rileyi conidia were obtained from the 10-day culture on PDA medium. The conidia were resuspended in a sterile bicarbonate buffer (9.5 mL of 0.2 M Na₂CO₃ mixed with 41.5 mL of 0.2 M NaHCO₃, pH 9.4). Then 1 µL of 10 mg/mL FITC fluorescein isothiocyanate (FITC; Sigma-Aldrich Korea) was added to the fungal pellet and placed on a shaker (170 rpm) for 30 min at room temperature under darkness (34). The conidia were washed four times with ice-cold PBS containing 0.02% EDTA and resuspended with 1 mL of TC100 insect medium (Welgen, Gyeongsan, Korea) and stored at -20°C.

For the phagocytosis assay, L5 larvae were surface-sterilized with 70% ethanol and injected with 2 µL of FITC-labeled conidia. At 30, 60, and 90 min PI, the larvae were bled through a clipped proleg onto a slide glass containing 5 µL of ACB. Phagocytic rates were determined by assessing the ratio of 100 hemocytes with or without ingested conidia under a fluorescence microscope (Leica, Wetzlar, Germany).

After surface sterilization with 70% ethanol, L5 larvae were anesthetized with ice and injected with 2 µL of conidial suspension (1×10⁵ conidia/mL) through the inter-segmental membrane between the last two abdominal segments using a Hamilton micro-syringe (Reno, NV, USA) equipped with a 26-gauge needle. After 8 h of incubation at room temperature, the hemocytic nodules were counted by dissecting the larvae under a stereomicroscope (Stemi SV 11, Zeiss, Jena, Germany) at 50 × magnifications. To assess the effects of DEX on nodule formation in larvae infected with M. rileyi, larvae were co-injected with 10 µg of DEX along with the conidia. For the time-course experiment, the number of nodules formed was counted at selected time points (1~8 h) PI, 3 larvae/time point.

Hemocyte aggregation is functionally defined as a clump of at least four hemocytes attached to conidia or hyphae. Aggregation activity was determined as the ratio of the aggregated hemocytes among 100 randomly chosen cells. To determine the inhibitory effects of eicosanoid biosynthesis inhibitors, each S. exigua L5 larva was injected with 1 µL of DEX (10 µg), ESC (1 µg), or NAP (1 µg) along with 2 µL of Fura-8 (1 mM). For the rescue experiment, 1 µL of AA (10 µg) was co-injected with DEX and 1 µL of PGE₂ (1 µg) was co-injected with NAP. At 30 min PI, 2 µL of fungal conidia and hyphae were injected into individual insects. To determine the effect of eicosanoids on cellular immune reactions, individual S. exigua (L5) larvae were co-injected with 1 µL of DEX (10 µg) and 2 µL of Fura-8 (1 mM). At 30 min PI, the insects were injected again with 2 µL (1 µg) of prostaglandin D₂, prostaglandin E₂, prostaglandin I₂, thromboxane A₂, or thromboxane B₂ along with fungal conidia and hyphae. At 10 min PI, hemolymph was collected and microscopic slides were prepared for assessing Fura-positive hemocyte and aggregated hemocytes.

RNA extraction and cDNA preparation followed the method described by Vatanparast et al. (35). RT-PCR of Se-TXAS and SePGE₂R genes was conducted using DNA Taq polymerase (GeneALL, Seoul, Korea) with an initial heat treatment at 94°C for 5 min, followed by 35 cycles of DNA denaturation at 94°C for 30 s, primer annealing at 50°C for 30 s ( Table S4 ), and extension at 72°C for 30 s. The PCR reaction was completed with a final chain extension step at 72°C for 10 min. qPCR was conducted on a Real-time PCR thermal cycler (Step One Plus Real-Time PCR System, Applied Biosystems, Singapore) using Power SYBR Green PCR Master Mix (Life Technologies, Carlsbad, CA, USA) according to the guidelines of Bustin et al. (36). The reaction mixture (20 µL) contained 10 µL of PCR Master Mix, 5 µL sterile water, 3 µL of cDNA template (50 ng), and 1 µL each of forward and reverse primers ( Table S4 ). The temperature program for qPCR began with 95°C heat treatment for 10 min followed by 40 cycles of denaturation at 94°C for 30 s, annealing at 50°C for 30 s, and extension at 72°C for 30 s. The expression level of the ribosomal gene, L32, was used as a reference gene to normalize target gene expression levels. Quantitative analysis was performed using the comparative CT (2^-ΔΔCT) method (37).

For RNAi, double-stranded RNAs (dsRNAs) encoding TXA₂ synthase (dsTXAS), SePGE₂R (dsPGE₂R), and green fluorescence protein (dsCON) were prepared as described by Vatanparast et al. (38) using Megascript RNAi Kit (Ambion, Austin, TX, USA). dsRNAs were mixed with a transfection reagent Metafectene PRO (Biontex, Plannegg, Germany) at a 1:1 (v/v) ratio and incubated at room temperature for 30 min to form liposomes. dsRNA (1 µg) was injected into L5 larval hemocoel with a microsyringe. The RNAi efficiency was evaluated by RT-qPCR at the selected time points. Each treatment was replicated three times using independent RNA preparations.

Secretory PLA₂ (sPLA₂) activity in S. exigua larval plasma was determined at three different time points during incubation at 25°C using a commercial assay kit (sPLA₂ Assay Kit, Cayman Chemical, Ann Arbor, MI, USA) with diheptanoyl thio-phosphatidyl choline as the enzyme substrate following Vatanparast et al. (38). Cellular PLA₂ (cPLA₂) activity measurement in fat body preparations used the same kit, with a different substrate, arachidonyl thio-phosphatidyl choline. A spectrofluorometer (VICTOR multi-label Plate reader, PerkinElmer, Waltham, MA, USA) was used to measure enzyme activity. Each treatment was replicated with three biologically independent enzyme preparations. Specific enzyme activity (µmol/min/µg) was calculated by dividing absorbance change by the amount of total protein. Protein concentrations were determined following Bradford (39).

Hemolymph was collected from L5 larvae into ACB by cutting an abdominal proleg and aspirating the exuded hemolymph with glass capillaries (TW100-4, World Precision Instrument, Sarasota, FL, USA). Hemocytes were counted with a hemocytometer (Neubauer improved bright line, Superior Marienfield, Lauda-Königshofen, Germany) under a phase contrast microscope (BX41, Olympus, Tokyo, Japan). Each treatment was independently replicated three times. To evaluate the effect of TXB₂ on THC, test larvae were injected with TXB₂ (1 µg per larva). Hemolymph was collected at each min after TXB₂ treatment up to 10 min and assessed for THC as described.

Micro-aggregation is defined as the aggregation of four or more hemocytes. L5 larvae were injected with different PGs and bled at different time points. The hemolymph (20 µL) was mixed with 10 µL of TC100 insect cell culture medium on a cavity well microscope slide (Globe Scientific, Mahwah, NJ, USA) to observe micro-aggregation behavior. Each treatment was replicated three times. To observe hemocyte migration on glass slides, 1 µL of TXB₂ (1 µg) was added to 10 µL of hemocyte suspension obtained from naive larvae. The hemocyte behavior was monitored by fluorescence microscopy (Leica, Wetzlar, Germany) at 200×.

Larvae immunity was challenged with 1,000 conidia injected into hemocoels of L5 larvae and incubated at 25°C for 16 h. Control larvae were injected with sterile PBS. Fat body samples were collected into 15 mL tubes from 70 larvae and washed with cold (4°C) PBS. Each sample was homogenized three times (10 min per cycle) in PBS with an ultrasonicator (Bandelin Sonoplus, Berlin, Germany) at 75% power, and subsequently adjusted to pH 4.0 using 1 N HCl. Prostanoids were extracted with 500 μL of ethyl acetate. The combined ethyl acetate extracts were dried under nitrogen to approximately 50 μL and applied to a small silicic acid column (2 × 90 mm containing 30 mg of Type 60A, 100-200 mesh silicic acid, Sigma-Aldrich Korea). Extracts were sequentially eluted with 300 μL of polar solvents starting with 100% ethyl acetate, followed by ethyl acetate: acetonitrile (1:1, v:v), 100% acetonitrile, acetonitrile: methanol (1:1, v:v), and 100% methanol. The acetonitrile:methanol fraction was used to quantify TXB₂. Each treatment was replicated three times using independent sample collections.

LC-MS/MS was performed using a QTrap 4500 (AB Sciex, Framingham, MA, USA) equipped with an autosampler, a binary pump, and a column oven. The analytical column was an Osaka Soda C18 column (2.1 mm × 150 mm, 2.7 μm) maintained at 40°C (Osaka, Japan). The mobile phases consisted of 0.1% formic acid in water (A) and 0.1% formic acid in acetonitrile (B). The linear gradient was as follows: 30% B at 0 min, 30% B at 2 min, 65% B at 12 min, 95% B at 12.5 min, 95% B at 25.0 min, 30% B at 28.0 min, and 30% B at 30 min. The flow rate was 0.40 mL/min. The autosampler was set at 5°C and the injection volume was 10 μL. The LC-MS/MS was equipped with an electrospray ionization (ESI) source. ESI was performed in negative ion mode. After optimization, the source parameters were: source temperature at 400°C, curtain gas flow rate at 32 µL/min, ion source gas flow rate at 60 µL/min, and spray voltage adjusted to −4,500 V. Analyses were performed in Multiple Reaction Monitoring (MRM) detection mode using nitrogen as the collision gas. Peak detection, integration, and quantitative analysis were done using MassView1.1 software (AB Sciex).

The data for continuous dependent variables were subjected to one-way analysis of variance (ANOVA) using PROG GLM in the SAS program (40). Mortality assay by the leaf dipping method was analyzed by repeated measure ANOVA. All experiments were conducted in three biologically independent replicates and the means ± standard errors (SE) were plotted using Sigma Plot (Systat Software, Point Richmond, CA, USA). The means were compared with the least significant difference (LSD) test at a Type I error of 0.05.

Fungal spore colonies on the PDA medium began to grow slowly with a white velvety appearance (early stage) and irregular borders at 4 days, then turned to pale green or malachite green with sporulation at 7 days ( Figure S1A ). The vegetative and reproductive structures developed hyaline hyphae and conidiophores with smooth and septate walls. The branches, which formed near the septa, developed in clusters on the same point with 2~4 phalids exhibiting a short, rounded, and thickened base. Conidia were smooth and ellipsoidal, and arranged in chains.

The ITS region was amplified and sequenced to identify species in the fungal isolate ( Figure S1B ). In total, 734 bp were sequenced, including full-length 5.8S rRNA, ITS-1, and ITS-2 regions as well as partial sequences of 18S and 28S rRNAs. The sequences showed high similarities (>99%) with known ITS sequences of several M. rileyi strains ( Table S1 ). To evaluate the relationship of the isolate with other fungal strains, a phylogenetic tree was constructed using the ITS sequences ( Figure S1C ). The isolate was significantly (bootstrap value = 100%) clustered with another M. rileyi isolate.

The fungal isolate was pathogenic to tested insects, with considerable variation in the fungal virulence ( Table S2 ). The lethal median time (LT₅₀) was much faster against S. exigua (111.41 h) than T. molitor (182.65 h) or P. xylostella (136.35 h). The insecticidal activities of the fungal isolate were also more potent in S. exigua (76.7%) compared to other insects under the same fungal treatment at 7 days after treatment ( Figure S1D ).

Fungal virulence to S. exigua larvae differed by infection route. Topical application led to 40.0% mortality, which increased to 64% following spore injections ( Figure 1A ). M. rileyi virulence increased in newly molted larvae, before cuticular sclerotization ( Figure S2A ) or to integument treated with an organic solvent to remove the epicuticular layer ( Figure S2B ). Co-injections of spores with DEX led to substantially increased mortality ( Figure 1A ). The DEX treatment led to significantly suppressed phagocytosis ( Figure 1B ) and hemocyte nodule formation following fungal infection ( Figure 1C ). sPLA₂ and iPLA₂ activities significantly increased within 5 min post-fungal infection (PFI; Figure 1D ), with parallel increases in mRNAs encoding sPLA₂ and iPLA₂B, but not iPLA₂A ( Figure 1E ).

Hemolymph collected from M. rileyi-challenged larvae had labeled spores and hyphae connected to hemocytes ( Figure 2A ). Proportions of Fura-positive hemocytes increased with time, up to 10 min PFI, in correlation (r = 0.943; P < 0.0001) with numbers of hemocyte aggregates ( Figure 2B ). Fura-staining intensity also increased with time up to 10 min. Treating S. exigua larvae with the PLA₂ inhibitor, DEX, led to reduced proportions of Fura-positive hemocytes and aggregated hemocytes ( Figure 2C ).

Treating larvae with a nonspecific cyclooxygenase inhibitor, NAP, prior to the conidial challenge led to substantially suppressed proportions of Fura-positive hemocytes (down by about 40%) and hemocyte microaggregates (down by about 35%), both of which were reversed in larvae co-treated with NAP+PGE₂ ( Figure 3A ). Treatment with the lipoxygenase inhibitor ESC led to small but statistically significant reductions in both parameters, with Fura-positive proportions down by about 8% and aggregated hemocytes down by about 24%. Treating larvae with DEX led to reductions in proportions of Fura-positive hemocytes (down by about 20%) and aggregated hemocytes (down by about 50%). Figure 3B shows the influences of selected prostanoids on possible reversals of DEX treatments before the conidial challenge. Co-injecting DEX along with one of three PGs led to increases in both parameters ( Figures S3A–E ). PGE₂ treatment led to increased Fura-positive hemocyte proportions by about 56%; PGD₂ treatment increased by about 50%; PGI₂ treatment did not reverse the DEX effect. TXA₂. TXB₂ treatments led to very high proportions of Fura-positive hemocytes, up to just over 60% for TXA₂ and nearly 80% for TXB₂. Parallel experiments with proportions of aggregated hemocytes led to similar results. Figure 3C shows the outcomes of treating larvae with specific inhibitors of thromboxane synthesis, DAZ, and a thromboxane receptor antagonist, TTB. Compared to controls, co-injecting conidia+DAZ and, in a separate group of experimental larvae, conidia+TTB led to significant decreases in proportions of Fura-positive hemocytes (down by about 10% after DAZ treatment and by about 20% after TTB treatment) and aggregated hemocytes (down by about 20% after DAZ treatment and by about 35% after TTB treatment).

Figure 4A outlines the biosynthesis of TXA₂/TXB₂ from AA in S. exigua. Two POXs, SePOX-F and SePOX-H, convert AA into PGH₂, which is converted into TXA₂. TXA₂ can convert into TXB₂ spontaneously. Figure 4B shows that the M. rileyi challenge did not influence the time-dependent accumulation of mRNAs encoding SePOX-A (upper left), while accumulations of mRNAs encoding SePOX-F and SePOX-H were significantly elevated beginning 16 h PFI (upper right and lower left panels). Figure 4B (lower right) shows that the accumulation of mRNAs encoding Se-TXAS, the enzyme that converts AA into TXA₂, was also significantly elevated at 16 h PFI.

We estimated concentrations of TXB₂ in fat body (L5 larvae) by LC-MS/MS. Chemical identification of TXB₂ was confirmed by two specific ion peaks, which matched the ion peaks in an authentic TXB₂ chemical standard ( Figure S4 ). TXB₂ was detected at 0.28 ± 0.02 ng/g in naïve larvae, which was significantly increased to 0.61 ± 0.05 ng/g in M. rileyi-challenged larvae ( Figure 5A ). Se-TXAS expression was silenced by injecting dsRNA specific to the gene into experimental larvae ( Figure 5B ). Injecting M. rileyi into dsRNA-treated larvae led to significantly enhanced pathogenicity, which was reversed in larvae co-injected with conidia+TXB₂ ( Figure 5C ).

Figure 6A reports the influence of TXB₂ injections (1 μg/larva) on total hemocyte numbers over 10 h PI. The blue line represents DMSO-treated control larvae and the red line shows total hemocyte counts increase over time PI. Figure 6B shows reports that DEX injections led to no visible changes in DIC, nor Ca²⁺ signaling. Co-injecting DEX+TXB₂ visible DIC and Fura staining. Figure 6C shows that separate treatments with the indicated prostanoids led to substantial increases in proportions of microaggregated hemocytes with the exception of PGI₂ treatments, which did not influence microaggregation. The thromboxane injections led to statistically significant higher proportions of microaggregated hemocytes, compared to controls and the three PG treatments. Figure 6D reports proportions of microaggregated hemocytes steadily increase as a function of time over 60 min after TXB₂ injections to a maximal level at about 60% of total hemocytes. This stimulatory activity of TXB₂ exhibited a dose-dependency ( Figure 6E ).

Within the S. exigua genome there are 37 predicted GPCR candidates ( Table S3 ), which were used to be compared with mammalian PG receptors. In particular, Figure 7A shows that a S. exigua PGE₂ receptor (Se-PGE₂R) is clustered with human, house mouse, and zebrafish thromboxane receptors. Injecting a dsRNA construct designed to the SePGE₂R sequence led to severe reductions in mRNAs encoding Se-PGE₂R, down by about 80% ( Figure 7B ). Figure 7C shows that silencing Se-PGE₂R led to reduced accumulations of mRNAs encoding PGE₂R and reduced Ca²⁺ signaling, seen in microphotographs as severe reductions in Fura-positive hemocytes and the accompanying histogram. Co-injections of dsPGE₂R+TXA₂ and, separately, dsPGE₂R+TXB₂, did not restore FURA staining. Similar images and the accompanying histogram document the same outcomes with proportions of microaggregated hemocytes.

The signaling pathway to up-regulate Ca²⁺ signal in hemocytes in response to TXB₂ was monitored by treating larvae with selected compounds that influence Ca²⁺ signaling in mammalian cells ( Figure 8 ). The micrographs in Figure 8A show that phospholipase C inhibitor (‘U-73122’), IP₃ receptor inhibitor (2-APB), and ryanodine receptor inhibitor (‘DAN’) significantly suppressed Ca²⁺ signaling and hemocyte aggregation following the M. rileyi challenge. The accompanying histograms show the quantitative values. In contrast, a sarco/endoplasmic reticulum Ca²⁺-ATPase (SERCA) inhibitor (‘TPG’) did not suppress the Ca²⁺ signal. The micrographs in Figure 8B and their accompanying histograms similarly show that co-injections with TXB₂ did not reverse the influences of the drug treatments.