Ecdysone Titer Determined by 3DE-3β-Reductase Enhances the Immune Response in the Silkworm Free

Insects use a highly effective immune system to defend themselves against foreign microorganisms, and this powerful system contributes to the successful evolution of insects (1). Unlike vertebrates, which have acquired immunity, insects only contain an innate immune system, including humoral and cellular reactions. Pathogen infection rapidly and strongly initiates the insect innate immune response. However, more and more studies showed that several immune-related genes can express without the stimulation of microorganisms, suggesting that other factors besides pathogen infection or injury could also regulate the insect immune response (2). Indeed, the Drosophila transcription factor Forkhead box O (dFOXO) in the insulin/insulin-like growth factor signaling pathway could regulate the expressions of the antimicrobial peptides (AMPs) under normal physiological conditions (3). The target of rapamycin cascade, another nutrient-dependent pathway in Drosophila, also has the potential impact on the innate immunity and the transcriptions of AMPs (4–6).

Indeed, the steroid hormone 20-hydroxyecdysone (20E), one of the two most important regulators of development, metamorphosis, and reproduction in insects, can also stimulate the insect innate immune response (7, 8). For example, 20E could promote the expressions of immune-related genes, especially AMP genes, in infected culture cells and animals (9–14). Furthermore, 20E controls the humoral innate immune response through two different ways in Drosophila. First, 20E upregulates the transcription of the peptidoglycan recognition protein (PGRP)-LC, and then induces all AMP genes through the immune deficiency (IMD) pathway; second, 20E regulates the expression of the AMPs, such as Diptericin, Metchnikowin, and Drosomycin, by the ecdysteroid-related transcription factors independent of PGRP-LC (15). Moreover, 20E also promotes insect cellular immunity. The expressions of the prophenoloxidases could be induced in the Anopheles gambiae cell line 4a-3B (16). Injection with 20E increases the phagocytic activity of Drosophila plasmatocytes (17), and 20E signaling also enhances hemocyte motility, encapsulation, and nodulation (18–21).

However, how insects regulate 20E titer to affect immunity after pathogen attack remains unknown. In insects, the biosynthesis of the 20E is mainly from the cholesterol pathway in the prothoracic glands (8). Nevertheless, ecdysone can also be directly synthesized from 3-dehydroecdysone (3DE) in other tissues (22–24). In immature stages of most lepidopteran species, 3DE is the major product of the prothoracic glands. After release from the glands, 3DE is rapidly reduced to ecdysone by 3DE-3β-reductase in hemolymph, and then the ecdysone is converted into 20E by 20-hydroxylase in the target tissues (22–26). 3DE-3β-reductase is considered to be an important enzyme in the biosynthesis of the ecdysone (27). A previous study demonstrated that Trichoplusia ni 3DE-3β-reductase gene was strongly induced by the pathogen infection (28). This means that insects may actively increase hormone level through 3DE-3β-reductase after stimulation of the microorganisms. In this study, we functionally characterize the 3DE-3β-reductase gene in the silkworm and describe how the silkworm 3DE-3β-reductase regulates ecdysone titer to enhance the immune response.

The strain DaZao of the domesticated silkworm was reared on fresh mulberry leaves at 25°C under a 12/12-h light/dark photoperiod. The day 3 larvae of the fifth instar silkworm were used for all experiments. The fifth instar is the last larval stage for the DaZao strain, and the duration of this instar is ∼7 d. Moreover, the fifth instar larvae begin wandering behavior late in day 7 and purge their gut contents when the ecdysone titer increases. About forty-eight hours later, ecdysone reaches the peak.

The silkworm ovarian cell line (BmN) was maintained in TC-100 insect cell culture medium (Invitrogen) supplemented with 10% FCS (PAA Laboratories, Pasching, Austria).

Escherichia coli strain DH5α (Gram-negative bacteria) and Bacillus bombyseptieus (Gram-positive bacteria) were used for infection experiments. Both bacteria were cultured on petri dishes containing 2.5% lysogeny broth (LB) medium and 1.0% agar (Sangon Biotech, Shanghai, China) at 37°C. Before injection, bacterial colony was grown in LB liquid culture at 37°C overnight. The cells were collected after centrifugation, and the pellets were washed and resuspended with saline. Then, the cell numbers were measured.

E. coli (10⁴ cells/larva) or B. bombyseptieus (10³ cells/larva) or saline was injected into the hemocoel of the silkworm. After incubation for different time points, hemocytes and fat body tissues were dissected on ice and immediately frozen and stored in liquid nitrogen, respectively. Every tissue sample was collected from five larvae.

To survey whether 20E may affect the silkworm immune, we first injected 20E (0.5 μg/larva) (Sigma-Aldrich, St. Louis, MO), and 3 h after treatment, the same cell numbers of bacteria were injected as shown above. DMSO was used as the control. The survival rate of the silkworm was measured after injection. The tissues were collected 24 h after microorganism infection. Additionally, CFU assays were performed to further test the effect of 20E according to a previous study with modification (29). Different concentrations of 20E (50, 500, and 5000 ng/larva) and bacteria were coinjected into the silkworm larvae. DMSO was also used as the control. After 0, 12, and 24 h, 200 μl hemolymph from each treatment was diluted in series (10⁻¹, 10⁻², 10⁻³) with PBS. Then the dilutions were plated on LB plates and incubated at 37°C overnight. Then the colonies were counted.

Because the 20E signal is transduced by the ecdysone receptor (EcR) and ultraspiracle complex, we can block the 20E action by downregulating the expression level of the EcR gene (30, 31). The specific primers containing T7 promoter for the silkworm EcR gene were the same as described in Tian et al. (31). dsRNAs were generated according to our previous paper (32). On the second day of the fifth instar, larvae were used to perform an RNA interference (RNAi) experiment. Ten-microliter solutions containing 30 μg EcR dsRNA were injected into each larva. To ensure the effective RNAi, another 30 μg EcR dsRNA was injected into each larva after 24 h. The same concentration of enhanced GFP (EGFP) dsRNA was used as controls. Six hours after the second injection, E. coli (10⁴ cells/larva) or B. bombyseptieus (10³ cells/larva) was used to infect the treated silkworm. The survival rates of the silkworm were surveyed after injection. The tissues were collected 24 h after microorganism infection.

To obtain the full-length cDNA of the silkworm 3DE-3β-reductase genes, a GeneRacer kit (Invitrogen, Carlsbad, CA) was used according to the manufacturer’s instruction. The procedure was shown in our previous study (33). All the amplified products were purified with an agarose gel purification kit (Omega Bio-tek, Norcross, GA) and then cloned into pEASY-T1-Clone vectors (TransGen Biotech, Beijing, China). Orientation and presence of the inserted cDNA were confirmed by sequencing of the recombinant plasmid.

The silkworm 3DE-3β-reductase gene was cloned into the prokaryotic expression vector pGEX-4T-1 and transformed into E. coli BL21 (DE3) strain. The 3DE-3β-reductase gene was induced by isopropyl β-d-1-thiogalactopyranoside, and the recombinant protein with the GST tag was purified using a GSTrap FF column (GE Healthcare, Uppsala, Sweden) according to the manufacturer’s instruction. The polyclonal Abs against 3DE-3β-reductase were produced by immunizing mice with the purified proteins as described (33).

Sections of the fat body were dissected from the day 3 larvae of the fifth instar silkworm. The sections were fixed in 4% paraformaldehyde at room temperature for 3–4 h. They were then embedded in paraffin and were cut to 5- to 6-μm-thick sections. For hemocytes, 200 μl hemolymph was collected and mixed with 2 vol PBS. Two hundred–microliter cell suspensions were moved on the slides for 20 min and then fixed with 4% paraformaldehyde for 20 min at room temperature. Immunohistochemical staining was performed following the previous study (33).

The cDNA of the coding sequence excluding the signal peptide region of the silkworm 3DE-3β-reductase gene (GenBank ID no. KU233524) was cloned into pIZ/V5-His vector (Invitrogen). The constructed plasmid was confirmed by sequencing. Then the recombinant plasmids were transfected into a Bombyx mori cell line BmN using X-tremeGENE HP DNA transfection reagent (Roche Diagnostics, Indianapolis, IN). The transfected method was used according to the manufacturer’s instructions. Forty-eight hours after transfection, the cells were collected and centrifuged at 6000 × g for 10 min at 4°C and resuspended by 150 mM phosphate buffer (pH 6.8). Then the cell suspension was sonically treated by an ultrasonic cell crusher (Ningbo Scientz Biotechnology, Ningbo City, China). The sample was cooled in a bath of an iced water mixture during sonic treatment. The supernatant of the cell extracts was collected by centrifugation at 12,000 × g for 20 min at 4°C. Because the recombinant proteins contain the His-tag, we used an affinity nickel resin column Ni-NTA superflow (Qiagen, Hilden, German) to purify the His-tag fusion protein. The purified protein was dialyzed with 75 mM phosphate buffer (pH 6.8) and then quantified using the BCA protein assay and visualized after SDS-PAGE by staining with Coomassie blue.

Based on the cDNA sequence of the silkworm 3DE-3β-reductase gene, we designed specific primers containing the T7 promoter sequence. The primers were designed by Primer Premier 5 software and listed in Supplemental Table I (34). The methods to generate and inject dsRNA are the same as shown above.

To further confirm 3DE-3β-reductase activity in the silkworm immunity, we performed rescue experiments by injecting the recombinant 3DE-3β-reductase protein and 20E into the gene knockdown silkworm, respectively. Six hours after the second dsRNA injection, the recombinant protein (1 μg/larva) was injected. The same concentration of BSA was used as the control. Three hours later, E. coli (10⁴ cells/larva) or B. bombyseptieus (10³ cells/larva) was injected to infect the silkworm. Insect 3DE-3β-reductase can convert 3DE to ecdysone, which is the precursor of the active 20E. Hence, we could inject 20E to rescue the 3DE-3β-reductase gene knockdown larvae. Similar to the injection of the recombinant protein, we injected 20E (0.5 μg/larva) after the second dsRNA injection, followed by infection with the bacteria. The DMSO was used as the control. For both rescue experiments, the survival rates of the silkworm were surveyed, and the fat body tissues were dissected at 24 h after bacterial infection.

For each treatment shown above, the stored tissues were grinded in liquid nitrogen to powders. Total RNA was extracted by the Ultrapure RNA kit (Beijing CoWin Biotech, Beijing, China) and treated with DNase I (Takara Bio, Shiga, Japan) to remove the genomic DNA contamination. The RNA was quantified by the UV spectrophotometer, and then 1 μg RNA was reverse transcribed to the first strand of cDNA by the EasyScript one-step gDNA removal and cDNA synthesis SuperMix kit (TransGen Biotech). The specific primers were designed and used in the quantitative real-time PCR analysis (Supplemental Table I). The quantitative real-time PCR was performed using a real-time PCR detection system (CFX96, Bio-Rad, Hercules, CA) with a SsoAdvanced SYBR Green supermix kit (Bio-Rad). The PCR was carried out as follows: 30 s at 95°C, followed by 40 cycles of 5 s at 95°C and 40 s at 60°C. The silkworm translation initiation factor 4A gene was used as the reference gene.

For RNAi and bacterial infection experiments, 3DE-3β-reductases were detected at protein level using Western blotting. Total proteins from tissues or cells were extracted and quantified. SDS-PAGE was used to separate proteins. The polyclonal Ab against 3DE-3β-reductase (at a dilution of 1:8000) shown above was used to perform this assay. The method was also described previously (33).

For ecdysteroid measurements, ecdysteroids were extracted from the silkworm hemolymph after bacterial infection at different time points. Briefly, 100 μl hemolymph was collected from the treated silkworm (three to five individuals) and added with 9 vol methanol. After centrifugation at 12,000 × g for 10 min, an aliquot of supernatant was combined and dried at 70°C, and then the dried extract was dissolved with 150 μl enzyme immunoassay (EIA) buffer (0.4 M NaCl, 1 mM EDTA, and 0.1% BSA in 0.1 M phosphate buffer) (Sangon Biotech). Ecdysteroid levels were quantified via competitive EIA (Cayman Chemical, Ann Arbor, MI) using anti-20E rabbit antiserum (Cayman Chemical), 20E acetylcholinesterase tracer (Cayman Chemical) and standard 20E (Sigma-Aldrich). The acetylcholinesterase activity was quantified by Ellman’s reagent (Cayman Chemical), and the absorbance at 405 nm was detected with an ELx800 absorbance microplate reader (BioTek, Winooski, VT).

All the experiments shown above were independently repeated three times. All statistical analyses in this study were performed in the statistical R package (35).

Previous studies have demonstrated that some stress factors could elevate the 20E titer (36–38). Thus, we examined whether pathogen stress could increase the 20E titer in the silkworm. We first challenged the day 3 larvae of the fifth-instar silkworm with E. coli and B. bombyseptieus, respectively. At different time points after infection, the hemolymph of the treated silkworm was collected and the level of the 20E concentration was measured by the competitive EIA assay. Our results showed that the 20E equivalents were significantly elevated after bacterial infection (12 h postinfection: saline, 33.568 ± 2.133 ng/ml; E. coli, 45.198 ± 1.457 ng/ml, p < 0.001; B. bombyseptieus, 47.832 ± 1.646 ng/ml, p < 0.001; 24 h postinfection: saline, 33.946 ± 1.779 ng/ml; E. coli, 54.305 ± 1.646 ng/ml, p < 0.001; B. bombyseptieus, 56.646 ± 1.886 ng/ml, p < 0.001) (Fig. 1A). This suggests that the silkworm could increase the 20E titer in the context of pathogen attack.

To ascertain whether the elevated 20E may induce or reduce the silkworm immunity, we injected 20E into the silkworm before microbial infection. When different pathogens were used to challenge the silkworm, the survival rates of the silkworm injected with 20E were significantly higher than those of the control (log-rank test: E. coli, 20E versus DMSO, p = 0.0134; B. bombyseptieus, 20E versus DMSO, p = 0.0237) (Fig. 1B). Then, CFU assays were performed to further examine the effect of 20E on the silkworm immunity. From Fig. 1C, we found that 20E treatment could significantly enhance silkworm to kill both bacteria. Note that the low 20E concentration (50 ng/larva) has a similar effect. This indicated that ecdysone could efficiently influence the immune system of the silkworm.

Similar to other insects, silkworm also contains Toll and IMD signaling pathways to stimulate the expressions of different AMP genes, such as cecropin, gloverin, and moricin, although it is still unclear which AMPs are specifically regulated by which pathway (many AMPs can be upregulated by both positive and negative bacteria) (39–41). To assess the effect of 20E on silkworm innate immunity, we surveyed the transcriptional responses of several immune-related genes, including AMP genes (Gloverin and Morcin), adaptor of the Toll receptor (Myd88), peptidoglycan recognition protein (PGRP-S6), and cellular immune-related gene (C-type lectin). Under the normal condition (saline), 20E could significantly induce the expressions of all tested genes excepted for the Toll signaling adaptor Myd88 gene (Fig. 1D). After injection with different bacteria, we obtained similar results, which further show that 20E has the potential to upregulate silkworm immunity.

The effect of the ecdysone is thought to be mediated by EcR. Several studies have proven that RNAi-mediated depletion of EcR gene could prevent 20E signal transduction (13, 15, 31). We then knocked down the expression of the silkworm EcR gene using dsRNA to further investigate the role of ecdysone on the silkworm immunity. The larvae treated with EcR RNAi exhibited a lower ability to defend against microorganisms (log-rank test: E. coli, ds-EGFP versus ds-EcR, p = 0.0115; B. bombyseptieus, ds-EGFP versus ds-EcR, p = 0.0323) (Fig. 2A). As expected, the expressions of the immune-related genes were reduced in EcR RNAi larvae (Fig. 2B–F), suggesting that the 20E signaling pathway via EcR is an important regulator for the silkworm immune response after bacterial infection.

Taken together, these results suggest that 20E could promote the expressions of several immune-related genes and enhance the resistance of the silkworm against pathogen through the activity of EcR.

A previous study showed that the whole length of the silkworm 3DE-3β-reductase gene is 972 bp, and the gene encodes a nonsecretory protein containing 323 aa (42). However, cotton leafworm (Spodoptera littoralis) 3DE-3β-reductase protein has a signal peptide (26). Therefore, we used RACE to obtain the full length of the silkworm 3DE-3β-reductase gene. As shown in Supplemental Fig. 1, the complete length of the coding sequence is 1032 bp. The resulting cDNA sequence encodes a protein of 343 aa with a signal peptide. We further surveyed the expression pattern of the gene and found that the 3DE-3β-reductase protein was expressed in the hemocyte and fat body (Supplemental Fig. 2). This is consistent with a previous study (43). The hemocyte and fat body are the two important immune tissues or organs. We therefore first investigated the induced expression pattern of the silkworm 3DE-3β-reductase gene after pathogen infection. In hemocyte, both E. coli and B. bombyseptieus did not induce the expression of the 3DE-3β-reductase gene (Supplemental Fig. 3A). In contrast, in fat body, both E. coli and B. bombyseptieus stimulated the 3DE-3β-reductase gene expression levels from 6 to 48 h postinfection (Fig. 3A). Additionally, we also confirmed this result at the protein level using Western blotting (Fig. 3B). A previous study suggested that T. ni 3DE-3β-reductase gene was mainly induced in fat body by bacteria (28). This indicates that the 3DE-3β-reductase can rapidly take part in the bacterial-infected silkworm fat body immune system, and this immune response may be common at least in the superfamily Bombycoidea (including both B. mori and T. ni).

To ascertain the function of the 3DE-3β-reductase gene in the silkworm immunity, we performed an RNAi experiment to knock down the expression of the gene. Twenty-four hours after dsRNA injection, compared with the control, the protein level of 3DE-3β-reductase significantly decreased (Fig. 3B). Note that dsRNA could not completely reduce the expression of 3DE-3β-reductase due to relative low efficiency of RNAi in lepidopteran insects. When E. coli or B. bombyseptieus was used to challenge the silkworms, the survival rates of the silkworms treated with 3DE-3β-reductase RNAi were significantly lower than those of the control (log-rank test: E. coli, ds-3β-red versus ds-EGFP, p = 0.0248; B. bombyseptieus, ds-3β-red versus ds-EGFP, p = 0.0142) (Fig. 3C). Previous studies showed that 3DE-3β-reductase plays role in the insect developmental processes (22–24, 27). To exclude the possibility that RNAi itself may affect larval mortality, the survival rate of the RNAi knockdown silkworm was examined under normal circumstance. In the feeding stage, the 3DE-3β-reductase RNAi had no effect on the normal development of the silkworm (log-rank test: ds-3β-red versus ds-EGFP, p = 0.901) (Fig. 3C). In total, the higher mortality of the 3DE-3β-reductase RNAi larvae indicates that this gene is essential for the silkworm immune response.

We then wanted to know how 3DE-3β-reductase takes part in the silkworm immunity. According to previous reports, 3DE-3β-reductase can convert the 3-dehydro-ecdysteroid to ecdysteroid (22–26). If the expression level of the 3DE-3β-reductase gene was downregulated, its enzymatic products might also decrease. Therefore, we collected the hemolymph from the 3DE-3β-reductase and EGFP RNAi larvae at 24 h after bacterial infection and measured the 20E titer as shown above. Compared with the control (ds-EGFP), the 20E equivalents significantly decreased in the 3DE-3β-reductase RNAi silkworms (E. coli, 50.05 ± 2.53 versus 41.81 ± 2.02 ng/ml, p = 0.013; B. bombyseptieus, 49.37 ± 2.58 versus 40.99 ± 3.19 ng/ml, p = 0.026) (Fig. 3D). As 20E could regulate the expressions of the immune-related genes, we further surveyed their transcription levels in 3DE-3β-reductase RNAi larvae. The injection with saline into the 3DE-3β-reductase or EGFP RNAi larvae could not induce the expressions of the tested genes. In contrast, the silkworms pretreated with 3DE-3β-reductase RNAi prior to bacterial challenge showed a significant decrease in the expression levels of the immune-related genes (Fig. 3E).

Additionally, we performed a rescue experiment to further investigate the function of the 3DE-3β-reductase. First, we expressed the recombinant 3DE-3β-reductase protein with His-tag in the silkworm cell line BmN. We purified the protein by nickel affinity chromatography column. The SDS-PAGE and Western blotting analyses showed the purified protein (Supplemental Fig. 3B). We then did the rescue experiment by injecting the purified recombinant protein into the gene knockdown larvae. Moreover, the enzymatic product of the 3DE-3β-reductase is ecdysteroid, and therefore we also injected 20E to rescue the RNAi silkworm. Fig. 4A shows that injection with either 3DE-3β-reductase protein or 20E significantly reduced the mortality of RNAi silkworm after bacterial infection compared with the control (BSA or DMSO). Injection with the active 3DE-3β-reductase protein could elevate the 20E titer after 12 h of treatment (E. coli, 36.96 ± 2.24 versus 47.51 ± 1.92 ng/ml, p = 0.0037; B. bombyseptieus, 37.98 ± 1.72 versus 48.47 ± 2.62 ng/ml, p = 0.0069) (Fig. 4B). Consistent with the increase of the 20E titer, we found that the transcription level of EcR, which is the first factor to pass the 20E signal, is significantly upregulated in the RNAi larvae injected with the 3DE-3β-reductase protein. Meanwhile, the expressions of the immune-related genes were strongly promoted (Fig. 4C).

Taken together, our results suggested that the silkworm could actively increase the expression level of the 3DE-3β-reductase, which may elevate the 20E titer in the hemolymph and thus enhance the silkworm immune response after bacterial infection.

Hormones are important regulators of many biological processes, including metabolism, development, and reproduction. Increasing evidence showed that hormones can also systemically regulate immunity in vertebrates (44). In mammals, glucocorticoids are the main neuroendocrine to regulate the immune system (45). Several studies showed that stressors can rapidly result in the systemic release of glucocorticoids, and thereby alter the mammalian innate immune and inflammatory response through the glucocorticoid receptor (44, 46, 47). In insects, molting hormone 20E titers can also increase when they are exposed to different stressors, such as starvation, heat treatment, and sleep deprivation (36–38). However, whether pathogen stress can change the insect 20E titer is still unclear. Our results showed an almost 40% increase in the 20E titer after bacterial infection. Therefore, similar to its role in starvation and sleep deprivation, elevated 20E might have a function as a stress hormone in pathogen attack.

Ecdysone was thought to be an essential regulator in insect development, metamorphosis, and reproduction. Several previous studies suggested that 20E could promote both cellular and humoral innate immunity, and this enhancement requires EcR in Drosophila (13–15, 20). However, knowledge about the effects of 20E on immune response is very limited in lepidopteran insects. A recent study showed that the transcription levels of most Helicoverpa armigera immune-related genes, including pattern recognition receptors and AMPs, significantly increased during the wandering stage (48). Furthermore, injection with 20E, even at very low concentrations, could strongly induce the expressions of these genes (49). Interestingly, for B. mori, two previous studies obtained opposite results. One study suggested that 20E suppressed the expressions of the silkworm AMP genes (31). In contrast, another study suggested that 20E treatment significantly upregulated the expressions of AMP genes as shown in H. armigera (49). Therefore, we further surveyed the effect of the elevated 20E titer after pathogen attack.

In this study, we used the larvae treated with exogenous 20E and bacterial infection to survey the effect of the ecdysone. All tested immune-related genes except for the Myd88 gene were induced in the silkworm larvae coinjected with 20E and bacteria. Myd88, an adaptor of the Toll receptor, is involved in the Toll-mediated innate immune responses (50). These results are consistent with a previous study in which most of the genes in the Toll signaling pathway cannot be up- or downregulated during the silkworm wandering stage (49). Therefore, 20E may have no effect on the silkworm Toll pathway. However, 20E depresses the transcription levels of several Drosophila Toll-related genes, such as Toll ligand spätzle and Toll transcription factor dorsal (51). The mechanism to explain this difference between the silkworm and fruit fly remains unclear. 20E signaling exerts its effects through the EcR and ultraspiracle complex, which binds to the promoters of target genes (52, 53). EcR is the primitive receptor to transduce the 20E signal. When the EcR gene was downregulated by RNAi, the survival rate of the silkworm and the transcription levels of immune-related genes significantly decreased (Fig. 2). Taken together, bacterial attack could increase 20E titer level, which could enhance the silkworm immunity mediated by EcR. Note that 20E concentration increased to be <50 ng/ml after infection, which is still too low to affect the silkworm normal development (54). Although the concentration is not very high, it is enough to modulate the immune system of the silkworm (Fig. 1C). Additionally, a recent work also showed that the injection of a low concentration of 20E could enhance the expressions of several immune-related genes in another lepidopteran insect (48). Thus, we speculate that insects may use some mechanisms to balance immune and developmental processes. Additionally, our results also demonstrated that the elevated 20E titer could significantly increase the ability of the silkworm to defend against the bacterial infection (Figs. 1, 2), supporting the results of a previous study (49).

As shown above, pathogen attack, starvation, sleep deprivation, and heat treatment could increase the insect 20E titer. However, the mechanism to regulate the 20E titer under environmental stress remains unclear. In this study, we showed, to our knowledge for the first time, that the elevated 20E titer after bacterial infection is generated by the induced 3DE-3β-reductase. 20E is mainly synthesized via a series of hydroxylation and oxidation steps from cholesterol. In this pathway, 3DE is the late product (8). Meanwhile, 3DE is also the intermediate product of the ecdysone 3-epimerizaion pathway (55). Lepidopteran molting gland contains a large amount of 3DE. When insects have the need, 3DE can be released and directly converted into ecdysone by 3DE-3β-reductase in hemolymph (24). In other words, ecdysone can be rapidly synthesized from the stored 3DE in a one-step reaction that is much faster than the traditional synthetic pathway from cholesterol. This may be a reason why insects use this pathway to synthesize 20E to activate immune response.

Because the 20E synthesis pathway from 3DE also exists in Diptera (56), whether our observation is the same case in Drosophila deserves investigation in the future. Additionally, a previous study suggested that bacterial infection induced the expression of the 3DE-3β-reductase gene in T. ni (28) and that the regulation of the 20E titer determined by 3DE-3β-reductase may be common at least in lepidopteran insects. Note that the mechanism to regulate the expression of the silkworm 3DE-3β-reductase gene after bacterial infection is still unclear. It will be of major interest to further study whether some known immune pathways (IMD, Toll pathway) or other signaling pathways have cross-talk with the development related pathways, which may regulate the expression of the silkworm 3DE-3β-reductase gene.

Based on our results, it can be speculated that after silkworm suffered from bacterial infection, 3DE was released from the prothoracic glands to hemolymph, fat body, and other tissues. Moreover, the expression of the silkworm 3DE-3β-reductase gene was induced in the fat body. A previous study showed that 3DE-3β-reductase with enzymatic activity was mainly present in the silkworm hemolymph (43). Indeed, we also detected the induced protein in the silkworm hemolymph after bacterial infection. Therefore, it is likely that 3DE-3β-reductase protein with signal peptide formed in fat body and then was transferred to hemolymph. Then the active 3DE-3β-reductase can convert 3DE into ecdysone, which, in turn, is rapidly hydroxylated to the active 20E. Finally, the high 20E concentration mediated by EcR function will elevate the expressions of several AMP genes, which could enhance the silkworm humoral immunity. Meanwhile, 20E can also induce the cellular immune response through C-type lectin (Fig. 5). Although the elevated 20E titer is limited, it can efficiently enhance the silkworm immune system. Note that this limited increase of the 20E titer cannot affect the normal development of the silkworm. Therefore, we thought that the silkworm evolved a very smart strategy to strengthen its immunity by increasing 20E titer, which is not enough to effect its normal development. Collectively, our findings reveal an important role of the 20E synthesis pathway from 3DE in enhancing immune response and represent a first step, to our knowledge, toward the understanding of interaction mechanisms between insect hormone and immunity.

We thank Deng-Wei Qi for help in preparing the polyclonal Abs against 3DE-3β-reductase and Dr. Shi-Hong Gu for constructive suggestions on our experiments.

The authors have no financial conflicts of interest.