Identification of a juvenile-hormone signaling inhibitor via high-throughput screening of a chemical library

Insecticide resistance has recently become a serious problem in the agricultural field. Development of insecticides with new mechanisms of action is essential to overcome this limitation. Juvenile hormone (JH) is an insect-specific hormone that plays key roles in maintaining the larval stage of insects. Hence, JH signaling pathway is considered a suitable target in the development of novel insecticides; however, only a few JH signaling inhibitors (JHSIs) have been reported, and no practical JHSIs have been developed. Here, we established a high-throughput screening (HTS) system for exploration of novel JHSIs using a Bombyx mori cell line (BmN_JF&AR cells) and carried out a large-scale screening in this cell line using a chemical library. The four-step HTS yielded 69 compounds as candidate JHSIs. Topical application of JHSI48 to B. mori larvae caused precocious metamorphosis. In ex vivo culture of the epidermis, JHSI48 suppressed the expression of the Krüppel homolog 1 gene, which is directly activated by JH-liganded receptor. Moreover, JHSI48 caused a parallel rightward shift in the JH response curve, suggesting that JHSI48 possesses a competitive antagonist-like activity. Thus, large-scale HTS using chemical libraries may have applications in development of future insecticides targeting the JH signaling pathway.

In current pest control, integrated pest management (IPM) is a promising eco-friendly approach that combines different control methods, rather than using only chemical insecticides, to suppress pest populations to lower levels than would cause economic damage 1 . However, chemical insecticides still play a pivotal role in modern agriculture pest control and are indispensable, even in the IPM system. Recently, resistance to existing insecticides has become a major concern, and the number of usable insecticides has decreased [2][3][4] . Thus, the development of insecticides with novel mechanisms of action is essential for the maintenance of stable agricultural production.
Insect growth regulators (IGRs), which inhibit insect-specific development, are a class of environmentally friendly insecticides compatible with the concept of IPM 5 . Juvenile hormone (JH) is an insect-specific hormone that suppresses precocious metamorphosis during the larval stage 6 , and which has different chemical structures in different insects. Hence, JH signaling pathway may be a suitable target for developing novel IGRs with selectivity to target pests and safety to the ambient environment 5,7,8 . Indeed, JH signaling activators (JHSAs), such as JH agonists (pyriproxyfen and fenoxycarb), which suppress larval-pupal and nymph-adult metamorphosis in holometabolous and hemimetabolous insects, respectively, are often used against agricultural and sanitary insect pests 7,8 . JHSAs are efficacious in controlling small and short-lived insects, such as thrips and aphids, by impairing their metamorphosis and reducing the number of next generations. However, the application of JHSAs to the larvae of relatively large and long-lived insects, such as moths and beetles, may elongate the larval stage, thereby causing the damage to spread. In contrast, JH signaling inhibitors (JHSIs) induce precocious metamorphosis and are expected to reduce feeding-related damage to agricultural products and subsequent generations. Accordingly, JHSIs are theoretically superior to JHSAs as they have applications in the control of both long-and short-lived insect pests; however, only a few JHSIs 9 and some inhibitors of JH biosynthesis 7  www.nature.com/scientificreports/ The molecular mechanisms of the JH signaling pathway in target cells have been elucidated in recent studies [10][11][12][13][14] . Briefly, JH that has entered target cells is received by a JH receptor, methoprene-tolerant protein (Met). Met proteins are a family of basic helix-loop-helix Per-ARNT-Sim (bHLH-PAS) transcription factors. Then, JH-liganded Met forms a heterodimer with steroid receptor coactivator protein (SRC), which also belongs to the family of bHLH-PAS transcription factors. The JH/Met/SRC complex activates the Krüppel homolog 1 gene (Kr-h1), a C 2 H 2 zinc-finger type transcription factor, via the JH response element upstream of Kr-h1 (kJHRE). Then, Kr-h1 blocks the expression of the pupal and adult specifier genes, broad-complex (BR-C) and ecdysone-induced protein 93F (E93), thereby preventing larva from precocious pupal and adult development. Based on this information, we can develop a high-throughput screening (HTS) system to conveniently evaluate the activation and inhibition activities of chemical compounds and subsequently explore novel seed compounds for JHSAs and JHSIs.
The silkworm (Bombyx mori) is a model lepidopteran insect with many available research tools, including genome information, transgenic methods, and cell lines [15][16][17] . In this study, we aimed to develop a practical JHSI by establishing a JH screening system for HTS using a B. mori cell line, in which the JH signaling pathway had been characterized [18][19][20][21][22] . Then, we carried out large-scale screening using this HTS system and succeeded in finding a novel JHSI.

Results and discussions
Establishment of a JH screening system. A previous study reported the identification of inhibitors of Met/SCR complex formation from plant compounds using a yeast two-hybrid system 9 . Treatment of mosquitos with these inhibitors caused defects in ovary development, whereas no effects were observed in larval development 9 . Here, we propose a screening system using an insect cell line to explore novel JHSAs and JHSIs. Because the Kr-h1 transcript has been reported to be induced by JH in most insect cell lines 18,21,[23][24][25] , the intrinsic factors involved in JH signaling, such as Met and SRC, are thought to be sufficiently expressed in these cell lines. In the presence of targeting insect cells, JHSA and JHSI activities were evaluated by introduction of a JH response element (JHRE)-reporter into the cells (the JHRE screening system). In this study, we established a JHRE screening system using a B. mori cell line as model lepidopteran insect.
First, we constructed a reporter vector for stable cell lines in the JHRE screening system (Fig. 1A). To reduce the background of the luc2 reporter in the JHRE reporter plasmid, as described previously (pGL4.14_− 2165 to − 2025 and − 49 to + 116) 18 , the luc2 reporter gene was swapped for a luc2P reporter gene containing the first degradation sequence (PEST) 26 . Moreover, this construct contained a hRlucP reference gene, which was continuously driven by the BmA3 promoter 27 , as a reference reporter to evaluate the cytotoxicity of compounds (Fig. 1A).
This plasmid was transfected into B. mori cells (BmN cells), and the cells were selected by hygromycin for establishment of the stable line (BmN_JHRE-Fluc and A3-Rluc, BmN_JF&AR). Dose-dependent increases in reporter activities were observed in cells treated with JH I; the median effective concentration (EC 50 ) was 3.7 × 10 −10 M, whereas the reporter activity was barely detectable in the absence of JH I (Fig. 1B). This analysis clearly demonstrated that BmN_JF&AR cells were responsive to subnanomolar concentrations of natural JH I, similar to the response levels of BmKr-h1 transcripts and transient reporter assays 18 .
Screening for JHSAs and JHSIs using this cell line is shown schematically in Fig. 1C. In the JHSA assay (left), induction of Fluc luminescence by a test compound indicates that the compound possessed JHSA activity. Meanwhile, in the JHSI assay (right), if Fluc luminescence was reduced when the cells were simultaneously treated with JH and a test compound, the compound possesses JHSI activity. False-positive results were caused by cytotoxicity of the compound and could be excluded by measurement of Rluc luminescence. In this study, we focused on exploration of JHSIs using BmN_JF&AR cells.

HTS of JHSIs.
To identify JHSIs from a chemical library, we performed HTS using a four-step hit validation assay in BmN_JF&AR cells (Fig. 1D). We used 1 nM JH I in JHSI screening based on the dose-response to JH I in BmN_JF&AR cells (Fig. 1B). The plate layout used for each screening is shown in Supplementary Fig. S1. JHSI activity was calculated by the inhibition rate [InH (%)], which was evaluated according to whether a test compound inhibits the reporter activity of 1 nM JH I. Therefore, positive and negative controls were set as dimethyl sulfoxide [DMSO] alone, and 1 nM JH I in DMSO, respectively. The positive and negative controls yielded consistent results in all screenings (Fig. 2), and the performance was qualitatively assessed by Z' factor analysis between the positive and negative controls. The average Z' factor values of the first to fourth screenings were 0.81 ± 0.03, 0.83 ± 0.06, 0.86 ± 0.02, and 0.90 ± 0.02 (Supplementary Table S1), respectively, indicating that our screening was a highly qualitative and reproducible assay.
First, we carried out random screening with a core library consisting of structurally diverse compounds (9600 compounds, n = 1), provided by the Drug Discovery Initiative (DDI, The University of Tokyo) (Figs. 1D, 2A). The screening yielded 120 hit compounds that inhibited 1 nM JH I activity by an inhibition rate (InH) of at least 26% at a compound concentration of 5 μM ( Fig. 2A). The activities of selected compounds were re-evaluated in confirmation tests during the second screening (n = 4) (Fig. 2B). Overall, the results of the first screening were highly reproducible. We excluded compounds that inhibited JH I activity at less than InH 20% or weakened cell adherence to the plate, resulting in 75 hit compounds (Fig. 2B).
Next, dose-response assays (n = 4) narrowed the hit compounds of the second screening (Fig. 2C). For example, as shown in the inset in Fig. 2C, inhibition of 1 nM JH I activity was decreased by treatment with the compounds in a concentration-dependent manner. This screening yielded 47 hit compounds with an InH of greater than 10% at 1 nM. For the fourth screening (n = 4), analogs of 47 hit compounds in the third screening were extracted from another ~ 210,000 compounds in the DDI chemical library, yielding 279 compounds. This identified approximately six analogs per hit compound in the third screening. These analogs were evaluated at Scientific Reports | (2020) 10:18413 | https://doi.org/10.1038/s41598-020-75386-x www.nature.com/scientificreports/ three concentrations (0.05, 0.5, and 5 μM) and were selected as having InHs of greater than 15% at 0.5 μM and − 10% at 0.05 μM (Fig. 2D). Ultimately, 69 compounds were identified as candidate JHSIs that inhibited the reporter activities of 1 nM JH I in BmN_JF&AR cells (Fig. 2D).

Characteristics of the screened compounds.
To characterize the compounds used in the fourth screening, we generated heat maps (Fig. 3A). Compounds of the fourth screening were assigned ADT numbers (ADT: analog and dose-response test), the detail for which are presented in the Fig. 3A legend. In the fourth screening, many additional analogs were observed with higher JHSI activity compared to those of the original compounds identified in the third screening, were obtained by the fourth screening, including the ADT8, ADT9, and ADT13 groups (Fig. 3A). In approximately half of the ADT groups, including ADT2, ADT7, and ADT10, the activities of the analogs were converted into JHSA activities by exchanging with slightly different functional groups (Fig. 3A). For example, ADT16-3 and ADT16-5 exhibited JHSA activities, whereas ADT16-1, ADT16-2, and ADT16-4 had JHSI activities (Fig. 3B). Interestingly, ADT5 compounds had a 4-phenoxyphenoxymethyl skeleton, similar to that of the practical JH analogs pyriproxyfen and fenoxycarb 7 , and exhibited JHSI activities (Fig. 3C).  Supplementary Table S2.

JHSI activity in vitro
With regard to the cytotoxicity of the JHSIs, the highest InH (%) of Rluc luminescence was 42% for 5 μM JHSI27, while the others were all less than 20% ( Supplementary Fig. S2), suggesting that no false-positive results were obtained based on our criterion (> InH 50%). All JHSIs were topically applied to third instar larvae on day 0 of B. mori, and precocious metamorphosis was observed as an index of JHSI activity in vivo (n = 10). Only JHSI48 {N-[2-(4-tert-butyl-2-chlorophenoxy) In the JHSI assay, JH and a test compound were simultaneously added to BmN_JF&AR cells, and the assay measured whether JH-dependent reporter activity was decreased (right). (D) Flow chart of HTS to identify JHSIs from the DDI chemical library using BmN_JF&AR cells. The DDI chemical library possessed 218,000 compounds in total, and the core library (9600 compounds) was composed of structurally diverse compounds for random screening. The core library was used from the first through third screenings, and analogs of the fourth screening were selected from another approximately 210,000 compounds based on the chemical structure of compounds screened in the third screening. The hit compounds selected by the four-step HTS using BmN_JF&AR cells were evaluated by topical application in B. mori larvae in in vivo assays.  (Fig. 4B), for which the median EC 50 was 3.2 × 10 −7 M (Fig. 4C) in BmN_JF&AR cells, induced precocious metamorphosis at the fourth instar larvae at a rate of 60% (Fig. 4D). The cocoon and pupae became miniature sized compared to those treated with DMSO alone (control); whereas the control larvae normally molted to the fifth instar and developed into normal pupae (Fig. 4E). Hence, we investigated the properties of JHSI48 in more detail. To exclude the possibility that precocious metamorphosis was caused by repression of JH biosynthesis in the corpora allata, we validated our findings by culture of epidermis ex vivo. A dose-dependent decrease in BmKr-h1 transcript expression was observed in the epidermis treated with JHSI48; the median EC 50 was 1.1 × 10 −6 M (Fig. 4F). JH binding assays have often been performed to verify direct interactions between Met and chemicals in previous studies [28][29][30] . However, since commercial production of radiolabeled JH required for the assay has been discontinued, we can no longer carry this out. Thus, we carried out dose-response testing with JH I in the presence of JHSIs, a method that is often used to validate either competitive or noncompetitive antagonist in pharmacology 31,32 . According to the pharmacological principle, in the presence of the competitive antagonist, the dose-response curve is shifted to the right in a parallel manner. Alternatively, noncompetitive antagonists, such as compounds that bind to an allosteric site of the receptor or irreversibly binding to the active site of the receptor, reduce the maximal effect that can be produced by the ligand. Reduced maximal effects by JH I were observed in the presence of nine compounds (JHSI8, 23, 24, 26, 27, 28, 29, 42, and 43), while JHSI48 and the others (59 compounds) showed a parallel rightward shift in the JH response curve ( Supplementary Fig. S3 and Fig. 4G). Taken together, these finding suggests that JHSI48 possesses a competitive antagonist-like activity. However, to clarify the precise action of JHSI48, further analysis is needed to demonstrate whether JHSI48 directly binds to the receptor.

Conclusion
We established a screening system for exploration of JHSIs and found JHSI48 as a novel JHSI, which were utilized as a seed of future insecticides targeting the JH signaling pathway. Future synthetic development to improve activity could yield a practical JHSI with applications in pest management. For example, treatment of young larvae with this compound induces precocious metamorphosis and is expected to reduce feeding-related damage to agricultural products and subsequent generations. Additionally, since JH has different chemical structures in different insects, this approach may afford species specificity depending on the specific synthetic development.

Establishment of a stable cell line. The BmN cell line (KATAKURA) derived from the ovaries of B. mori
was maintained at 25 °C in IPL-41 medium (Gibco, Invitrogen) containing 10% fetal bovine serum (HyClone). BmN cells were seeded at a density of 1.5 × 10 5 cells/well in 96-well plates 1 day before transfection. The cells were transfected with pGL4.14_JHREP-luc2P and BmA3P-hRlucP using FugeneHD (Promega). The cells were incubated for 6 h after transfection and selected with IPL-41 medium containing 100 μg/mL hygromycin (Invi-voGen) for 3 months, and the stable cell line (BmN_JF&AR) was then established.
Chemicals. JH I was purchased from SciTech. All compounds used for the HTS were supplied by the DDI of The University of Tokyo; approximately 220,000 compounds were included in the chemical library (https :// Figure 2. Scatter plots of HTS. JHSI activities were examined using the Dual-Luciferase reporter assay system in BmN_JF&AR cells. The JHSI activities of compounds in BmN_JF&AR cells were determined using the inhibition rate [InH (%)], where InH 0% and 100% indicated no activity and maximum JHSI activity, respectively. Gray points are wells containing 1 nM JH I and compound. The red and blue points indicate positive (only DMSO) and negative (JH I, 1 nM) controls, respectively. Plate layouts are described in Supplementary Fig. S1. (A) The core library (9600 compounds, 5 μM) was screened in the first screening (n = 1). The green shadow indicates a plate consisting of 320 compounds, with 16 positive control wells and 16 negative control wells, and the lower plot shows an enlarged image of a plate as an example. (B) The activities of compounds obtained by the first screening were confirmed in the second screening (confirmation test, n = 4). CT, compounds in confirmation test. (C) The dose-responses of compounds selected by the second screening were examined in the third screening (n = 4). The purple shadow focuses on a compound, and the doseresponse is indicated in the inset. (D) Analogs of the selected compounds in the third screening were evaluated in the fourth screening (analog and dose-response tests, n = 3).

HTS using BmN_JF&AR cells.
We used a core library from the DDI containing 9600 diverse compounds for a random screening (first screening). Analogs of compounds selected by the third screening were extracted from another ~ 210,000 compounds in the DDI chemical library and were used in the fourth screening. Compounds in DMSO were dispensed into 384-or 96-well plates using the POD Automation Platform (Labcyte). The layouts of the plates are shown in Supplementary Fig. S1. In the first, third, and fourth screenings, 10 μL In the second screening, 100 μL medium containing BmN_JF&AR cells (final density: 1.5 × 10 5 cells/well) and 100 μL of 2 nM JH I medium (final concentration: 1 nM JH I) were added to 96-well plates, and the reporter activities were measured as described previously 18 . The inhibition rate [InH (%)] was calculated using the following equation: InH (%) = 100 × (1 -[sample -mean p ]/[mean n -mean p ]), where mean p is the mean of Fluc or Rluc luminescence in the positive control (only DMSO), and mean n is the mean of the negative control (JH I in DMSO) 33 . To evaluate the accuracy of the screenings, the Z' factor values for each plate were calculated using the following equation; Z = 1 -(3SD p + 3SD n )/ (mean n -mean p ), where SD n is the standard deviation of the negative control, and SD p is the standard deviation of the positive control 34 . False-positive activity caused by the cytotoxicity of the compound was monitored by determining the InH (%) of Rluc luminescence, and compounds with greater than InH 50% were excluded from the hit compound list. www.nature.com/scientificreports/ Experimental animals and bioassays. B. mori (Kinsyu × Showa strain) was reared on an artificial diet (Nosan Corp) at 25 °C under a 12-h light/dark cycle. In in vivo assays, 0.5 μL hit compounds in DMSO (10 mM), which were selected by the fourth screening, were topically applied to the dorsal epidermis on day 0 of the third instar larvae, and precocious metamorphosis was observed as an index of JH antagonist activity.
dissected in phosphate-buffered saline (137 mM NaCl, 8 mM Na 2 HPO 4 , 2.7 mM KCl, and 1.5 mM KH 2 PO 4 , pH 7.4) and then cultured in Grace's insect medium (Gibco, Invitrogen) at 25 °C, as previously described 21 . The integuments were incubated in medium containing 10 nM JH I and different concentrations of JHSI48 for 4 h.
qPCR. Total RNA was extracted from the epidermis using an RNeasy Plus mini kit (Qiagen) and was used to synthesize cDNA with a PrimeScript RT reagent kit (Takara Bio). The primers used to quantify BmKr-h1 (common region of two isoforms) and BmRp49 were described previously (Supplementary Table S2) 18 . BmRp49 was used as an internal reference. Reactions were performed in a 10-μL volume containing template cDNA derived from 1 ng total RNA, 5 μL SYBR Premix Ex Taq (Takara Bio), and each primer (0.2 μM) using a LightCycler 480 real-time thermal cycler (Roche). The PCR conditions were as follows: 95 °C for 5 min, followed by 55 cycles of 95 °C for 5 s and 60 °C for 20 s. Relative gene expression was determined by the 2 −ΔΔCt method 35 .
Identification of competitive and non-competitive antagonist. To verify whether JHSIs are competitive or non-competitive antagonists, pharmacological assays 31,32 via dose-response testing in the presence of JHSIs were carried out by the same method as described for HTS. JH I was added at concentrations of JH I 0, 10 -10 , 10 -9 , 10 -8 , 10 -7 , and 10 -6 M, and JHSIs were tested at 0, 0.5, and 5 μM. To monitor the reduction in the maximal effect caused by non-competitive antagonist, the luminescence of only 10 -6 M JH I was compared to that of 0.5 μM of JHSIs using Student's t tests. p < 0.001 was regarded as non-competitive antagonist.