Circadian regulation of night feeding and daytime detoxification in a formidable Asian pest Spodoptera litura

Abstract

Voracious feeding, trans-continental migration and insecticide resistance make Spodoptera litura among the most difficult Asian agricultural pests to control. Larvae exhibit strong circadian behavior, feeding actively at night and hiding in soil during daytime. The daily pattern of larval metabolism was reversed, with higher transcription levels of genes for digestion (amylase, protease, lipase) and detoxification (CYP450s, GSTs, COEs) in daytime than at night. To investigate the control of these processes, we annotated nine essential clock genes and analyzed their transcription patterns, followed by functional analysis of their coupling using siRNA knockdown of interlocked negative feedback system core and repressor genes (SlituClk, SlituBmal1 and SlituCwo). Based on phase relationships and overexpression in cultured cells the controlling mechanism seems to involve direct coupling of the circadian processes to E-boxes in responding promoters. Additional manipulations involving exposure to the neonicotinoid imidacloprid suggested that insecticide application must be based on chronotoxicological considerations for optimal effectiveness.

Fig. 1. Behavioral rhythms of S. litura larvae during 6LD1 to 6LD3.

a Locomotion rhythm measured by recording the percentage of active larvae; b amount of consumed food; and c excrement weight under LD 12:12 (red solid line) and DD (blue dashed line) conditions. Nine groups each with 10 individuals were used for feeding behavior statistics. Counts were made for 6LD1-6LD3 larvae at 3 h intervals. Locomotion activity rhythm was measured according to the percentage of “active” larvae (see Methods); larvae were observed every 20 min, for a total of 9 values obtained at 3 h intervals. White and black shading indicates the photophase and scotophase. Error bars represent SEM.

Fig. 2. Oscillation of detoxification gene expression in the fat body and midgut.

a Clustering of genes with the same expression pattern in the midgut (Cluster A11) and fat body (Cluster B12). The expression values were processed using Mfuzz software with the method of homogenization by log10. b The percentage of detoxification genes with peak expression at specific time points in the fat body and midgut. c Expression heatmap of selected detoxification genes. d RT-qPCR analysis of representative detoxification genes from each of the three major families, P450-082 (top), GST022 (middle), and COE051 (bottom). Each experiment was performed with fat bodies from 3 6LD2 larvae and was repeated independently three times. Error bars represent SEM, photophase is represented by white rectangles, and scotophase by black rectangles and black shading.

Fig. 3. RT-qPCR analyses of core circadian gene transcription in larval brain.

Amounts of transcripts of SlituClk (a), SlituBmal1 (b), SlituPer (c), SlituTim (d), and SlituCwo (e) during 6LD1 to 6LD3 under LD 12:12 (red solid line) and DD (blue dashed line) after entrainment by LD 12:12. Brains were collected from 3 groups with 10 individuals each at 3 h intervals for 72 h under the two conditions. The expression profile of the reference gene SlituActin3 (black dotted line) is shown after homogenization with log10. Light and black shade represent the photophase and scotophase. The results are given as mean ± SEM of three independently repeated experiments.

Fig. 4. Effects of knockdown of circadian core genes on expression of detoxification genes and imidacloprid sensitivity.

Relative expression of detoxification genes was measured by RT-qPCR after injection with a combined siRNAs for SlituClk and SlituBmal1 (“2 core”, red) or siRNA for GFP (control, blue); and b siRNA for SlituCwo (red) or GFP (control, blue). c Effect of artificial diet containing imidacloprid (30 μg/g) on larvae after the injection of siRNA with [I] combined SlituClk + Bmal1 or [II] GFP as control. d Time course of imidacloprid sensitivity after knockdown. Larvae were fed artificial diet supplemented with imidacloprid (30 μg/g) 24 h after knockdown by siRNA injection; “affected” larvae were recorded every 6 h. Larvae were scored as “affected” when they rounded up and did not move or feed and excreted shapeless feces. Several hours later, some recovered from this “suspended” state and some died. Three groups of 10 6LD1 individuals were prepared for siRNA injection and each value represents the average of three experiments with SEM error bars. The level of statistically significant difference was set at *P value < 0.05, **P value < 0.01, ***P value < 0.001 and ****P value < 0.0001.

Fig. 5. E-box annotation and location in the regulatory region of detoxification genes.

a Location of multiple E-boxes in the regulatory regions of clustered detoxification genes. Different types of E-boxes are annotated 3 kb upstream from transcriptional start sites (TSS) and shown with different colored rectangles. Daytime activated detoxification genes (red) and night activated detoxification genes (blue) are shown for a given cluster. b The number of detoxification genes relative to the distance between the E-box sequence and the TSS. c Annotated motifs of canonical E-boxes recognized by specific transcription factors.

Fig. 6. Effects of core circadian gene overexpression on detoxification gene transcription through E-box binding in cultured cells.

a Location of canonical (CACGTG, red) and non-canonical (CACCTG, yellow; CATGTG, green; CAACTG, orange) E-boxes in 5′ regulatory regions of SlituCOE051, SlituGST035, and SlituP450-082. b Relative luciferase activity for the promoters of SlituP450-082, SlituGST035, and SlituCOE051 induced by co-transfection of Spli-221 cells with SlituClk (OE Clk), SlituBmal1 (OE Bmal1), SlituPer (OE Per), SlituCwo (OE Cwo), and combined SlituClk + Bmal1 (OE 2 core) overexpression vectors (red). Control vectors overexpress EGFP (blue). c Relative luciferase activity of mutated (TGTACT, red) and normal (CACGTG, Ctrl, blue) E-box sequences for SlituP450-082 induced by co-transfection of SlituClk and SlituBmal1. The results are given as mean ± SEM of three repeated experiments and statistically significant difference was set at *P value < 0.05, **P value < 0.01 and ***P value < 0.001.

Fig. 7. Topical treatment of imidacloprid on 4th instar larvae.

a Outline of day–night settings and timing of imidacloprid treatment. b The effect of imidacloprid exposure on normal larvae. The experiments were performed during daytime (10 a.m.–4 p.m., red) and night (10 p.m.–4 a.m., blue) in normal LD (12 h:12 h) and DD (continued darkness) conditions based on a preliminary LD₅₀ test (see Methods). Larvae were scored as “affected” when they rounded up, stiffened and did not move when touched, as if dead (“suspended animation”). Three groups of 10 4LD2 larvae were used; each group was treated once during the day or night. The mean percentage of larvae affected after direct exposure to imidacloprid solution on their dorsal surface is shown ± SEM with three repeated tests and statistically significant difference was set at *P value < 0.05, **P value < 0.01 and ***P value < 0.001.