MAPK-dependent hormonal signaling plasticity contributes to overcoming Bacillus thuringiensis toxin action in an insect host

The arms race between entomopathogenic bacteria and their insect hosts is an excellent model for decoding the intricate coevolutionary processes of host-pathogen interaction. Here, we

demonstrate that the MAPK signaling pathway is a general switch to trans-regulate differential expression of aminopeptidase N and other midgut genes in an insect host, diamondback moth

(Plutella xylostella), thereby countering the virulence effect of Bacillus thuringiensis (Bt) toxins. Moreover, the MAPK cascade is activated and fine-tuned by the crosstalk between two

major insect hormones, 20-hydroxyecdysone (20E) and juvenile hormone (JH) to elicit an important physiological response (i.e. Bt resistance) without incurring the significant fitness costs

often associated with pathogen resistance. Hormones are well known to orchestrate physiological trade-offs in a wide variety of organisms, and our work decodes a hitherto undescribed

function of these classic hormones and suggests that hormonal signaling plasticity is a general cross-kingdom strategy to fend off pathogens.

Insects are the most speciose group of animals, with approximately 5.5 million species estimated to exist on earth1. Many insects are agricultural pests regularly causing 10–20% crop losses

and with on-going global warming, they post a real threat to world food security2. The gram-positive bacterium Bacillus thuringiensis (Bt) can produce protein toxins to kill different

insects with high host specificity and environmental safety3, which makes it the most successful biopesticide for the last few decades4. Transgenic crops expressing Bt toxins (Bt crops) have

become the cornerstone of bioinspired pest control technology, with >100 million hectares planted globally in 20185. Although Bt products have provided unprecedented economic,

environmental, and social benefits3,6,7,8,9,10, the rapid evolution of Bt resistance in at least nine insect species in the field has seriously eroded their potential4,11,12,13,14.

Unraveling the molecular mechanisms of Bt resistance has important implications for the sustainable utilization of Bt-based technology15,16,17.

Bt Cry toxins exert toxicity in insect larval midguts via a multi-step process requiring protoxin activation, toxin–receptor interaction, toxin oligomerization, membrane insertion, and pore

formation18,19. Alterations of midgut receptors such as cadherin (CAD), aminopeptidase N (APN), alkaline phosphatase (ALP), and ABC transporters (e.g., ABCC2) disrupt toxin binding and are

generally associated with high-level resistance to Bt Cry toxins in insects20,21. The diamondback moth, Plutella xylostella (L.), is one of the most devastating and cosmopolitan agricultural

pests22. It was the first insect to develop field-evolved resistance to Bt biopesticides23, and the availability of complete whole genome information24 renders it an excellent model to

probe how insect hosts withstand Bt infection during host–pathogen interaction. Previously, field-evolved resistance to Bt Cry1Ac toxin in P. xylostella has been linked to both a

cis-mutation in the PxABCC2 gene25 and MAPK-mediated differential expression of PxmALP, PxABCB1, PxABCC1-3, and PxABCG1 genes26,27,28. Although we found that the MAPK signaling pathway can

alter the expression of multiple midgut genes related to Cry1Ac resistance in P. xylostella, its downstream response gene repertoires and upstream activation signals remained mysterious29.

APN proteins are a class of endoproteases catalyzing the cleavage of neutral amino acids from the N-terminus of protein or polypeptide substrates30. They belong to the M1 family

(metallo-type) of zinc-dependent aminopeptidases that are implicated in many physiological processes of diverse organisms31. In the insect midgut, besides their important role in food

digestion, APNs were the first identified functional receptors of Bt Cry toxins20,30. Moreover, we had previously detected several differentially expressed midgut PxAPN genes in Bt

Cry1Ac-resistant P. xylostella32,33. However, whether the differential expression of these PxAPN genes associated with Cry1Ac resistance was also trans-regulated by the MAPK cascade in P.

xylostella was unclear.

Insect endocrinologists have studied insect hormones for more than a century, and they have discovered that two major insect hormones, juvenile hormone (JH) and 20-hydroxyecdysone (20E), act

antagonistically with each other to coordinately orchestrate insect life-history traits including growth, development, and reproduction34,35,36,37. Moreover, JH and 20E are multifunctional

players that can also participate in insect immune defense to pathogenic infection38,39, and the MAPK signaling pathway is involved in this pleiotropic hormone signaling network35,40. Since

exogenous hormone treatments can alter APN gene expression in insects41, we also wanted to test whether altered levels of insect hormones can activate the MAPK cascades thereby

trans-regulating the differential expression of PxAPN and potentially other midgut genes to confer Cry1Ac resistance in P. xylostella.

In this study, we further probe the mechanism of Bt Cry toxin resistance in P. xylostella, starting by looking at the potential role of APN genes, and we confirm that MAPK-mediated

differential expression of APN and other midgut genes does lead to Cry1Ac resistance in P. xylostella. More importantly, we uncover that the MAPK cascade is activated and modulated by an

enhanced pleiotropic hormone signaling pathway. This study provides a model for understanding how intracellular signaling networks shape the expression landscape of midgut genes causing Bt

resistance/tolerance in insects within the context of balancing growth-defense tradeoffs.

Based on currently available transcriptome and genome databases of P. xylostella, 18 M1 aminopeptidase genes including 15 APN genes were identified in silico, and their full-length cDNA

sequences were successfully cloned except for PxAPN3b (Supplementary Table 1 and Supplementary Fig. 1a). A representative lepidopteran APN protein contains six common features (Supplementary

Fig. 1b), including the characteristic gluzincin aminopeptidase motif GAMEN and the zinc-binding/gluzincin motif HEX2HX18E located in the peptidase_M1 domain which are conserved in nearly

all of these M1 aminopeptidases (Supplementary Table 1 and Supplementary Fig. 1c). We found that the APN1-12 gene cluster possesses highly conserved synteny in both gene order and

orientation in different lepidopteran insects, indicating that it has undergone tandem gene duplication during insect genome evolution (Fig. 1a). Although the paralogous PxAPN1-12 genes show

similar features including exon number, size, and intron phase (Supplementary Fig. 1d), they share relatively low protein sequence similarity (Supplementary Fig. 1e), implying their

evolutionary and functional diversity. A model-based phylogenetic analysis demonstrates that lepidopteran APN proteins cluster into 13 classes and are evolutionarily conserved in each class.

Sister phylogenetic relationships were also observed between APN1 and APN3 and between APN5 and APN6, suggesting close protein structure and functional similarities within these pairs

(Supplementary Table 2 and Fig. 1b).

a Synteny analysis of APN1-12 genes among four lepidopteran insects. b Phylogenetic analysis of the currently available 340 lepidopteran APN and other M1 aminopeptidases by maximum

likelihood method based on the optimized LG+G model at 495 aligned amino acid positions. c The constitutive transcription profiles of PxAPN and other M1 aminopeptidase genes in midgut

tissues of fourth-instar larvae from all the Bt-susceptible and -resistant P. xylostella strains as determined by qPCR analysis. For each gene, the expression fold changes are color-coded

according to the gradient, magenta and green rectangles indicate significant up- and down-regulation, respectively (ratio >1.5-fold in either direction), whereas yellow rectangles indicate

no significant transcription variations. Genes are organized according to their phylogenetic tree constructed by the maximum likelihood method based on the optimized LG+G+I model at 690

aligned amino acid positions. d The relative expression levels of PxAPN1 and PxAPN3a proteins in BBMV samples of fourth-instar larvae from different strains. Both the detection of PxAPN

protein levels by Western blots (upper row) and quantitative estimation of band intensity by densitometry (graph) are presented. Data are presented as mean values (c) and mean values ± SEM

(d), n = 3 biologically independent samples, *p