Bio-olfacticides: More Food and less pesticides in insect control – DEMETER

This multidisciplinary project involves a variety of approaches: high-throughput sequencing, bioinformatics, molecular biology, genome editing, electrophysiology, machine learning, and behavioral studies. Modern sequencing methods combining Illumina and PacBio, associated with powerful bioinformatic tools, have enabled the assembly of the genome of S. littoralis and the annotation of ORs of this species. Their phylogenetic analysis combined with RNAseq (transcriptome sequencing) approaches on different tissues, developmental stages (caterpillars/moths) and sexes (males/females) allowed the selection of ORs potentially important in key behaviors (sex pheromone receptors, receptors to plant odors). The functional characterization of these target ORs (ie the identification of the odorants that they detect) combines functional genomic approaches that we have developed on this insect: heterologous expression in a host insect or in vitro, genome editing by CRISPR/Cas9 to inactivate OR in vivo. Quantitative structure-to-activity relationship (QSAR) models are applied to the selected ORs, and used to virtually screen large databases of molecules, predicting new ligands. The predictions are subject to experimentation, by testing the functional response of target ORs to the proposed ligands using electrophysiology. If active, the new ligands are tested on the insect olfactory behavior.

We sequenced the first complete genome of S. littoralis, which allowed us to establish by expert manual curation the complete OR repertoire of this species (~ 80 ORs). We also identified the taste receptors, also important for food decision making by insects, highlighting impressive expansions (more than 250 genes) which would be linked to the polyphagous diet of these herbivores. To characterize the most promising ORs in terms of crop protection applications, we conducted the first large-scale functional study of ORs in a crop pest insect. Through this systematic approach, and guided by RNAseq and phylogenetic analyzes, we identified the sex pheromone receptors of this species, particularly relevant targets of “pherocides”. From an evolutionary perspective, we have shown that moth sex pheromone receptors have appeared at least twice during evolution, revolutionizing the commonly held idea that pheromone receptors have a unique evolutionary origin. Our functional screen also identified receptors involved in the olfactory-guided behavior of caterpillars to plants, on which we built and applied models of computational predictions of agonists and / or antagonists. In silico screens of virtual libraries of molecules have proposed new ligands, whose action on ORs and the behavior of insects has then been verified experimentally. This is the first time that such a combination of chemo-informatics and experimentation has been conducted on a Lepidoptera, and the success rates (30 to 90%) are very promising, concretely demonstrating the effectiveness of in silico approaches to identify new semiochemicals active on insects.

The in silico models, fed by the experimental data generated, will be improved by learning, and will also be developed on other key moth ORs. Among the molecules confirmed to be active on the receptors and inducing a behavior of interest for pest biocontrol (attraction, for instance for trapping; repellence), those presenting the best potential for exploitation (inexpensive, non-toxic) will be selected for large scale tests (e.g. in greenhouse or in the field). Ultimately, such an approach may be extended to other pests. It should be noted that for narrow-spectrum ORs, such as sex pheromone receptors, the in silico predictions are not efficient because the models are not fed with sufficient data. An approach based on the OR structure would make sense in this case. Recent developments in cryo-electron microscopy have enabled the resolution at the atomic scale of the very first three-dimensional structures of insect ORs and the identification of odorant binding sites. Receptor-targeted modeling will complement the ligand-targeted approaches implemented in this project, to broaden the range of bio-olfacticides. In a broader perspective, taste receptors may also be considered, as they could be targeted to complete the panoply of bio-olfacticides with «bio-gustaticides«.

Main scientific publications

Caballero-Vidal G., Bouysset C., Grunig H., Fiorucci S., Montagné N., Golebiowski J., and Jacquin-Joly E. (2020) Machine learning decodes chemical features to identify novel agonists of a moth odorant receptor. Sci. Reports 10:1655.

Bastin-Héline L., de Fouchier A., Cao S., Koutroumpa F., Caballero-Vidal G., Robakiewicz S., Monsempes C., François M.C., Ribeyre T., de Cian A., Walker W.B., Wang G., Jacquin-Joly E. & Montagné N. (2019) A novel lineage of candidate pheromone receptors for sex communication in moths. eLife 8:e49826

Licon CC, Bosc G, Sabri M, Mantel M, Fournel A, Bushdid C, Golebiowski J, Robardet C, Plantevit M, Kaytoue M, Bensafi M (2019) Chemical features mining provides new descriptive structure-odor relationships. Plos Comput. Biol. 15(4):e1006945

de Fouchier A., Sun X., Caballero-Vidal G., Travaillard S., Jacquin-Joly E. & Montagné N. (2018) Behavioral effect of plant volatiles binding to Spodoptera littoralis larval odorant receptors. Front. Behav. Neurosci. 12: 264

Bushdid C., de March C.A., Matsunami H., Golebiowski J. (2018) Numerical Models and In Vitro Assays to Study Odorant Receptors. Methods in Molec. Biol. 1820:77-93

Bushdid C., de March C.A., Fiorucci S., Matsunami H., Golebiowski J. (2018) Agonists of G protein-coupled odorant receptors are predicted from chemical features. The J. Phys. Chem. Letters 9, 2235-2240

de Fouchier A., Walker W.B., Montagné N., Steiner C., Binyameen M., Schlyter F., Chertemps T., Maria A., François M.C., Monsempes C., Anderson P., Hansson B.S., Larsson M. C., Jacquin-Joly E. (2017) Functional evolution of Lepidoptera olfactory receptors revealed by deorphanization of a moth repertoire. Nature Comm. 8: 15709

Patent:

Jacquin-Joly E., de Fouchier A., Montagné N. (2018) “Pheromonal receptor of Spodoptera littoralis and identification of natural ligand of said receptor and uses thereof” INRA. n°16305329.1. Left by inrae in 2021 because of the lack of industrial partners

Context:
Noctuid moths represent an important group of insects that includes the most devastating pests on the planet. As olfaction underlies several of their behaviours that are critical for crop aggression - including reproduction, host selection and oviposition - this sensory modality appears as an attractive target for the development of innovative strategies to reduce the negative impact of these pests.
At the molecular level, the key actors involved in the recognition of chemical signals are the olfactory receptors (ORs), thus they appear as the best targets for the rational design of selective “bio-olfacticides”, molecules able to modify the OR response and thus the corresponding behavior. The nature of insect ORs provides interesting opportunities: 1) they are completely different from vertebrate ORs, allowing targeted and safe actions; 2) they are highly divergent between insects, allowing to act in a selective way to preserve beneficial insects; 3) they present a 7-transmembrane domain structure, similar to that of G-protein coupled receptors (GPCRs) on which pharmacologic know-how is well developed, for the design of therapeutic drugs (i.e. agonists/antagonists interfering with the binding site and thus activating/inhibiting the cellular response).

Objectives:
In this context, we propose to characterize ligands of the ORs of the noctuid moth Spodoptera littoralis, a model species in agronomy for which we have identified a large array of ORs, thanks to a previous project funded by ANR. Its genome we are currently assembling will complete this repertoire. We propose to set up an innovative automatized high-throughput functional screening approach to identify a vast array of ligand-receptor couples (for the first time in a crop pest). This breakthrough in the field of molecular bases of olfaction in insects outside Diptera will identify ORs involved in key behaviors (reproduction, oviposition, attraction, repellence) to target in the frame of crop pest protection. Ligand-based and receptor-based molecular modelling approaches will be used to design in silico potential agonists/antagonists/blockers (the “bio-olfacticides”). Their effects on the OR responses and in fine the behaviour of pest insects will be tested.

Methodology and expected results:
Classical bioinformatics tools associated with manual curation will be used to define full length OR gene models in the genome. The characterization of ligands (or deorphanization) for the ORs identified in S. littoralis will be performed in vitro via their expression in Xenopus oocytes coupled to two-electrode voltage clamp electrophysiology. We will benefit from a high throughput (HT) technology designed for molecular pharmacology studies on mammalian GPCRs, which allows screening of a very large array of natural and synthetic ligands. Sensitivity, specificity and response dynamic of the ORs will be assessed. The response parameters obtained will define the first “odor space” of a crop pest, which will be compared to the two odor spaces established previously in two model insects (Drosophila and the mosquito Anopheles gambiae). Molecular modelling coupled to chemoinformatics will be used to predict potential bio-olfacticides for interesting ORs. The efficiency of these bio-olfacticides will be further tested at the molecular (OR) level using the HT screening platform, at the neuron level using electroantennography and at the behavioural level in the lab using larval (servosphere) and adult (wind tunnel) dedicated assays. More than a proof-of-concept to validate bio-olfacticide efficiency on insect behaviour, this project will propose active molecules to improve crop production and reduce pesticide use, in a context of food safety and environment protection.