The paper in Nature Ecology & Evolution is here: http://go.nature.com/2xWngaF

The tobacco cutworm Spodoptera litura, a noctuid, is a widespread and highly pestiferous phylogenetic group which had received little attention in the genome community at the time. After deep consideration of “Why S. litura?” we decided to design the study as a direct comparison of genomic differences between the highly polyphagous tobacco cutworm, which feeds on more than 100 host plants, and the almost monophagous domesticated silkworm, Bombyx mori, which feeds only on mulberry (Morus) and its close relatives. We speculated that this approach would yield insights into the genomic underpinnings of host plant specialization and possibly reveal a kind of “master system” or strategy polyphagous insects have evolved to deal with their host plants’ diverse defenses against herbivory.

Our plan required the annotation of chemosensory and detoxification related gene families of B. mori and S. litura as a comparison between extreme examples. This turned out to be a huge effort, leading the original team to ask for the help of many specialists who contributed extensively to the analysis of olfactory (ORs) and gustatory receptors (GRs) and cytochrome P450s, glutathione S-transferase (GST) and carboxylesterase (COE), enzymes that insects typically use to detoxify plant allelochemicals and insecticides. Thus emerged a detailed picture of the expansion, chromosomal organization, and phylogenetic relatedness of numerous genes and gene families that enable enhanced detoxification of plant secondary metabolites and insecticides and survival of this pest on so many different hosts. We were also able to show that larval exposure to insecticidal toxins induced expression of detoxification genes, and, although germline RNAi (interference RNA) worked only sporadically in Lepidoptera, thanks to relatively recent advancements in techniques available for functional genomics, knock-down of representative genes using short interfering RNA (siRNA) reduced larval survival, consistent with their contribution to the insect’s natural pesticide tolerance. 

Figure 1 | S. litura caused serious damage on crops in North India.

In parallel with the project on S. litura, colleagues in France were developing a genome project for 2 strains of the closely related species, S. frugiperda, which are differentiated by feeding preference for rice vs corn, with similar questions in mind (Gouin et al., 2017). These comparisons will further strengthen understanding of the basis of these insects, extraordinary ecological adaptation, and the two groups agreed to release the published genome projects at the same time. An independent genome project on two other pestiferous noctuids, the old world bollworm, Helicoverpa armigera, and the corn earworm, H. zea (Pearce et al., 2017), recently reported expansion of similar gene families associated with detoxification of plant secondary compounds, though as expected details are somewhat different (Xu et al., 2016).

The Nature paper describing the evolution of Wolf to Dog analyzed by population genetics (Axelsson et al., 2013) greatly influenced us in thinking about a species dispersed over such a wide geographical area. We wanted to know to what extent have S. litura populations in Asia diverged or remained similar at the genome level? We collected S. litura from several locales and asked our colleagues, Hirohisa Kishino and Jiaqi Wu, to identify variable genome domains among them. Their analysis provided a clear picture of the substructure and variation among near and distant populations of S. litura encompassing a variety of habitats (Fig. 2). It also produced another outstanding result, namely, frequent communication between South India, South China and Japan (Fig. 3). This means this highly deleterious pest is expanding in east Asia through long distance migration, a wholly unexpected finding which will be important for designing successful ways to control it.

Figure 2 |  Examples of variable domains estimated from population genetics analysis. a, A big Cyp9a expansion in S. litura chr29. Middle part: Expression heatmap of Cyp9a genes induced by toxin treatment in three tissues (fb, fat body; mg, midgut; and mp, Malpighian tubule). Lower part: Diversity of each gene involved in the Cyp9a cluster domain including an ADH gene cluster among 16 local populations. b, The distribution of Tajima’s D of 5000 bp windows. Blue, distribution among the windows intersecting with GRs; red, distribution among windows intersecting with coding sequences (CDSs).

Figure 3  | Gene flow and diversifying selection on the S. litura genomes. a, F_ST-based cluster analysis of local populations. b, Genetic diversity and gene flows among local populations. The size of the circles represents the genetic diversity, π. Location pairs with high gene flow (F_ST<0.05) are connected by segments. The line width represents migration rates (4Nm=1/F_ST-1). 

Another novel finding was the massive expansion of bitter receptor genes (or bitter GRs). This led to the idea that functional connections between expanded bitter GRs and enhanced detoxification ability are essential for survival of highly polyphagous species like S. litura. Determining the large number of GRs in S. litura was not straightforward. Although we and the S. frugiperda group could identify more than 200 GR genes based on their sequences, their expression levels are so low it was difficult to find transcriptional evidence for many of them. Thus, in our first transcriptome analysis using RNA-seq we could find evidence for only 44 GR transcripts despite examining most perception organs. Subsequently, we reanalyzed the RNA-seq data using different sensitivity criteria and found 122 GR transcripts including 109 bitter GRs (Fig. 4). Among them are clear specializations within larval (maxilla) and adult (proboscis) taste sensing organs where we found distinct gene clusters of bitter GRs which we regard as examples of likely adaptation for responding to diverse host plant chemicals. As noted by one of the manuscript reviewers, determining bitter GR receptor ligand specificity and signaling pathways is critical for understanding the ability of this powerful pest to detect and activate detoxification mechanisms in so many host plants, and remains for future studies.

Figure 4 | Increase of GR transcripts by reanalysis of RNA-seq data of various tissues with different sensitivity criteria. Left panel (original analysis): We used as criteria that the false discovery rate was <0.01 in any tissue, leading to 44 GR transcripts. Right panel (reanalysis): Due to very low GR expression levels, we recorded all GRs with expression levels higher than 0.1 FPKM in any tissue, leading to 122 GR transcripts. LAnt, larval antenna; LEpi, larval epipharynx; LLeg, larval legs; LMax, larval maxilla; LMid, larval midgut; MAnt, moth antenna; MLeg, moth legs; MPG, moth pheromone glands; MPro, moth proboscis.

Through this project, we realized that verification of a correlation between a massive expansion of bitter GR receptors and extreme polyphagy of this noctuid species is critical to explain the mechanisms of polyphagy, insecticide tolerance and high adaptation to the many ecological niches of this pest. We think a comparison with S. picta could be most appropriate for this project (Fig. 5). S. picta, which separated from S. litura only 3Mya, is a specialist of Crinum, a lily family that produces specific toxic alkaloids, and their geographical distribution is very close or overlapping. Thus we will determine whether or not the S. picta genome has a massively expanded GR family, and perhaps elucidate which GRs are responsible for polyphagy vs host plant specialization. We will also compare detoxification related gene families between the two species.       

Figure 5  | Extremely different feeding habit between S. picta and S. litura. a, S. picta is a specialist of Crinum plants. b, S. litura feeds on over 100 different plants.

When this blog was written, genomes of four of the most notorious noctuid agricultural pests were reported. The availability of these genome sequences will pave the way for deeper analysis of factors involved not only in their wide host range but also their high degree of insecticide tolerance and their ever increasing resistance to insecticides. This research promises to lead the way for better ways to control them, whether by development of new insecticides or other novel as yet undetermined approaches.

References

Xu, W., Papanicolaou, A., Zhang, H. J. & Anderson, A. Expansion of a bitter taste receptor family in a polyphagous insect herbivore. Sci Rep 6, 23666 (2016).

Pearce, S. L. et al. Genomic innovations, transcriptional plasticity and gene loss underlying the evolution and divergence of two highly polyphagous and invasive Helicoverpa pest species. BMC Biol 15(1):63 (2017).

Gouin, A. et al. Two genomes of highly polyphagous lepidopteran pests (Spodoptera frugiperda, Noctuidae) with different host-plant ranges. Sci Rep 7 11816 (2017).

You, M. et al. A heterozygous moth genome provides insights into herbivory and detoxification. Nat Genet 45, 220-225 (2013).

Axelsson E. et al. The genomic signature of dog domestication reveals adaptation to a starch-rich diet. Nat 495, 360-364 (2013).