First draft genome assembly of the desert locust, Schistocerca gregaria.

Background: At the time of publication, the most devastating desert locust crisis in decades is affecting East Africa, the Arabian Peninsula and South-West Asia. The situation is extremely alarming in East Africa, where Kenya, Ethiopia and Somalia face an unprecedented threat to food security and livelihoods. Most of the time, however, locusts do not occur in swarms, but live as relatively harmless solitary insects. The phenotypically distinct solitarious and gregarious locust phases differ markedly in many aspects of behaviour, physiology and morphology, making them an excellent model to study how environmental factors shape behaviour and development. A better understanding of the extreme phenotypic plasticity in desert locusts will offer new, more environmentally sustainable ways of fighting devastating swarms. Methods: High molecular weight DNA derived from two adult males was used for Mate Pair and Paired End Illumina sequencing and PacBio sequencing. A reliable reference genome of Schistocerca gregaria was assembled using the ABySS pipeline, scaffolding was improved using LINKS. Results: In total, 1,316 Gb Illumina reads and 112 Gb PacBio reads were produced and assembled. The resulting draft genome consists of 8,817,834,205 bp organised in 955,015 scaffolds with an N50 of 157,705 bp, making the desert locust genome the largest insect genome sequenced and assembled to date. In total, 18,815 protein-encoding genes are predicted in the desert locust genome, of which 13,646 (72.53%) obtained at least one functional assignment based on similarity to known proteins. Conclusions: The desert locust genome data will contribute greatly to studies of phenotypic plasticity, physiology, neurobiology, molecular ecology, evolutionary genetics and comparative genomics, and will promote the desert locust's use as a model system. The data will also facilitate the development of novel, more sustainable strategies for preventing or combating swarms of these infamous insects.

First evidence of protein-carbohydrate regulation in a plant bug (Lygus hesperus).

Lygus bugs are highly polyphagous piercing/sucking insects found throughout North America. Collectively, they have been reported to feed on over 330 plant species (one of the broadest host range ever documented for a group of insects); they also feed on many economically important crops. Despite its prevalence across North America and status as a common pest in many agroecosystems, very little is known about how Lygus bugs regulate their intake of nutrients. In reality, little is known about nutrient regulation for most hemipterans, specifically non-phloem feeding species in the suborder Heteroptera. This likely reflects difficulties in developing adequate artificial diets for insects with piercing/sucking mouthparts. There is, however, an artificial diet for L. hersperus, and in this study we modified it and performed choice and no-choice experiments to determine how L. hesperus regulates its intake of two macronutrients - protein (p) and carbohydrates (c) - that are tightly linked to survival and performance in other insect herbivores. In choice experiments L. hesperus was allowed to select between two foods with different protein:carbohydrate ratios. We documented strong regulation for protein and carbohydrates, with late instar nymphs selecting a slightly protein-biased intake target (protein-carbohydrate ratio = 1.5:1). We also performed no-choice experiments, where nymphs were restricted to a single food. Here, the protein-carbohydrate ratio of their food had a strong impact on survival, which was highest for nymphs reared on the treatment with a protein-carbohydrate ratio closest to the self-selected intake target (determined by the choice experiments), but no significant impact on developmental time or mass gain. Our data are the first of their kind for a non-phloem feeding hemipteran and provide a starting point for more broadly understanding and further investigating the nutritional ecology/physiology of Lygus bugs. Our study also provides a framework for exploring nutrient regulation in other hemipterans and for optimizing artificial diets for piercing/sucking insects, especially heteropterans.

Nutrition affects insect susceptibility to Bt toxins.

Pesticide resistance represents a major challenge to global food production. The spread of resistance alleles is the primary explanation for observations of reduced pesticide efficacy over time, but the potential for gene-by-environment interactions (plasticity) to mediate susceptibility has largely been overlooked. Here we show that nutrition is an environmental factor that affects susceptibility to Bt toxins. Protein and carbohydrates are two key macronutrients for insect herbivores, and the polyphagous pest Helicoverpa zea self-selects and performs best on diets that are protein-biased relative to carbohydrates. Despite this, most Bt bioassays employ carbohydrate-biased rearing diets. This study explored the effect of diet protein-carbohydrate content on H. zea susceptibility to Cry1Ac, a common Bt endotoxin. We detected a 100-fold increase in LC50 for larvae on optimal versus carbohydrate-biased diets, and significant diet-mediated variation in survival and performance when challenged with Cry1Ac. Our results suggest that Bt resistance bioassays that use ecologically- and physiologically-mismatched diets over-estimate susceptibility and under-estimate resistance.

Revisiting macronutrient regulation in the polyphagous herbivore Helicoverpa zea (Lepidoptera: Noctuidae): New insights via nutritional geometry.

Insect herbivores that ingest protein and carbohydrates in physiologically-optimal proportions and concentrations show superior performance and fitness. The first-ever study of protein-carbohydrate regulation in an insect herbivore was performed using the polyphagous agricultural pest Helicoverpa zea. In that study, experimental final instar caterpillars were presented two diets - one containing protein but no carbohydrates, the other containing carbohydrates but no protein - and allowed to self-select their protein-carbohydrate intake. The results showed that H. zea selected a diet with a protein-to-carbohydrate (p:c) ratio of 4:1. At about this same time, the geometric framework (GF) for the study of nutrition was introduced. The GF is now established as the most rigorous means to study nutrient regulation (in any animal). It has been used to study protein-carbohydrate regulation in several lepidopteran species, which exhibit a range of self-selected p:c ratios between 0.8 and 1.5. Given the economic importance of H. zea, and it is extremely protein-biased p:c ratio of 4:1 relative to those reported for other lepidopterans, we decided to revisit its protein-carbohydrate regulation. Our results, using the experimental approach of the GF, show that H. zea larvae self-select a p:c ratio of 1.6:1. This p:c ratio strongly matches that of its close relative, Heliothis virescens, and is more consistent with self-selected p:c ratios reported for other lepidopterans. Having accurate protein and carbohydrate regulation information for an insect herbivore pest such as H. zea is valuable for two reasons. First, it can be used to better understand feeding patterns in the field, which might lead to enhanced management. Second, it will allow researchers to develop rearing diets that more accurately reflect larval nutritional needs, which has important implications for resistance bioassays and other measures of physiological stress.

Diet micronutrient balance matters: How the ratio of dietary sterols/steroids affects development, growth and reproduction in two lepidopteran insects.

Insects lack the ability to synthesize sterols de novo so they acquire this essential nutrient from their food. Cholesterol is the dominant sterol found in most insects, but in plant vegetative tissue it makes up only a small fraction of the total sterol profile. Instead, plants mostly contain phytosterols; plant-feeding insects generate the majority of their cholesterol by metabolizing phytosterols. However, not all phytosterols are readily converted to cholesterol, and some are even deleterious when ingested above a threshold level. In a recent study we showed that caterpillars reared on tobacco accumulating novel sterols/steroids exhibited reduced performance, even when suitable sterols were present. In the current study we examined how the dominant sterols (cholesterol and stigmasterol) and steroids (cholestanol and cholestanone) typical of the modified tobacco plants affected two insect herbivores (Heliothis virescens and Helicoverpa zea). The sterols/steroids were incorporated into synthetic diets singly, as well as in various combinations, ratios and amounts. For each insect species, a range of performance values was recorded for two generations, with the eggs from the 1st-generation adults as the source of neonates for the 2nd-generation. Performance on the novel steroids (cholestanol and cholestanone) was extremely poor compared to suitable sterols (cholesterol and stigmasterol). Additionally, performance tended to decrease as the ratio of the novel dietary steroids increased. We discuss how the balance of different dietary sterols/steroids affected our two caterpillar species, relate this back to recent studies on sterol/steroid metabolism in these two species, and consider the potential application of sterol/steroid modification in crops.

A dietary test of putative deleterious sterols for the aphid Myzus persicae.

The aphid Myzus persicae displays high mortality on tobacco plants bearing a transgene which results in the accumulation of the ketosteroids cholestan-3-one and cholest-4-en-3-one in the phloem sap. To test whether the ketosteroids are the basis of the plant resistance to the aphids, M. persicae were reared on chemically-defined diets with different steroid contents at 0.1-10 µg ml(-1). Relative to sterol-free diet and dietary supplements of the two ketosteroids and two phytosterols, dietary cholesterol significantly extended aphid lifespan and increased fecundity at one or more dietary concentrations tested. Median lifespan was 50% lower on the diet supplemented with cholest-4-en-3-one than on the cholesterol-supplemented diet. Aphid feeding rate did not vary significantly across the treatments, indicative of no anti-feedant effect of any sterol/steroid. Aphids reared on diets containing equal amounts of cholesterol and cholest-4-en-3-one showed fecundity equivalent to aphids on diets containing only cholesterol. Aphids were reared on diets that reproduced the relative steroid abundance in the phloem sap of the control and modified tobacco plants, and their performance on the two diet formulations was broadly equivalent. We conclude that, at the concentrations tested, plant ketosteroids support weaker aphid performance than cholesterol, but do not cause acute toxicity to the aphids. In plants, the ketosteroids may act synergistically with plant factors absent from artificial diets but are unlikely to be solely responsible for resistance of modified tobacco plants.

Plant phloem sterol content: forms, putative functions, and implications for phloem-feeding insects.

All eukaryotes contain sterols, which serve as structural components in cell membranes, and as precursors for important hormones. Plant vegetative tissues are known to contain mixtures of sterols, but very little is known about the sterol composition of phloem. Plants are food for many animals, but plant-feeding arthropods (including phloem-feeding insets) are unique among animals in that they have lost the ability to synthesize sterols, and must therefore acquire these essential nutrients from their food, or via endosymbionts. Our paper starts by providing a very brief overview of variation in plant sterol content, and how different sterols can affect insect herbivores, including those specializing on phloem. We then describe an experiment, where we bulk collected phloem sap exudate from bean and tobacco, and analyzed its sterol content. This approach revealed two significant observations concerning phloem sterols. First, the phloem exudate from each plant was found to contain sterols in three different fractions - free sterols, sterols conjugated to lipids (acylated), and sterols conjugated to carbohydrates (glycosylated). Second, for both plants, cholesterol was identified as the dominant sterol in each phloem exudate fraction; the remaining sterols in each fraction were a mixture of common phytosterols. We discuss our phloem exudate sterol profiles in a plant physiology/biochemistry context, and how it relates to the nutritional physiology/ecology of phloem-feeding insects. We close by proposing important next steps that will advance our knowledge concerning plant phloem sterol biology, and how phloem-sterol content might affect phloem-feeding insects.

Sterol/steroid metabolism and absorption in a generalist and specialist caterpillar: effects of dietary sterol/steroid structure, mixture and ratio.

Insects cannot synthesize sterols de novo, so they typically require a dietary source. Cholesterol is the dominant sterol in most insects, but because plants contain only small amounts of cholesterol, plant-feeding insects generate most of their cholesterol by metabolizing plant sterols. Plants almost always contain mixtures of different sterols, but some are not readily metabolized to cholesterol. Here we explore, in two separate experiments, how dietary phytosterols and phytosteroids, in different mixtures, ratios, and amounts, affect insect herbivore sterol/steroid metabolism and absorption; we use two caterpillars species - one a generalist (Heliothis virescens), the other a specialist (Manduca sexta). In our first experiment caterpillars were reared on two tobacco lines - one expressing a typical phystosterol profile, the other expressing high amounts/ratios of stanols and 3-ketosteroids. Caterpillars reared on the control tobacco contained mostly cholesterol, but those reared on the modified tobacco had reduced amounts of cholesterol, and lower total sterol/steroid body profiles. In our second experiment, caterpillars were reared on artificial diets containing known amounts of cholesterol, stigmasterol, cholestanol and/or cholestanone, either singly or in various combinations and ratios. Cholesterol and stigmasterol-reared moths were mostly cholesterol, while cholestanol-reared moths were mostly cholestanol. Moth tissue cholesterol concentration tended to decrease as the ratio of dietary cholestanol and/or cholestanone increased. In both moths cholestanone was metabolized into cholestanol and epicholestanol. Interestingly, M. sexta generated much more cholestanol than epicholestanol, while H. virescens did the opposite. Finally, total tissue steroid levels were significantly reduced in moths reared on diets containing very high levels of cholestanol. We discuss how dietary sterol/steroid structural differences are important with respect to sterol/steroid metabolism and uptake, including species-specific differences.

Plant sterols and host plant suitability for generalist and specialist caterpillars.

Insects, unlike plants and vertebrates, lack the ability to biosynthesize sterols. Cholesterol is typically the most common sterol found in plant-feeding insects, but it is rarely found in plants above trace levels, so plant-feeding insects must produce the cholesterol they need by metabolizing the sterols found in the plants they eat. Plant-feeding insects are, however, often limited in terms of which sterols can be converted to cholesterol. In the current study we used a transgenic tobacco plant line that displays high levels of atypical plant steroids, specifically stanols and ketone-steroids, to explore how novel steroid structural features affect performance in three economically important caterpillars (Heliothis virescens, Spodoptera exigua, and Manduca sexta). For each species we measured pupation success, larval development, pupal mass, pupal development, and eclosion success. For the two generalists species (H. virescens and S. exigua) we also measured egg production and egg viability. We then used these eggs to replicate the experiment, so that we could examine the effect of parental steroid dietary history on survival, growth and reproduction of 2nd-generation individuals. Significant negative effects of novel steroids on larval and pupal performance were observed for each caterpillar in the first generation, although these were often subtle, and were not consistent between the three species. In the second generation, larval survival estimated by 'pupation number/plant' on the tobacco plants with novel steroids was significantly reduced, while eclosion success was significantly lower for H. virescens. With respect to adult reproduction (i.e. egg production and egg viability) there were no observed differences in the first generation, but novel steroids significantly negatively impacted reproduction in the second generation. The findings from this study, when integrated into a simple population growth model, demonstrate the potential in using plants with modified steroids as a novel approach to manage populations of economically important caterpillar species.

Stability of AtVSP in the insect digestive canal determines its defensive capability.

We have previously demonstrated that Arabidopsis vegetative storage protein (AtVSP) is an acid phosphatase that has anti-insect activity in in vitro feeding assays [Liu et al., 2005. Plant Physiology 139, 1545-1556]. To investigate the functionality of AtVSP in planta as an anti-insect defense protein, we produced AtVSP-overexpressing as well as AtVSP-silenced transgenic Arabidopsis lines, and evaluated impact on the polyphagous American grasshopper Schistocerca americana. Grasshoppers showed no significant difference in weight gain and growth rate when feeding on wild type, overexpressing, or silenced lines, respectively. In addition, AtVSP protein was undetectable in either the midgut or frass of grasshoppers reared on transgenic plants suggesting that AtVSP was unable to withstand proteolytic degradation. To determine the stability of the AtVSP protein in grasshopper digestive canal, midgut extracts from various nymphal stages were incubated with bacterially expressed AtVSP for different periods of time. AtVSP was hydrolyzed rapidly by grasshopper midgut extract, in stark contrast with its fate when incubated with cowpea bruchid midgut extract. Multiple proteases have been detected in the midgut of grasshoppers, which may play important roles in determining the insect response to AtVSP. Results indicate that stability of an anti-insect protein in insect guts is a crucial property integral to the defense protein.