Apex Predator Nematodes and Meso-Predator Bacteria Consume Their Basal Insect Prey through Discrete Stages of Chemical Transformations

Abstract

Microbial symbiosis drives physiological processes of higher-order systems, including the acquisition and consumption of nutrients that support symbiotic partner reproduction. Metabolic analytics provide new avenues to examine how chemical ecology, or the conversion of existing biomass to new forms, changes over a symbiotic life cycle. We applied these approaches to the nematode Steinernema carpocapsae, its mutualist bacterium, Xenorhabdus nematophila, and the insects they infect. The nematode-bacterium pair infects, kills, and reproduces in an insect until nutrients are depleted. To understand the conversion of insect biomass over time into either nematode or bacterium biomass, we integrated information from trophic, metabolomic, and gene regulation analyses. Trophic analysis established bacteria as meso-predators and primary insect consumers. Nematodes hold a trophic position of 4.6, indicative of an apex predator, consuming bacteria and likely other nematodes. Metabolic changes associated with Galleria mellonella insect bioconversion were assessed using multivariate statistical analyses of metabolomics data sets derived from sampling over an infection time course. Statistically significant, discrete phases were detected, indicating the insect chemical environment changes reproducibly during bioconversion. A novel hierarchical clustering method was designed to probe molecular abundance fluctuation patterns over time, revealing distinct metabolite clusters that exhibit similar abundance shifts across the time course. Composite data suggest bacterial tryptophan and nematode kynurenine pathways are coordinated for reciprocal exchange of tryptophan and NAD⁺ and for synthesis of intermediates that can have complex effects on bacterial phenotypes and nematode behaviors. Our analysis of pathways and metabolites reveals the chemistry underlying the recycling of organic material during carnivory. IMPORTANCE The processes by which organic life is consumed and reborn in a complex ecosystem were investigated through a multiomics approach applied to the tripartite Xenorhabdus bacterium-Steinernema nematode-Galleria insect symbiosis. Trophic analyses demonstrate the primary consumers of the insect are the bacteria, and the nematode in turn consumes the bacteria. This suggests the Steinernema-Xenorhabdus mutualism is a form of agriculture in which the nematode cultivates the bacterial food sources by inoculating them into insect hosts. Metabolomics analysis revealed a shift in biological material throughout progression of the life cycle: active infection, insect death, and conversion of cadaver tissues into bacterial biomass and nematode tissue. We show that each phase of the life cycle is metabolically distinct, with significant differences including those in the tricarboxylic acid cycle and amino acid pathways. Our findings demonstrate that symbiotic life cycles can be defined by reproducible stage-specific chemical signatures, enhancing our broad understanding of metabolic processes that underpin a three-way symbiosis.

Keywords: animal-microbe symbiosis; food web; interkingdom interactions; kynurenine; metabolomics; transcriptomics; trophic hierarchies; tryptophan.

FIG 1

Trophic analyses reveal Steinernema nematodes feed on Xenorhabdus bacteria in vitro (A) and in vivo (B). Trophic isoclines are represented via numeric TP_glu-phe ⁰/₀₀ ratios. Specific bacterial cultures or animals are displayed as the different shapes shown.

FIG 2

Identities and broad functional categorization of transcripts differentially expressed between X. nematophila mutants and their isogenic parent strains. (A) Quantification of the number of differentially expressed transcripts and how many are considered metabolic (light gray, with percentages listed), as determined by KEGG annotation. |Signal fold change| of >2 was used as a cutoff for significance. BlastKOALA functional categorizations of the differential metabolic transcripts are adjacent. The color legend is organized by having the most common category listed first. (B) Breakdown of the specific amino acid and carbohydrate metabolism pathways that were affected by the mutations, compared for each strain, with number of genes listed. The relative abundance (higher in mutant or higher in wild type) of transcript signals in each comparison set of strains is indicated on the x axis.

FIG 3

Key moments in the EPNB life cycle mapped onto important molecules are indicative of the bioconversion of the insect cadaver. Metabolites with the top 15 VIP scores of >1 for component 1 averaged relative abundances were grouped together into 3 categories: purine and pyrimidine components, amino acid components, and other important molecules. Asterisks next to the metabolites represent significant difference in metabolite abundance (P < 0.05) from t tests comparing the uninfected stage to the time stages (with red representing early phase, yellow representing middle phase, and green representing late phase). Relative metabolite abundance in log scale is displayed on the y axis of the line graphs, and dashed lines are used to define the early, middle, and late stages of infections as defined in this study, the general characteristics of which are described. Images of representative living, dead, spent, and water-trapped insects are shown, with lines connecting each to the approximate time frame of bioconversion they represent. In addition, representative images of nematodes are shown (not to scale). Images shown for days 4, 6, 8, and 10 were taken of mixed nematode populations from cadavers dissected at those time points of the metabolomics experiment. An example of an adult female nematode undergoing endotokia matricida, in which juvenile nematodes hatch within and consume the mother, is shown. Blue arrows are used to indicate adult female nematodes and juvenile nematodes in select images.

FIG 4

Distinct chemical environments occur during bioconversion of an insect cadaver by S. carpocapsae and X. nematophila. Three-dimensional PLS-DA of time course infection metabolic profiles grouped according to stage of infection: uninfected (black) and early (red gradient), middle (yellow gradient), and late infected (green gradient) insects. Components contributing to the separation of the profiles are listed (in percentages) on the axes.

FIG 5

Schematic of the combined tryptophan/kynurenine metabolic pathway predicted for S. carpocapsae and X. nematophila. Predicted metabolic pathways of X. nematophila (green), S. carpocapsae (orange), or other organisms (black) are shown. Reactions and the enzymes predicted to catalyze them are indicated with arrows and protein names, respectively. Dashed block arrows indicate multiple pathway steps, and solid block arrows indicate a single pathway reaction. Fold changes in transcript abundances identified through the X. nematophila bacteria microarray analysis reported here (Data S2) or a publicly available S. carpocapsae nematode transcriptome analysis (Table 3) (75) are indicated in parentheses below the relevant protein name. For bacterial microarray data, shown are the fold differences of wild-type (or 1°) transcript abundance compared to the comparison strain noted. For S. carpocapsae, the ratio of transcript detected in the head versus the tail is shown. Metabolites indicated with red and blue font are those that were significantly elevated or reduced, respectively in the indicated stage (early, middle, or late; see the text) relative to the prior stage (see Fig. 3 and 6). ANT, anthranliate; CPAD5P, 1-(2-carboxyphenylamino)-1′-deoxy-d-ribulose 5-phosphate; CRSMT, chorismate; DHB, 2,3-dihydroxybenzoate; DHSK, 3-dehydroshikimate; DQT, 3-dehydroquinate; 3HAA, 3-hydroxyanthranilic acid; 3HK, 3-hydroxykynurenine; IND, indole; InGP, (3-indoyl)-glycerolphosphate; KA, kynurenic acid; KYN, kynurenine; NAD⁺, NAD; NFK, N-formylkynurenine; NPRAN, N-5-phospho-beta-d-ribosyl-anthranilate; PHE, phenylalanine; PRPP, phosphoribosyl pyrophosphate; QA, quinolinic acid; SME, shikimate; TRP, tryptophan; TYR, tyrosine.

FIG 6

Hierarchical clustering analysis of detected metabolites revealed 10 clusters, within each of which the metabolites displayed similar rates of change over the infection. (A) Dendrogram corresponding to Spearman correlation values for each metabolite. (B) Identified metabolite clusters as depicted by numbers 1 to 10 with similar log(rate of change) over the life cycle for the time phase compared to the previous time phase: early compared to uninfected (early), middle compared to early (middle), and late compared to middle (late). Metabolites with a red gradient exhibit an increased molecular abundance shift, and metabolites with a blue gradient exhibit a decreased molecular abundance shift.