Gut Bacteria Associated With Monochamus saltuarius (Coleoptera: Cerambycidae) and Their Possible Roles in Host Plant Adaptations

Abstract

Monochamus saltuarius (Coleoptera: Cerambycidae) is an important native pest in the pine forests of northeast China and a dispersing vector of an invasive species Bursaphelenchus xylophilus. To investigate the bacterial gut diversity of M. saltuarius larvae in different host species, and infer the role of symbiotic bacteria in host adaptation, we used 16S rRNA gene Illumina sequencing and liquid chromatography-mass spectrometry metabolomics processing to obtain and compare the composition of the bacterial community and metabolites in the midguts of larvae feeding on three host tree species: Pinus koraiensis, Pinus sylvestris var. mongolica, and Pinus tabuliformis. Metabolomics in xylem samples from the three aforementioned hosts were also performed. Proteobacteria and Firmicutes were the predominant bacterial phyla in the larval gut. At the genus level, Klebsiella, unclassified_f__Enterobacteriaceae, Lactococcus, and Burkholderia-Caballeronia-Paraburkholderia were most dominant in P. koraiensis and P. sylvestris var. mongolica feeders, while Burkholderia-Caballeronia-Paraburkholderia, Dyella, Pseudoxanthomonas, and Mycobacterium were most dominant in P. tabuliformis feeders. Bacterial communities were similar in diversity in P. koraiensis and P. sylvestris var. mongolica feeders, while communities were highly diverse in P. tabuliformis feeders. Compared with the other two tree species, P. tabuliformis xylems had more diverse and abundant secondary metabolites, while larvae feeding on these trees had a stronger metabolic capacity for secondary metabolites than the other two host feeders. Correlation analysis of the association of microorganisms with metabolic features showed that dominant bacterial genera in P. tabuliformis feeders were more negatively correlated with plant secondary metabolites than those of other host tree feeders.

Keywords: borer; host adaptation; host-microbe interaction; intestinal bacterial composition; metabolomics; microbiota.

FIGURE 1

Venn diagrams of OTUs shared between intestines of larvae feeding on the three host tree species. PK, midguts of larvae feeding on P. koraiensis; PT, midguts of larvae feeding on P. tabuliformis; PS, midguts of larvae feeding on P. sylvestris var. mongolica. All of these abbreviations apply to the following figures.

FIGURE 2

α-diversity and β-diversity of midgut bacteria in M. saltuarius feeding on different host trees. (A) Significant differences in the Chao index (richness estimator). (B) Significant differences in the Shannon index (diversity estimator). (C) Significant differences in the Simpson index (diversity estimator) (Student’s t-test; * P < 0.05, **P < 0.01, ***P < 0.001). (D) Principal coordinate analysis based on bray_curtis distances generated from OTU tables. Ovals of different colors represent different groupings (adonis; P = 0.001).

FIGURE 3

Taxonomic composition of highly abundant bacterial phyla and genera in the gut microbiota of larvae feeding on different host trees. (A) Relative abundance of dominant microbial phyla (abundance ≥ 1%). (B) Relative abundance of dominant microbial genera (abundance ≥ 2%). The relative percent abundance of bacterial genera is represented by different colors.

FIGURE 4

Differences in intestinal bacteria of M. saltuarius feeding on different host tree species at the phylum and genus level. (A) Differences in abundance of dominant bacterial phyla (the top five are shown). (B) Differences in abundance of dominant bacterial genera (the top ten are shown) (Kruskal–Wallis H-test; *P < 0.05, **P < 0.01, *** P < 0.001).

FIGURE 5

Comparison of predicted KEGG ortholog group counts at level 3 and enzyme levels in different host tree species. (A) Heat map of the most abundant 30 categories at level 3; each column corresponds to an M. saltuarius sample, and each row corresponds to a specific category. (B) Relative abundance of three highly abundant predicted KEGG enzymes involved in cellulose or limonene and pinene degradation in the guts of larvae feeding on different host tree species (one-way ANOVA; *P < 0.05, **P < 0.01, ***P < 0.001).

FIGURE 6

Comparison of relative metabolite abundance in wood tissue from different host tree species. (A) Heat map of metabolites annotated in the pathway “metabolism of terpenoids and polyketides.” (B) Heat map of metabolites annotated in the pathway “biosynthesis of other secondary metabolites.” Each column corresponds to a host tree wood tissue sample, and each row corresponds to a specific metabolite. T_PK, xylem tissue from P. koraiensis; T_PT, xylem tissue from P. tabuliformis; T_PS, xylem tissue from P. sylvestris var. mongolica. All of these abbreviations apply to the following figures.

FIGURE 7

Comparison of relative abundances of compounds annotated in three KEGG pathways. (A) Wood tissue samples. (B) Larval gut samples. BSM, biosynthesis of other secondary metabolites; MTP, metabolism of terpenoids and polyketides. Relative abundances were calculated by the sum normalization method.

FIGURE 8

Correlation analysis between genera abundant in the gut and gut metabolite contents annotated as putative plant defense substances. (A) Correlations between the top five abundant genera in larvae feeding on all host tree species and compounds annotated as related to plant defense substances. (B) Correlations between abundant genera in P. tabuliformis feeders (abundance ≥ 2%, except for the top five) and putative plant defense metabolites. Each row in the figure represents a metabolite, each column represents a genus, and each lattice represents a Pearson correlation coefficient between a component and a metabolite. Red represents positive correlations, while blue represents negative correlations (*P < 0.05, **P < 0.01, ***P < 0.001).