Biotype and host relatedness influence the composition of bacterial microbiomes in Schizaphis graminum aphids

Abstract

Introduction: The microbiome of greenbug aphid (Schizaphis graminum (Rondani)) was investigated in regard to greenbug biotype, collection date, host species, and host cultivar.

Methods: DNA samples were collected from biotypes E and K feeding on 17 cultivars belonging to five host plant species, namely wheat, barley, rye, sorghum, and the goatgrass Aegilops triuncialis. Samples were taken immediately before infestation and two, four, and eight days thereafter. The V5-V7 hypervariable region of 16S rDNA was PCR amplified, Illumina sequenced, and aligned to a curated database of bacterial 16S rDNA sequences.

Results and discussion: The almost universal intracellular endosymbiont of aphids, Buchnera aphidicola, comprised 78.24 to 99.99% of the read counts among samples, largely because of its high copy number of genomes per bacteroid. Abundant non-Buchnera genera included Pseudomonas, Rhodanobacter, Massilia, and Enterobacter. Read counts of eight of 78 examined genera were more than 90% restricted to a single replicate of a single treatment. Shannon entropy was highest in biotype K and on the barley host, but it did not vary significantly among dates post infestation. Unweighted UniFrac distances most significantly varied with biotype, host plant species, infestation time, and almost all of their interactions. Weighted UniFrac and Jaccard distances varied less significantly. By counts of differentially populated genera, the factors biotype, host plant species, infestation time, and host plant resistance genes to greenbug, were consecutively less important. Functional analysis with PICRUSt2 illustrated a diminution of respiratory electron transport and long-chain fatty acids in the Buchnera endosymbiont, reflecting adaptation to an intracellular environment.

Keywords: 16S rRNA; RNA-seq; Schizaphis graminum; gut microflora; microbiome; symbiont.

Figure 1

Rarefaction survey of OTU richness, subsampling from 25,000 to 850,000 reads in increments of 25,000 reads.

Figure 2

Barplot of 30 most abundant genera sorted by greenbug biotype and days post infestation.

Figure 3

Barplot of 30 most abundant genera sorted by host plant species.

Figure 4

Barplot of 30 most abundant genera sorted by host plant species and cultivar.

Figure 5

Spearman’s correlation of observed relative abundance among species. More positive values indicate greater strength of association. Significance testing was also performed for each pair of taxa, and asterisks indicate: ^*p < 0.05, ^**p < 0.01, and ^***p < 0.001.

Figure 6

Zero-radius OTUs observed within the total microbiome data, grouped and colored by host plant species. The median value and first and third quartiles in each group are also illustrated. Asterisks indicate statistical significance: *p < 0.05; **p < 0.01; ***p < 0.001; ns, p > 0.05.

Figure 7

Principal coordinates analysis based on unweighted UniFrac distances distinguished by greenbug biotype. The distributions for biotypes E and K are mostly encircled.