Variable microbiomes between mosquito lines are maintained across different environments

Abstract

The composition of the microbiome is shaped by both environment and host in most organisms, but in the mosquito Aedes aegypti the role of the host in shaping the microbiome is poorly understood. Previously, we had shown that four lines of Ae. aegypti harbored different microbiomes when reared in the same insectary under identical conditions. To determine whether these lines differed from each other across time and in different environments, we characterized the microbiome of the same four lines of Ae. aegypti reared in the original insectary and at another institution. While it was clear that the environment influenced the microbiomes of these lines, we did still observe distinct differences in the microbiome between lines within each insectary. Clear differences were observed in alpha diversity, beta diversity, and abundance of specific bacterial taxa. To determine if the line specific differences in the microbiome were maintained across environments, pair-wise differential abundances of taxa was compared between insectaries. Lines were most similar to other lines from the same insectary than to the same line reared in a different insectary. Additionally, relatively few differentially abundant taxa identified between pairs of lines were shared across insectaries, indicating that line specific properties of the microbiome are not conserved across environments, or that there were distinct microbiota within each insectary. Overall, these results demonstrate that mosquito lines under the same environmental conditions have different microbiomes across microbially- diverse environments and host by microbe interactions affecting microbiome composition and abundance is dependent on environmentally available bacteria.

Fig 1. Diversity of the microbiome in individuals reared at the LSTM or UTMB insectaries.

Structure of bacterial communities was determined by deep sequencing the V3-V4 region of the bacterial 16S gene in adults from 4 different lines of Ae aegypti (Galveston, Thailand, Iquitos, Juchitan) reared in two different insectaries at UTMB and LSTM. The Bacterial community structure is represented (A) by the species richness index Chao1 and (B) by principal component analysis of Bray-Curtis dissimilarity index. Mosquito lines reared at LSTM are shown in purple, mosquitoes reared to UTMB are shown in turquoise.

Fig 2. Alpha and beta diversity metrics for the Ae. aegypti lines reared at the LSTM and UTMB insectaries.

(A, C): The species richness (Chao1 index) was calculated from 20 individuals from each line (Galveston, Thailand, Iquitos, Juchitan) at each insectary. The level of species richness differed between individuals from the LSTM (ANOVA, p-value < 0.001) and UTMB (ANOVA, p-value < 0.001). (B,D):The dissimilarities between the 4 different lines of Ae. aegypti was analyzed by principal component analysis of Bray-Curtis dissimilarity index. The bacterial community structure of the lines differed in individuals from LSTM (PERMANOVA, p-value = 0.001) and UTMB (PERMANOVA, p-value = 0.001).

Fig 3. Upset plot showing the number of genera that are shared between lines in the LSTM and UTMB insectaries.

Each mosquito strain in the different insectary (one per row at the bottom half of the image) is treated as a ’set’ with an identified number of bacterial taxa (’Set Size’). The various permutations of intersections are denoted by the ball-and-stick diagram at the bottom of the image, and size of these intersections denoted by the bar graph at the top of the image (’Intersection Size’). Rows are colored based on mosquito strain (middle-left), and further divided into the rearing Insectary (bottom left). 23 taxa are shared across all the different strains between the different insectaries, constituting a potential ’core’ set of bacteria. This is then followed by five (Persicitalea, Janthinobacterium, Rahnella, Luteolibacter, Verrucomicrobium) and three taxa (Sphingopyxis, Burkholderia-Caballeronia-Paraburkholderia, Methyloversatilis) that appeared unique to the rearing insectary. All other permutations of intersects contain two or fewer bacterial taxa.

Fig 4. Relative abundance of bacteria in each line.

The dominant bacterial genera are different between insectaries and between lines. The relative abundance of the 20 most abundant genera in shown for 20 individuals from each line at LSTM and UTMB insectaries. Bacterial genera were assigned to OTUs clustered with a 97% sequence identity cutoff and taxonomically classified with the SILVA database.

Fig 5. Pairwise comparisons of mosquito lines reared in each insectary.

The microbiomes of mosquito lines reared in the same insectary are more similar compared to those reared in a different insectary. Results from a pairwise differential abundance analysis are shown for each pair of lines as the percent of genera that are significantly different between the pairs after correcting for multiple comparisons. A light blue color indicates a higher degree of dissimilarity between the lines.

Fig 6. Few genera have conserved differences in abundance between lines at both insectaries.

Venn diagrams show the overlap in specific genera that are differentially abundant between each pair of lines (Galveston, Iquitos, Thailand, or Juchitan) at each insectary (LSTM or UTMB). The ID of genera differentially abundant between each pairwise comparison was compared between insectaries. The shared genera represent a genus that is differentially abundant between the lines in both environments. The ID of shared genera can be found in S3 Table.