Insights into the potential for mutualistic and harmful host-microbe interactions affecting brown alga freshwater acclimation

Abstract

Microbes can modify their hosts' stress tolerance, thus potentially enhancing their ecological range. An example of such interactions is Ectocarpus subulatus, one of the few freshwater-tolerant brown algae. This tolerance is partially due to its (un)cultivated microbiome. We investigated this phenomenon by modifying the microbiome of laboratory-grown E. subulatus using mild antibiotic treatments, which affected its ability to grow in low salinity. Low salinity acclimation of these algal-bacterial associations was then compared. Salinity significantly impacted bacterial and viral gene expression, albeit in different ways across algal-bacterial communities. In contrast, gene expression of the host and metabolite profiles were affected almost exclusively in the freshwater-intolerant algal-bacterial communities. We found no evidence of bacterial protein production that would directly improve algal stress tolerance. However, vitamin K synthesis is one possible bacterial service missing specifically in freshwater-intolerant cultures in low salinity. In this condition, we also observed a relative increase in bacterial transcriptomic activity and the induction of microbial genes involved in the biosynthesis of the autoinducer AI-1, a quorum-sensing regulator. This could have resulted in dysbiosis by causing a shift in bacterial behaviour in the intolerant algal-bacterial community. Together, these results provide two promising hypotheses to be examined by future targeted experiments. Although they apply only to the specific study system, they offer an example of how bacteria may impact their host's stress response.

Keywords: brown algae; holobiont; host-microbiome interaction; low salinity acclimation; meta-transcriptomics; metabolic networks; metabolite profiling; virome.

FIGURE 1

Overview of the experimental setup comprising three different microbial communities and two salinity conditions. All conditions were derived from the same starter culture, but antibiotic (ATB) treatments were carried out with three different ATB mixes (see Section 2), leaving hosts with different microbial communities and different salinity tolerance, termed “microbial communities (MC) 1–3”. Pie charts in the metabarcoding part show the ten most abundant genera in each microbial community. The low‐salinity response of 5 replicate algae with each of these communities was examined. 100% NSW, natural seawater; 15% NSW, 15% NSW in distilled H₂O (volume/volume); m, month(s); w, week(s)

FIGURE 2

Overview of metatranscriptomic and metagenomic data obtained and the analysis pipeline. Despite extensive sequencing efforts, only 0.7% of reads mapped to bacterial genomes assembled from the metagenome. The percentages correspond to the percentage of total raw reads at the start

FIGURE 3

(a) Principal component analysis (PCA) of algal gene expression. (b) PCA of expression of bacterial reactions. (c) PCA of viral transcripts. Samples with MC3 differ from samples with MC1 and MC2 in all three panels, but samples with MC1 and MC2 differ mainly regarding bacterial gene expression.

FIGURE 4

(a) Differentially expressed algal genes. The figure shows the number of differentially expressed algal genes with each of the microbial communities in 15% NSW compared to 100% NSW (MC1, MC2, and MC3); algae with MC1 + MC2 jointly compared to those with MC3 in 100% NSW (microbiome effect), the differences between algae with MC1 and MC2, and the difference in the low salinity‐response of algae with MC3 compared to those with MC1+ MC2 (interaction term, crossed arrows). Numbers in parentheses correspond to the number of overrepresented GO terms associated with the differentially regulated genes. (b) Similar analysis as A, but on the bacterial transcriptome; in this case, the analysis was based on differentially expressed metabolic reactions rather than genes. (c) Similar analysis as (a) and (b), but concerning viral sequences. Samples with MC3 exhibited the strongest host response, while the bacterial responses were equally pronounced for all three MCs

FIGURE 5

Heat map based on hierarchical clustering of normalized transcriptional activity per bacterial bin in each microbial community and condition from Table S8. The figure shows that different bacteria were active, depending on microbial community and, to a lesser extent, salinity. Clustering was based on the Pearson correlation coefficient using the average linkage method. Unit variance scaling was applied to row data, that is, for each row, the average was adjusted to 0 and the standard deviation to 1. Red indicates high relative transcriptomic activity in the given condition, and blue indicates low activity. “15%” corresponds to the treatment with 15% natural seawater (NSW), “100%” to the treatment with 100% NSW. Bins labelled “full” are predicted to be ≥90% complete, and bins labelled “partial” < 90%. The predicted completeness in % is given after the “c” in the bin name

FIGURE 6

Box plot of relative abundances of viral sequences belonging to different families in all samples (MC1–MC3 = microbial community 1–3) showing the variability of the viral transcript abundance depending on microbial community and salinity. Values are given in % of the total number of viral reads per sample. p‐values correspond to the results of an ANOVA across all six conditions (4 replicates each). Lowercase letters indicate significant differences between treatments (Tukey's HSD test). 15% corresponds to the treatment with 15% NSW, 100% to the treatment with 100% NSW

FIGURE 7

Upset plot of (a) metabolites predicted to be producible by the active bacterial communities in the different conditions, and (b) metabolites predicted to be newly producible by the algal host if given access to the metabolites producible by the bacteria. The bars represent the number of metabolites shared between the different communities and conditions. Most metabolites are producible in all communities, but MC3–15% has the lowest number of producible metabolites. A detailed list of metabolites is available in Table S11

FIGURE 8

Heat map based on the abundance of each metabolite tested as significant in at least one condition. It shows clear differences in metabolite profiles obtained by GC–MS analysis depending on microbial community and for algae with MC3 depending on salinity. Clustering was based on the Pearson correlation coefficient using the average linkage method. Unit variance scaling was applied to row data, that is, for each row, the average was adjusted to 0 and the standard deviation to 1. 15% corresponds to the treatment with 15% NSW, 100% to the treatment with 100% NSW. “?” indicates features with reverse match score (R) < 800, “??” features with R < 700. Features were labelled “unknown” for R < 600