Phenotypic plasticity in heterotrophic marine microbial communities in continuous cultures

Abstract

Phenotypic plasticity (PP) is the development of alternate phenotypes of a given taxon as an adaptation to environmental conditions. Methodological limitations have restricted the quantification of PP to the measurement of a few traits in single organisms. We used metatranscriptomic libraries to overcome these challenges and estimate PP using the expressed genes of multiple heterotrophic organisms as a proxy for traits in a microbial community. The metatranscriptomes captured the expression response of natural marine bacterial communities grown on differing carbon resource regimes in continuous cultures. We found that taxa with different magnitudes of PP coexisted in the same cultures, and that members of the order Rhodobacterales had the highest levels of PP. In agreement with previous studies, our results suggest that continuous culturing may have specifically selected for taxa featuring a rather high range of PP. On average, PP and abundance changes within a taxon contributed equally to the organism's change in functional gene abundance, implying that both PP and abundance mediated observed differences in community function. However, not all functional changes due to PP were directly reflected in the bulk community functional response: gene expression changes in individual taxa due to PP were partly masked by counterbalanced expression of the same gene in other taxa. This observation demonstrates that PP had a stabilizing effect on a community's functional response to environmental change.

Figure 1

Transcripts binned according to function (a), taxon (b) or both function and taxon (c, d) in different treatments (indicated as x and y) were used to quantitatively estimate the community functional change (a), changes in the actively transcribing community composition (b), PP of individual taxa in size-normalized taxon bins (c) and TP in taxon bins with gene counts multiplied by the taxon transcript abundance before calculating Bray–Curtis dissimilarities (d).

Figure 2

Functional stability can be caused by two scenarios: (a) abundance changes of taxa that express the same function at the same relative transcription level (that is, taxa that are functionally redundant) and (b) changes in relative transcription level of the same function within taxa that do not change their overall transcript abundance (that is, taxa that feature PP). Numbers in brackets indicate the fraction of bars occupied by function α in order to authenticate the schematically displayed functional stability.

Figure 3

Dendrograms (Bray–Curtis dissimilarity, Ward's linkage) based on normalized count data after functional (a) and taxonomic (b) binning of the sequence data. Technical replicates from Exp_SwDi are labeled with filled circles and technical replicates from Exp_CyDi are labeled with open circles (modified from Beier et al., 2014).

Figure 4

Histogram displaying the variability of PP in 46 (a) and 99 (b) taxa (>2000 transcripts each) in Exp_SwDi and Exp_CyDi, respectively. The magnitude of PP was estimated by Bray–Curtis dissimilarities of functional bins for the same taxon in different treatments within an experiment. Increasing Bray–Curtis values indicate increasing PP.

Figure 5

Boxplot displaying PP of taxa from Exp_SwDi and Exp_CyDi after grouping taxa according to their phylogenetic order. Numbers in brackets following the order names indicate the number of taxa affiliated with the order. Orders with <5 taxa were excluded from analysis. Open circles indicate outliers with values exceeding by 1.5-fold the distance of the interquartile range.