Phytoplankton release massive amounts of dissolved organic matter (DOM) into the water column during recurring blooms in coastal waters and inland seas. The released DOM encompasses a complex mixture of both known and unknown compounds, and is a rich nutrient source for heterotrophic bacteria. The metabolic activity of bacteria during and after phytoplankton blooms can hence be expected to reflect the characteristics of the released DOM. We therefore investigated if bacterioplankton could be used as "living sensors" of phytoplankton DOM quantity and/or quality, by applying gene expression analyses to identify bacterial metabolisms induced by DOM. We used transcriptional analysis of two Baltic Sea bacterial isolates (Polaribacter sp. BAL334 [Flavobacteriia] and Brevundimonas sp. BAL450 [Alphaproteobacteria]) growing with DOM from axenic cultures of the dinoflagellate Prorocentrum minimum. We observed pronounced differences between the two bacteria both in growth and the expressed metabolic pathways in cultures exposed to dinoflagellate DOM compared with controls. Differences in metabolic responses between the two isolates were caused both by differences in gene repertoire between them (e.g. in the SEED categories for membrane transport, motility and photoheterotrophy) and the regulation of expression (e.g. fatty acid metabolism), emphasizing the importance of separating the responses of different taxa in analyses of community sequence data. Similarities between the bacteria included substantially increased expression of genes for Ton and Tol transport systems in both isolates, which are commonly associated with uptake of complex organic molecules. Polaribacter sp. BAL334 showed stronger metabolic responses to DOM harvested from exponential than stationary phase dinoflagellates (128 compared to 26 differentially expressed genes), whereas Brevundimonas sp. BAL450 responded more to the DOM from stationary than exponential phase dinoflagellates (33 compared to 6 differentially expressed genes). These findings suggest that shifts in bacterial metabolisms during different phases of phytoplankton blooms can be detected in individual bacterial species and can provide insights into their involvement in DOM transformations.

Neurotoxic methylmercury (MeHg) is formed by microbial methylation of inorganic divalent Hg (Hg-II) and constitutes severe environmental and human health risks. The methylation is enabled by hgcA and hgcB genes, but it is not know nif the associated molecular-level processes are rate-limiting or enable accurate prediction of MeHg formation in nature. In this study, we investigated the relationships between hgc genes and MeHg across redox-stratified water columns in the brackish Baltic Sea. We showed, for the first time, that hgc transcript abundance and the concentration of dissolved Hg-II-sulfide species were strong predictors of both the Hg-II methylation rate and MeHg concentration, implying their roles as principal joint drivers of MeHg formation in these systems. Additionally, we characterized the metabolic capacities of hgc(+) microorganisms by reconstructing their genomes from metagenomes (i.e., hgc(+) MAGs), which highlighted the versatility of putative Hg-II methylators in the water column of the Baltic Sea. In establishing relationships between hgc transcripts and the Hg-II methylation rate, we advance the fundamental understanding of mechanistic principles governing MeHg formation in nature and enable refined predictions of MeHg levels in coastal seas in response to the accelerating spread of oxygen-deficientzones.

Organic pollutants (OPs) are critically toxic, bioaccumulative and globally widespread. Moreover, several OPs negatively influence aquatic wildlife. Although bacteria are major drivers of the ocean carbon cycle and the turnover of vital elements, there is limited knowledge of OP effects on heterotrophic bacterioplankton. We therefore investigated growth and gene expression responses of the Baltic Sea model bacterium Rheinheimera sp. BAL341 to environmentally relevant concentrations of distinct classes of OPs in 2-h incubation experiments. During exponential growth, exposure to a mix of polycyclic aromatic hydrocarbons, alkanes and organophosphate esters (denoted MIX) resulted in a significant decrease (between 9% and 18%) in bacterial abundance and production compared with controls. In contrast, combined exposure to perfluorooctanesulfonic acids and perfluorooctanoic acids (denoted PFAS) had no significant effect on growth. Nevertheless, MIX and PFAS exposures both induced significant shifts in gene expression profiles compared with controls in exponential growth. This involved several functional metabolism categories (e.g. stress response and fatty acids metabolism), some of which were pollutant-specific (e.g. phosphate acquisition and alkane-1 monooxygenase genes). In stationary phase, only two genes in the MIX treatment were significantly differentially expressed. The substantial direct influence of OPs on metabolism during bacterial growth suggests that widespread OPs could severely alter biogeochemical processes governed by bacterioplankton.

Phytoplankton release massive amounts of dissolved organic matter (DOM) into the water column during recurring blooms in coastal waters and inland seas. The released DOM includes dissolved organic carbon, nitrogen and phosphorus, in a complex mixture of both known and unknown compounds, and is a rich nutrient source for heterotrophic bacteria. The metabolic activity of heterotrophic bacteria during and after phytoplankton blooms can hence be expected to reflect the characteristics of the released DOM. With this in mind, we wanted to investigate if bacterioplankton could be used as “living sensors” of phytoplankton DOM quantity and quality, and to trace the flow of nutrients in the ecosystem. We used transcriptional activity from Baltic Sea bacterial isolates (Polaribacter sp. BAL334 (Flavobacteriia) and Brevundimonas sp. BAL450 (Alphaproteobacteria)) exposed to DOM derived from the dinoflagellate Prorocentrum minimum in exponential and stationary growth phases respectively. We observed strong responses both in terms of physiology – bacterial abundance – and the expressed metabolic pathways – e.g. Membrane Transport, Fatty Acids, Lipids and Isoprenoids – of the populations in samples exposed to dinoflagellate DOM compared with controls. Particularly striking was the increased expression of Ton and Tol transport systems, commonly associated with uptake of complex molecules, in both isolates. Equally important were the differences in metabolic responses between the two isolates, caused by differences in gene repertoire between them, emphasizing the importance of separating the responses of different taxa in analyses of community sequence data. Differences in response to DOM sourced from exponentially and stationary growing dinoflagellates were less pronounced, although not absent, than differences between the bacterial isolates. This suggests that shifts in metabolism during the different phases of a phytoplankton bloom might be detectable in individual bacterial populations. To conclude, our work opened a door to the future use of bacterioplankton as living sensors of environmental status, particularly with respect to phytoplankton blooms.