Internal carbon recycling by heterotrophic prokaryotes compensates for mismatches between phytoplankton production and heterotrophic consumption.

Molecular observational tools are great for characterizing the composition and genetic endowment of microbial communities, but cannot measure fluxes, which are critical for the understanding of ecosystems. To overcome these limitations, we use a mechanistic inference approach to estimate dissolved organic carbon (DOC) production and consumption by phytoplankton operational taxonomic units (OTUs) and heterotrophic prokaryotic amplicon sequences variants (ASVs), and infer carbon fluxes between members of this microbial community from Western English Channel time-series data. Our analyses focus on phytoplankton spring and summer blooms, as well as bacteria summer blooms. In spring blooms, phytoplankton DOC production exceeds heterotrophic prokaryotic consumption, but in bacterial summer blooms heterotrophic prokaryotes consume 3 times more DOC than produced by the phytoplankton. This mismatch is compensated by heterotrophic prokaryotic DOC release by death, presumably from viral lysis. In both types of summer blooms, large amounts of the DOC liberated by heterotrophic prokaryotes are re-used through internal recycling., with fluxes between different heterotrophic prokaryotes being at the same level as those between phytoplankton and heterotrophic prokaryotes. Contextualized, internal recycling accounts for approximately 75% and 30% of the estimated net primary production (0.16 vs 0.22 and 0.08 vs 0.29 µmol l-1 d-1) in bacteria and phytoplankton summer blooms, respectively, and thus represents a major component of the Western English Channel carbon cycle. We conclude that internal recycling compensates for mismatches between phytoplankton DOC production and heterotrophic prokaryotic consumption, and we encourage future analyses on aquatic carbon cycles to consider fluxes between heterotrophic prokaryotes, specifically internal recycling.

Genome reduction occurred in early Prochlorococcus with an unusually low effective population size.

In the oligotrophic sunlit ocean, the most abundant free-living planktonic bacterial lineages evolve convergently through genome reduction. The cyanobacterium Prochlorococcus responsible for 10% global oxygen production is a prominent example. The dominant theory known as "genome streamlining" posits that they have extremely large effective population sizes (Ne) such that selection for metabolic efficiency acts to drive genome reduction. Because genome reduction largely took place anciently, this theory builds on the assumption that their ancestors' Ne was similarly large. Constraining Ne for ancient ancestors is challenging because experimental measurements of extinct organisms are impossible and alternatively reconstructing ancestral Ne with phylogenetic models gives large uncertainties. Here, we develop a new strategy that leverages agent-based modeling to simulate the changes in the genome-wide ratio of radical to conservative nonsynonymous nucleotide substitution rate (dR/dC) in a possible range of Ne in ancestral populations. This proxy shows expected increases with decreases of Ne only when Ne falls to about 10 k - 100 k or lower, magnitudes characteristic of Ne of obligate endosymbiont species where drift drives genome reduction. Our simulations therefore strongly support a scenario where the primary force of Prochlorococcus genome reduction is drift rather than selection.

Microcystin aids in cold temperature acclimation: Differences between a toxic Microcystis wildtype and non-toxic mutant.

For Microcystis aeruginosa PCC 7806, temperature decreases from 26 °C to 19 °C double the microcystin quota per cell during growth in continuous culture. Here we tested whether this increase in microcystin provided M. aeruginosa PCC 7806 with a fitness advantage during colder-temperature growth by comparing cell concentration, cellular physiology, reactive oxygen species damage, and the transcriptomics-inferred metabolism to a non-toxigenic mutant strain M. aeruginosa PCC 7806 ΔmcyB. Photo-physiological data combined with transcriptomic data revealed metabolic changes in the mutant strain during growth at 19 °C, which included increased electron sinks and non-photochemical quenching. Increased gene expression was observed for a glutathione-dependent peroxiredoxin during cold treatment, suggesting compensatory mechanisms to defend against reactive oxygen species are employed in the absence of microcystin in the mutant. Our observations highlight the potential selective advantages of a longer-term defensive strategy in management of oxidative stress (i.e., making microcystin) vs the shorter-term proactive strategy of producing cellular components to actively dissipate or degrade oxidative stress agents.

Cool temperature acclimation in toxigenic Microcystis aeruginosa PCC 7806 and its non-toxigenic mutant.

For Microcystis aeruginosa PCC 7806, temperature decreases from 26° C to 19° C double the microcystin quota per cell during growth in continuous culture. Here we tested whether this increase in microcystin provided M. aeruginosa PCC 7806 with a fitness advantage during colder-temperature growth by comparing cell concentration, cellular physiology, and the transcriptomics-inferred metabolism to a non-toxigenic mutant strain M. aeruginosa PCC 7806 ΔmcyB. Photo-physiological data combined with transcriptomic data revealed metabolic changes in the mutant strain during growth at 19° C, which included increased electron sinks and non-photochemical quenching. Increased gene expression was observed for a glutathione-dependent peroxiredoxin during cold treatment, suggesting compensatory mechanisms to defend against reactive oxygen species are employed in the absence of microcystin in the mutant. Our observations highlight the potential selective advantages of a longer-term defensive strategy in management of oxidative stress (i.e., making microcystin) vs the shorter-term proactive strategy of producing cellular components to actively dissipate or degrade oxidative stress agents.

A Reduction of Transcriptional Regulation in Aquatic Oligotrophic Microorganisms Enhances Fitness in Nutrient-Poor Environments.

In this review, we consider the regulatory strategies of aquatic oligotrophs, microbial cells that are adapted to thrive under low-nutrient concentrations in oceans, lakes, and other aquatic ecosystems. Many reports have concluded that oligotrophs use less transcriptional regulation than copiotrophic cells, which are adapted to high nutrient concentrations and are far more common subjects for laboratory investigations of regulation. It is theorized that oligotrophs have retained alternate mechanisms of regulation, such as riboswitches, that provide shorter response times and smaller amplitude responses and require fewer cellular resources. We examine the accumulated evidence for distinctive regulatory strategies in oligotrophs. We explore differences in the selective pressures copiotrophs and oligotrophs encounter and ask why, although evolutionary history gives copiotrophs and oligotrophs access to the same regulatory mechanisms, they might exhibit distinctly different patterns in how these mechanisms are used. We discuss the implications of these findings for understanding broad patterns in the evolution of microbial regulatory networks and their relationships to environmental niche and life history strategy. We ask whether these observations, which have emerged from a decade of increased investigation of the cell biology of oligotrophs, might be relevant to recent discoveries of many microbial cell lineages in nature that share with oligotrophs the property of reduced genome size.

Differences in the regulatory strategies of marine oligotrophs and copiotrophs reflect differences in motility.

Aquatic bacteria frequently are divided into lifestyle categories oligotroph or copiotroph. Oligotrophs have proportionately fewer transcriptional regulatory genes than copiotrophs and are generally non-motile/chemotactic. We hypothesized that the absence of chemotaxis/motility in oligotrophs prevents them from occupying nutrient patches long enough to benefit from transcriptional regulation. We first confirmed that marine oligotrophs are generally reduced in genes for transcriptional regulation and motility/chemotaxis. Next, using a non-motile oligotroph (Ca. Pelagibacter st. HTCC7211), a motile copiotroph (Alteromonas macleodii st. HOT1A3), and [14 C]l-alanine, we confirmed that l-alanine catabolism is not transcriptionally regulated in HTCC7211 but is in HOT1A3. We then found that HOT1A3 took 2.5-4 min to initiate l-alanine oxidation at patch l-alanine concentrations, compared to <30 s for HTCC7211. By modelling cell trajectories, we predicted that, in most scenarios, non-motile cells spend <2 min in patches, compared to >4 min for chemotactic/motile cells. Thus, the time necessary for transcriptional regulation to initiate prevents transcriptional regulation from being beneficial for non-motile oligotrophs. This is supported by a mechanistic model we developed, which predicted that HTCC7211 cells with transcriptional regulation of l-alanine metabolism would produce 12% of their standing ATP stock upon encountering an l-alanine patch, compared to 880% in HTCC7211 cells without transcriptional regulation.

Response to Comment on "Models predict planned phosphorus load reduction will make Lake Erie more toxic".

Huisman et al. claim that our model is poorly supported or contradicted by other studies and the predictions are "seriously flawed." We show their criticism is based on an incomplete selection of evidence, misinterpretation of data, or does not actually refute the model. Like all ecosystem models, our model has simplifications and uncertainties, but it is better than existing approaches hat ignore biology and do not predict toxin concentration.

Models predict planned phosphorus load reduction will make Lake Erie more toxic.

Harmful cyanobacteria are a global environmental problem, yet we lack actionable understanding of toxigenic versus nontoxigenic strain ecology and toxin production. We performed a large-scale meta-analysis including 103 papers and used it to develop a mechanistic, agent-based model of Microcystis growth and microcystin production. Simulations for Lake Erie suggest that the observed toxigenic-to-nontoxigenic strain succession during the 2014 Toledo drinking water crisis was controlled by different cellular oxidative stress mitigation strategies (protection by microcystin versus degradation by enzymes) and the different susceptibility of those mechanisms to nitrogen limitation. This model, as well as a simpler empirical one, predicts that the planned phosphorus load reduction will lower biomass but make nitrogen and light more available, which will increase toxin production, favor toxigenic cells, and increase toxin concentrations.

SAR11 Cells Rely on Enzyme Multifunctionality To Metabolize a Range of Polyamine Compounds.

In the ocean surface layer and cell culture, the polyamine transport protein PotD of SAR11 bacteria is often one of the most abundant proteins detected. Polyamines are organic cations at seawater pH produced by all living organisms and are thought to be an important component of dissolved organic matter (DOM) produced in planktonic ecosystems. We hypothesized that SAR11 cells uptake and metabolize multiple polyamines and use them as sources of carbon and nitrogen. Metabolic footprinting and fingerprinting were used to measure the uptake of five polyamine compounds (putrescine, cadaverine, agmatine, norspermidine, and spermidine) in two SAR11 strains that represent the majority of SAR11 cells in the surface ocean environment, "Candidatus Pelagibacter" strain HTCC7211 and "Candidatus Pelagibacter ubique" strain HTCC1062. Both strains took up all five polyamines and concentrated them to micromolar or millimolar intracellular concentrations. Both strains could use most of the polyamines to meet their nitrogen requirements, but polyamines did not fully substitute for their requirements of glycine (or related compounds) or pyruvate (or related compounds). Our data suggest that potABCD transports all five polyamines and that spermidine synthase, speE, is reversible, catalyzing the breakdown of spermidine and norspermidine, in addition to its usual biosynthetic role. These findings provide support for the hypothesis that enzyme multifunctionality enables streamlined cells in planktonic ecosystems to increase the range of DOM compounds they metabolize. IMPORTANCE Genome streamlining in SAR11 bacterioplankton has resulted in a small repertoire of genes, yet paradoxically, they consume a substantial fraction of primary production in the oceans. Enzyme multifunctionality, referring to enzymes that are adapted to have broader substrate and catalytic range than canonically defined, is hypothesized to be an adaptation that increases the range of organic compounds metabolized by cells in environments where selection favors genome minimization. We provide experimental support for this hypothesis by demonstrating that SAR11 cells take up and metabolize multiple polyamine compounds and propose that a small set of multifunctional enzymes catalyze this metabolism. We report that polyamine uptake rates can exceed metabolic rates, resulting in both high intracellular concentrations of these nitrogen-rich compounds (in comparison to native polyamine levels) and an increase in cell size.

Dynamic carbon flux network of a diverse marine microbial community.

The functioning of microbial ecosystems has important consequences from global climate to human health, but quantitative mechanistic understanding remains elusive. The components of microbial ecosystems can now be observed at high resolution, but interactions still have to be inferred e.g., a time-series may show a bloom of bacteria X followed by virus Y suggesting they interact. Existing inference approaches are mostly empirical, like correlation networks, which are not mechanistically constrained and do not provide quantitative mass fluxes, and thus have limited utility. We developed an inference method, where a mechanistic model with hundreds of species and thousands of parameters is calibrated to time series data. The large scale, nonlinearity and feedbacks pose a challenging optimization problem, which is overcome using a novel procedure that mimics natural speciation or diversification e.g., stepwise increase of bacteria species. The method allows for curation using species-level information from e.g., physiological experiments or genome sequences. The product is a mass-balancing, mechanistically-constrained, quantitative representation of the ecosystem. We apply the method to characterize phytoplankton-heterotrophic bacteria interactions via dissolved organic matter in a marine system. The resulting model predicts quantitative fluxes for each interaction and time point (e.g., 0.16 µmolC/L/d of chrysolaminarin to Polaribacter on April 16, 2009). At the system level, the flux network shows a strong correlation between the abundance of bacteria species and their carbon flux during blooms, with copiotrophs being relatively more important than oligotrophs. However, oligotrophs, like SAR11, are unexpectedly high carbon processors for weeks into blooms, due to their higher biomass. The fraction of exudates (vs. grazing/death products) in the DOM pool decreases during blooms, and they are preferentially consumed by oligotrophs. In addition, functional similarity of phytoplankton i.e., what they produce, decouples their association with heterotrophs. The methodology is applicable to other microbial ecosystems, like human microbiome or wastewater treatment plants.

Episodic Decrease in Temperature Increases mcy Gene Transcription and Cellular Microcystin in Continuous Cultures of Microcystis aeruginosa PCC 7806.

Microcystins produced during harmful cyanobacterial blooms are a public health concern. Although patterns are emerging, the environmental cues that stimulate production of microcystin remain confusing, hindering our ability to predict fluctuations in bloom toxicity. In earlier work, growth at cool temperatures relative to optimum (18°C vs. 26°C) was confirmed to increase microcystin quota in batch cultures of Microcystis aeruginosa NIES-843. Here, we tested this response in M. aeruginosa PCC 7806 using continuous cultures to examine temporal dynamics and using RNA-sequencing to investigate the physiological nature of the response. A temperature reduction from 26 to 19°C increased microcystin quota ∼2-fold, from an average of ∼464 ag µm-3 cell volume to ∼891 ag µm-3 over a 7-9 d period. Reverting the temperature to 26°C returned the cellular microcystin quota to ∼489 ag µm-3. Long periods (31-42 d) at 19°C did not increase or decrease microcystin quota beyond that observed at 7-9 d. Nitrogen concentration had little effect on the overall response. RNA sequencing indicated that the decrease in temperature to 19°C induced a classic cold-stress response in M. aeruginosa PCC 7806, but this operated on a different timescale than the increased microcystin production. Microcystin quota showed a strong 48- to 72-h time-lag correlation to mcy gene expression, but no correlation to concurrent mcy expression. This work confirms an effect of temperature on microcystin quota and extends our understanding of the physiological nature of the response.

Combining Molecular Observations and Microbial Ecosystem Modeling: A Practical Guide.

Advances in technologies for molecular observation are leading to novel types of data, including gene, transcript, protein, and metabolite levels, which are fundamentally different from the types traditionally compared with microbial ecosystem models, such as biomass (e.g., chlorophyll a) and nutrient concentrations. A grand challenge is to use these data to improve predictive models and use models to explain observed patterns. This article presents a framework that aligns observations and models along the dimension of abstraction or biological organization-from raw sequences to ecosystem patterns for observations, and from sequence simulators to ecological theory for models. It then reviews 16 studies that compared model results with molecular observations. Molecular data can and are being combined with microbial ecosystem models, but to keep up with and take advantage of the full scope of observations, models need to become more mechanistically detailed and complex, which is a technical and cultural challenge for the ecological modeling community.

Circadian clock helps cyanobacteria manage energy in coastal and high latitude ocean.

The circadian clock coordinates cellular functions over the diel cycle in many organisms. The molecular mechanisms of the cyanobacterial clock are well characterized, but its ecological role remains a mystery. We present an agent-based model of Synechococcus (harboring a self-sustained, bona fide circadian clock) that explicitly represents genes (e.g., kaiABC), transcripts, proteins, and metabolites. The model is calibrated to data from laboratory experiments with wild type and no-clock mutant strains, and it successfully reproduces the main observed patterns of glycogen metabolism. Comparison of wild type and no-clock mutant strains suggests a main benefit of the clock is due to energy management. For example, it inhibits glycogen synthesis early in the day when it is not needed and energy is better used for making the photosynthesis apparatus. To explore the ecological role of the clock, we integrate the model into a dynamic, three-dimensional global circulation model that includes light variability due to seasonal and diel incident radiation and vertical extinction. Model output is compared with field data, including in situ gene transcript levels. We simulate cyanobaceria with and without a circadian clock, which allows us to quantify the fitness benefit of the clock. Interestingly, the benefit is weakest in the low latitude open ocean, where Prochlorococcus (lacking a self-sustained clock) dominates. However, our attempt to experimentally validate this testable prediction failed. Our study provides insights into the role of the clock and an example for how models can be used to integrate across multiple levels of biological organization.

Fresh Ideas Bloom in Gut Healthcare to Cross-Fertilize Lake Management.

Harmful bacteria may be the most significant threat to human gut and lake ecosystem health, and they are often managed using similar tools, like poisoning with antibiotics or algicides. Out-of-the-box thinking in human microbiome engineering is leading to novel methods, like engineering bacteria to kill pathogens, "persuade" them not to produce toxins, or "mop up" their toxins. The bacterial agent can be given a competitive edge via an exclusive nutrient, and they can be engineered to commit suicide once their work is done. Viruses can kill pathogens with specific DNA sequences or knock out their antibiotic resistance genes using CRISPR technology. Some of these ideas may work for lakes. We critically review novel methods for managing harmful bacteria in the gut from the perspective of managing toxic cyanobacteria in lakes, and discuss practical aspects such as modifying bacteria using genetic engineering or directed evolution, mass culturing and controlling the agents. A key knowledge gap is in the ecology of strains, like toxigenic vs nontoxigenic Microcystis, including allelopathic and Black Queen interactions. Some of the "gut methods" may have future potential for lakes, but there presently is no substitute for established management approaches, including reducing N and P nutrient inputs, and mitigating climate change.

Heterotrophic substrate specificity in the aquatic environment: The role of microscale patchiness investigated using modelling.

Recent observations of natural bacterial communities show high genetic diversity in organic carbon uptake systems (microdiversity) and specificity in substrate species taken up. This seemingly contradicts a large body of literature from laboratory experiments under nutrient-limiting conditions, where mixed substrate use is the rule. This apparent discrepancy can be resolved by realizing that bacteria in the natural aquatic environment encounter nutrients as high-concentration patches. They may live in an ecologically nutrient-limiting environment, but they are rarely in a biologically nutrient-limited state. Rather they switch between non-growing and nutrient-replete states. During nutrient-replete growth the metabolism is saturated and assimilating additional substrates does not increase the growth rate, but carrying the assimilation system constitutes a cost. Consequently, the specialist strategy is beneficial, which is consistent with observations from laboratory experiments. When the bacteria are not growing, the added cost also reduces the fitness of the generalist species. A simple mathematical model encompassing the relevant mechanisms is developed and parameterized realistically based on the literature. The model predicts that, under pulsed conditions, specialization is beneficial when the metabolic cost of an additional uptake system is more than ~0.5% of the total metabolic cost, which is a reasonable estimate and illustrates that this is a plausible hypothesis.

Carbon limitation drives GC content evolution of a marine bacterium in an individual-based genome-scale model.

An important unanswered question in evolutionary genomics is the source of considerable variation of genomic base composition (GC content) even among organisms that share one habitat. Evolution toward GC-poor genomes has been considered a major adaptive pathway in the oligotrophic ocean, but GC-rich bacteria are also prevalent and highly successful in this environment. We quantify the contribution of multiple factors to the change of genomic GC content of Ruegeria pomeroyi DSS-3, a representative and GC-rich member in the globally abundant Roseobacter clade, using an agent-based model. The model simulates 2 × 108 cells, which allows random genetic drift to act in a realistic manner. Each cell has a whole genome subject to base-substitution mutation and recombination, which affect the carbon and nitrogen requirements of DNA and protein pools. Nonsynonymous changes can be functionally deleterious. Together, these factors affect the growth and fitness. Simulations show that experimentally determined mutation bias toward GC is not sufficient to build the GC-rich genome of DSS-3. While nitrogen availability has been repeatedly hypothesized to drive the evolution of GC content in marine bacterioplankton, our model instead predicts that DSS-3 and its ancestors have been evolving in environments primarily limited by carbon.

From Genes to Ecosystems in Microbiology: Modeling Approaches and the Importance of Individuality.

Models are important tools in microbial ecology. They can be used to advance understanding by helping to interpret observations and test hypotheses, and to predict the effects of ecosystem management actions or a different climate. Over the past decades, biological knowledge and ecosystem observations have advanced to the molecular and in particular gene level. However, microbial ecology models have changed less and a current challenge is to make them utilize the knowledge and observations at the genetic level. We review published models that explicitly consider genes and make predictions at the population or ecosystem level. The models can be grouped into three general approaches, i.e., metabolic flux, gene-centric and agent-based. We describe and contrast these approaches by applying them to a hypothetical ecosystem and discuss their strengths and weaknesses. An important distinguishing feature is how variation between individual cells (individuality) is handled. In microbial ecosystems, individual heterogeneity is generated by a number of mechanisms including stochastic interactions of molecules (e.g., gene expression), stochastic and deterministic cell division asymmetry, small-scale environmental heterogeneity, and differential transport in a heterogeneous environment. This heterogeneity can then be amplified and transferred to other cell properties by several mechanisms, including nutrient uptake, metabolism and growth, cell cycle asynchronicity and the effects of age and damage. For example, stochastic gene expression may lead to heterogeneity in nutrient uptake enzyme levels, which in turn results in heterogeneity in intracellular nutrient levels. Individuality can have important ecological consequences, including division of labor, bet hedging, aging and sub-optimality. Understanding the importance of individuality and the mechanism(s) underlying it for the specific microbial system and question investigated is essential for selecting the optimal modeling strategy.

Community Biological Ammonium Demand: A Conceptual Model for Cyanobacteria Blooms in Eutrophic Lakes.

Cyanobacterial harmful algal blooms (CyanoHABs) are enhanced by anthropogenic pressures, including excessive nutrient (nitrogen, N, and phosphorus, P) inputs and a warming climate. Severe eutrophication in aquatic systems is often manifested as non-N2-fixing CyanoHABs (e.g., Microcystis spp.), but the biogeochemical relationship between N inputs/dynamics and CyanoHABs needs definition. Community biological ammonium (NH4+) demand (CBAD) relates N dynamics to total microbial productivity and NH4+ deprivation in aquatic systems. A mechanistic conceptual model was constructed by combining nutrient cycling and CBAD observations from a spectrum of lakes to assess N cycling interactions with CyanoHABs. Model predictions were supported with CBAD data from a Microcystis bloom in Maumee Bay, Lake Erie, during summer 2015. Nitrogen compounds are transformed to reduced, more bioavailable forms (e.g., NH4+ and urea) favored by CyanoHABs. During blooms, algal biomass increases faster than internal NH4+ regeneration rates, causing high CBAD values. High turnover rates from cell death and remineralization of labile organic matter consume oxygen and enhance denitrification. These processes drive eutrophic systems to NH4+ limitation or colimitation under warm, shallow conditions and support the need for dual nutrient (N and P) control.

Neutral Evolution and Dispersal Limitation Produce Biogeographic Patterns in Microcystis aeruginosa Populations of Lake Systems.

Molecular observations reveal substantial biogeographic patterns of cyanobacteria within systems of connected lakes. An important question is the relative role of environmental selection and neutral processes in the biogeography of these systems. Here, we quantify the effect of genetic drift and dispersal limitation by simulating individual cyanobacteria cells using an agent-based model (ABM). In the model, cells grow (divide), die, and migrate between lakes. Each cell has a full genome that is subject to neutral mutation (i.e., the growth rate is independent of the genome). The model is verified by simulating simplified lake systems, for which theoretical solutions are available. Then, it is used to simulate the biogeography of the cyanobacterium Microcystis aeruginosa in a number of real systems, including the Great Lakes, Klamath River, Yahara River, and Chattahoochee River. Model output is analyzed using standard bioinformatics tools (BLAST, MAFFT). The emergent patterns of nucleotide divergence between lakes are dynamic, including gradual increases due to accumulation of mutations and abrupt changes due to population takeovers by migrant cells (coalescence events). The model predicted nucleotide divergence is heterogeneous within systems, and for weakly connected lakes, it can be substantial. For example, Lakes Superior and Michigan are predicted to have an average genomic nucleotide divergence of 8200 bp or 0.14%. The divergence between more strongly connected lakes is much lower. Our results provide a quantitative baseline for future biogeography studies. They show that dispersal limitation can be an important factor in microbe biogeography, which is contrary to the common belief, and could affect how a system responds to environmental change.