From hairballs to hypotheses–biological insights from microbial networks

FEMS Microbiology Reviews, Год журнала: 2018, Номер 42(6), С. 761 - 780

Опубликована: Июль 25, 2018

Microbial networks are an increasingly popular tool to investigate microbial community structure, as they integrate multiple types of information and may represent systems-level behaviour. Interpreting these is not straightforward, the biological implications network properties unclear. Analysis allows researchers predict hub species interactions. Additionally, such analyses can help identify alternative states niches. Here, we review factors that result in spurious predictions address emergent be meaningful context microbiome. We also give overview studies analyse new hypotheses. Moreover, show a simulation how affected by choice environmental factors. For example, consistent across tools, heterogeneity induces modularity. highlight need for robust inference suggest strategies infer more reliably.

Язык: Английский

---

Strength of species interactions determines biodiversity and stability in microbial communities

Nature Ecology & Evolution, Год журнала: 2020, Номер 4(3), С. 376 - 383

Опубликована: Фев. 10, 2020

Язык: Английский

---

Deciphering microbial interactions in synthetic human gut microbiome communities

Molecular Systems Biology, Год журнала: 2018, Номер 14(6)

Опубликована: Июнь 1, 2018

Article21 June 2018Open Access Transparent process Deciphering microbial interactions in synthetic human gut microbiome communities Ophelia S Venturelli Corresponding Author [email protected] orcid.org/0000-0003-2200-1963 Department of Biochemistry, University Wisconsin-Madison, Madison, WI, USA Search for more papers by this author Alex V Carr Environmental Genomics and Systems Biology, Lawrence Berkeley National Laboratory, Berkeley, CA, Garth Fisher Ryan H Hsu orcid.org/0000-0001-8221-224X California Institute Quantitative Biosciences, Rebecca Lau Benjamin P Bowen Susan Hromada Trent Northen Adam Arkin Bioengineering, Energy Biosciences Institute, Information *,1, Carr2,‡,‡, Fisher2,‡, Hsu3, Lau2, Bowen2, Hromada1, Northen2 Arkin2,3,4,5 1Department 2Environmental 3California 4Department 5Energy ‡These authors contributed equally to work ‡Correction added on 27 2018 after first online publication: C was corrected *Corresponding author. Tel: +1 608 263 7017; E-mail: Molecular Biology (2018)14:e8157https://doi.org/10.15252/msb.20178157 See also: Abreu et al (June 2018) PDFDownload PDF article text main figures. Peer ReviewDownload a summary the editorial decision including letters, reviewer comments responses feedback. ToolsAdd favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Abstract The ecological forces that govern assembly stability microbiota remain unresolved. We developed generalizable model-guided framework predict higher-dimensional consortia from time-resolved measurements lower-order assemblages. This method employed decipher diverse community. show pairwise are major drivers multi-species community dynamics, as opposed higher-order interactions. inferred network exhibits high proportion negative frequent positive Ecological responsive recipient species were discovered network. Our model demonstrated prevalent interaction topology enables robust coexistence implementing feedback loop balances disparities monospecies fitness levels. could generate history-dependent initial proportions frequently do not originate bistability. Measurements extracellular metabolites illuminated metabolic capabilities potential molecular basis In sum, these methods defined roles human-associated intestinal design principles communities. Synopsis Analysis shows dynamics. study reveals drivers, metabolite hub ecologically sensitive organisms A data-driven pipeline is used construct predictive dynamic anaerobic Design stable history-dependence elucidated. profiles analyzed each organism. highlights challenges using phylogenetic exo-metabolomic "signals" functions. Introduction Microbes have evolved occupy nearly every environment Earth, spanning extreme environments such acid mine drains hot springs multicellular organisms. dense collection microorganisms inhabits gastrointestinal tract (Lozupone al, 2012; Earle 2015; Tropini 2017) performs numerous functions impact physiology, nutrition, behavior, development (Ley 2005; Fischbach Sonnenburg, 2011; Foster McVey Neufeld, 2013; Louis 2014; Sharon Rooks Garrett, 2016). Functions partitioned among genetically distinct populations interact perform complex chemical transformations exhibit emergent properties colonization resistance at level. Such collective realized combined operating multiple time spatial scales be achieved single population. degree structuring varies across length scales: At macroscale hundreds micrometers, bacteria cluster into habitats, whereas scale intermixing members has been observed (Donaldson Mark Welch 2017). composed bacterial species, majority which span Firmicutes, Bacteroidetes, Actinobacteria phyla 2006). Constituent strains shown persist an individual over long periods time, demonstrating (Faith 2013). Perturbations system dietary shifts or antibiotic administration can shift point alternative state (Relman, 2012). While identities co-occurrence relationships individuals elucidated (Faust 2012), we lack quantitative understanding how shape assembly, stability, response perturbations. For example, enable dominant Bacteroidetes well understood (Fischbach 2011). resilience microbiomes, capacity recover perturbations, strongly linked diversity. Indeed, reduction diversity associated with diseases, suggesting high-dimensional functionally heterogeneous ecosystem promotes health (Sommer Understanding factors influencing implications targeted interventions modulate states. Central problem inferring unknown developing tools temporal changes behaviors environmental stimuli. Cooperation competition feedbacks influence functional activities stability. Negative dominate inter-relationships aquatic microcosms (Foster Bell, However, prevalence cooperation occupying other remains elusive. Direct resources space, biomolecular warfare, production toxic waste products (Hibbing 2010). Positive stem secreted utilized member detoxification environment. Pairwise modified third organism, leading effects (Bairey driver large structure function, represent key nodes manipulated control states (Gibson Predicting dynamics step toward organizational Computational models different resolutions analyze Raes, Dynamic computational investigate structure, dynamical systems theory parameter sensitivity (Astrom Murray, Generalized Lotka–Volterra (gLV) ordinary differential equation represents limited number parameters deduced time-series data. Here, develop systematic modeling experimental interrogate mediating assembly. Time-resolved assemblages train revealed Specific context identified exhibited overall community, specific Firmicutes displayed outgoing sub-network variations parameters. showed due slow convergence steady composition networks enriched exo-metabolomics profiling, data pinpointed set predicted mediate failed forecast influential modulating Together, results combinations couplings realize repertoire behaviors. Results Probing aimed dissect reduced complexity Actinobacteria, Proteobacteria. To end, ecology encompassing Bacteroides thetaiotaomicron (BT), ovatus (BO), uniformis (BU), vulgatus (BV), Blautia hydrogenotrophica (BH), Collinsella aerofaciens (CA), Clostridium hiranonis (CH), Desulfovibrio piger (DP), Eggerthella lenta (EL), Eubacterium rectale (ER), Faecalibacterium prausnitzii (FP), Prevotella copri (PC) designed mirror natural (Fig 1A; Qin These contribute significantly implicated diseases (Watterlot 2008; Larsen 2010; Thota Fujimoto Haiser Scher Table 1). Figure 1. Experimental high-throughput characterization Phylogenetic tree 12-member analysis performed concatenated alignment single-copy marker genes obtained via PhyloSift (preprint: Darling 2014). Maximum likelihood trees generated default options. bar substitutions per site alignment. Schematic study. Species approximately 1:1 19:1 based absorbance 600 nm (OD600) microtiter plates liquid-handling robotic manipulation. Approximately 12 h, samples collected multiplexed 16S rRNA gene sequencing (black circles). Relative abundance measured V3–V4 region dual-indexed primers compatible Illumina platform (stacked plot, bottom right). Serial transfers 24 intervals (purple bars, top) transferring aliquot fresh media 1:20 dilution. parallel, OD600 performed. Download figure PowerPoint associations previous literature. Arrows pointing up down denote associations, respectively Association(s) Inflammatory autoimmune disease (↑) (Scher 2013), autism (↓) (Kang 2013) (BV) Ulcerative colitis (Bamba 1995) (BU) Metabolic/immunological dysfunction (Gauffin Cano 2012) (BO) Type I diabetes (Giongo 2011) (BT) (↑)(Bloom (FP) Crohn's 2008), inflammatory bowel (Segain 2000), Celiac (De Palma 2010) (BH) Healthy colon (Nava (ER) II (Larsen (CA) Colon cancer (Moore Moore, 1995), rheumatoid arthritis (Chen 2016) (EL) Cardiac drug (Haiser (Thota 2011), (DP) Regressive (Finegold (CH) None reported Synthetic arrayed chamber automated procedure (see Materials Methods). rich Methods) selected support growth all monospecies. serially transferred 24-h prevent lag phases being eliminated allow approach monitoring many cell generations. Further, serial also reflect recurrent perturbations diet colonic transit 1B). Multiplexed 12-h elucidate stages. relative computed sum read counts organism divided total reads condition Since construction aided absolute information (Bucci 2016; Widder 2016), biomass monitored 30 min (OD600). Cellular traits adhesion, size, (Stevenson addition, counting colony-forming units (CFU) biased dormant sub-populations, selection solid vs. liquid media, stage (Jansson Prosser, 1997; Volkmer Heinemann, Ou trained estimated thus automatically accounts any biases. infer interactions, (66 combinations) ratio values (PW1 dataset, Appendix Fig S1, Dataset EV1). Monospecies sequencing, broad range rates, carrying capacities (M S2). single-species dominance (Appendix S1). distribution PW1 provided insight variability presence second S3A). Absolute normalized maximum value evaluate baseline species. CH lowest coefficient variation (CV), indicating levels S3B). remaining bimodal (FP, DP, PC, BH, CA), long-tail (ER EL), and/or CV (ER, PC), altered further probe consortia, 15 2B, EV2) inoculated (95% A, 5% B, wherein percentages reversed) characterized our workflow (PW2 S4). classified following categories threshold 72 h: (i) dominance; (ii) both persisted above threshold; (iii) history dependence whereby mapped structures; (iv) did quantitatively satisfy thresholds cases 1–3. subset category weak potentially attributed biological replicates. qualitative 51 (t = 0) final h) PW2 reciprocal percentages, S5A). 50, 24, 12% dominance, coexistence, dependence, S5B). 2. Model training generalized function fits T3 represented points lines, respectively. subplot, x- y-axis, Stars datasets mean squared errors greater than 0.15. Error bars 1 s.d. least three Temporal 95% B values. Time Data lines T3, Inferred inter-species coefficients gLV T3. Gray green edges (αij < > coefficients. edge width node size magnitude (xe −μiαii−1), highlight significant less 1e-5 displayed. Construction growth, intra-species sensitivity. equations given by: where n, μ, αii, αij intra-species, coefficients, minimize overfitting data, regularized estimation implemented penalized Three sets evaluated capability: (T1) M; (T2) M, PW1; (T3) PW1, PW2. regularization (λ) scanned balance goodness fit sparsity S6). parameterized captured 2A B). accurately EL; CA; BO, CH; ER, BH; BH error between Thresholding yielded densely connected 77% pairs interaction. connectivity varied 75 79% ranging 1e-6 1e-3. Interaction 1e-3 expected change rate Of 56 21% positive, 2C). arise resource competition, by-products. secretion (BO, BV, BU, BT) net network, EL, positively stimulated S7A). unidirectional (−/0, 36%), bidirectional (−/−, 32%), (+/−, 26%) S7B). contribution full dictated coupled incoming Therefore, prediction about role requires simulation model. FP five determine abundance, examined 6-member FP, CH, DP. required alter twofold, dual moderately increased h S7C). enhanced conditions. Corroborating notion, Figs S3 Strong deciphered enhancement productivity compared null representing productivities integral S8). 38% twofold predictions models, consistent BU; BO; BT, BV; BO lower model, mutual inhibitory topologies BT; ER comparison (EL, ER) mutualism, augment productivity. Hierarchical clustering similar patterns S9A pattern relationships. distantly related display (e.g., DP PC CA) closely exhi

Язык: Английский

---

Toxicity drives facilitation between 4 bacterial species

Proceedings of the National Academy of Sciences, Год журнала: 2019, Номер 116(32), С. 15979 - 15984

Опубликована: Июль 3, 2019

Competition between microbes is extremely common, with many investing in mechanisms to harm other strains and species. Yet positive interactions species have also been documented. What makes help or each currently unclear. Here, we studied the 4 bacterial capable of degrading metal working fluids (MWF), an industrial coolant lubricant, which contains growth substrates as well toxic biocides. We were surprised find only neutral Using mathematical modeling further experiments, show that this community likely due toxicity MWF, whereby species' detoxification benefited others by facilitating their survival, such they could grow degrade MWF better when together. The addition nutrients, reduction toxicity, more instead resulted competitive behavior. Our work provides support stress gradient hypothesis showing how harsh, environments can strongly favor facilitation microbial mask underlying interactions.

Язык: Английский

---

Trophic Interactions and the Drivers of Microbial Community Assembly

Current Biology, Год журнала: 2020, Номер 30(19), С. R1176 - R1188

Опубликована: Окт. 1, 2020

Despite numerous surveys of gene and species content in heterotrophic microbial communities, such as those found animal guts, oceans, or soils, it is still unclear whether there are generalizable biological ecological processes that control their dynamics function. Here, we review experimental theoretical advances to argue networks trophic interactions, which the metabolic excretions one primary resource for another, constitute central drivers community assembly. Trophic interactions emerge from deconstruction complex forms organic matter into a wealth smaller intermediates, some released environment serve nutritional buffet community. The structure emergent network rate at resources supplied many features assembly, including relative contributions competition cooperation emergence alternative states. Viewing assembly through lens also has important implications spatial communities well functional redundancy taxonomic groups. Given ubiquity across environments, they impart common logic can enable development more quantitative predictive ecology.

Язык: Английский

---

Use and abuse of correlation analyses in microbial ecology

The ISME Journal, Год журнала: 2019, Номер 13(11), С. 2647 - 2655

Опубликована: Июнь 28, 2019

Язык: Английский

---

Massively parallel screening of synthetic microbial communities

Proceedings of the National Academy of Sciences, Год журнала: 2019, Номер 116(26), С. 12804 - 12809

Опубликована: Июнь 11, 2019

Microbial communities have numerous potential applications in biotechnology, agriculture, and medicine. Nevertheless, the limited accuracy with which we can predict interspecies interactions environmental dependencies hinders efforts to rationally engineer beneficial consortia. Empirical screening is a complementary approach wherein synthetic are combinatorially constructed assayed high throughput. However, assembling many combinations of microbes logistically complex difficult achieve on timescale commensurate microbial growth. Here, introduce kChip, droplets-based platform that performs rapid, massively parallel, bottom-up construction communities. We first show kChip enables phenotypic characterization across conditions. Next, screen ∼100,000 multispecies comprising up 19 soil isolates, identified sets promote growth model plant symbiont

Язык: Английский

---

Interaction networks, ecological stability, and collective antibiotic tolerance in polymicrobial infections

Proceedings of the National Academy of Sciences, Год журнала: 2017, Номер 114(40), С. 10666 - 10671

Опубликована: Сен. 18, 2017

Significance In many infections, multiple microbial species are present simultaneously. Such polymicrobial infections can be viewed as small ecosystems. Do bacteria in these communities interact with each other? If so, do interactions affect the stability of ecosystem, particular, when antibiotics present? We focus on urinary tract and demonstrate that there ample ecological between different bacterial species, both presence absence antibiotics. further show they crucially ecosystem resilience to environmental perturbations such Understanding nature point toward ways disrupting infection ecosystems, which could potentially used a new strategy eradicate infective communities.

Язык: Английский

---

High-order interactions distort the functional landscape of microbial consortia

PLoS Biology, Год журнала: 2019, Номер 17(12), С. e3000550 - e3000550

Опубликована: Дек. 12, 2019

Understanding the link between community composition and function is a major challenge in microbial population biology, with implications for management of natural microbiomes design synthetic consortia. Specifically, it poorly understood whether functions can be quantitatively predicted from traits species monoculture. Inspired by study complex genetic interactions, we have examined how amylolytic rate combinatorial assemblages six starch-degrading soil bacteria depend on separate functional contributions each their interactions. Filtering our results through theory biochemical kinetics, show that this simple additive absence interactions among members. For about half combinatorially assembled consortia, dominated pairwise higher-order other half, despite presence strong competitive We explain mechanistic basis these findings propose quantitative framework allows us to effect behavioral dynamics Our suggest robustness consortium critically affects ability predict bottom-up engineer ecosystem communities.

Язык: Английский

---

Microbial communities as dynamical systems

Current Opinion in Microbiology, Год журнала: 2018, Номер 44, С. 41 - 49

Опубликована: Июль 21, 2018

Nowadays, microbial communities are frequently monitored over long periods of time and the interactions between their members explored in vitro. This development has opened way to apply mathematical models characterize community structure dynamics, predict responses perturbations explore general dynamical properties such as stability, alternative stable states periodicity. Here, we highlight role systems theory exploration communities, with a special emphasis on generalized Lotka–Volterra (gLV) equations. In particular, discuss applications, assumptions limitations gLV model, mention modifications address these review stochastic extensions. The models, together generation series data, can improve design control communities.

Язык: Английский

---