Project Topics
Get Complete Project Material File(s) Now! »
Microbiome functional diversity and redundancy, core microbiome
Clustered organisms display “core” and “flexible” genomes, which are respectively common to all organisms and organism-specific. Core genomes tend to perform central functions, such as DNA replication or protein synthesis while flexible genomes encode generally accessory functions such as transport reactions, and are suspected to control important aspects in the definition, diversification and overlap of species’ ecological niches (Fuhrman, 2009; Curiel Yuste et al., 2014).
Consequently, microbial communities display both a functional diversity and redundancy, which has its importance in how the community is structured. Functional aspects have to be considered when designing core microbiotas or microbial factories, respectively in order to mimic the function of a whole community with a reduced pool of species, or to optimize a biological process (Toju et al., 2018). Redundant functions will be likely to allow broader species choices, while specific ones will likely be more restrictive. In addition, functions taken alone are not enough : since a species might be influenced by another, ecological interactions between microbes must be taken into account, which implies to dive into the study of microbiomes’ assembly rules.
Processes behind communities assembly
A community is defined as a group of organisms representing multiple species living in a specified place and time (Vellend, 2010). Community ecology is the discipline seeking to analyze how biological assemblages are structured (which species and their abundances), what are their functional interactions and how community structure changes in space and time (Konopka, 2009). Understanding how communities assemble has been a central question since the early days of ecology, and the term “assembly rule” was introduced by Diamond (1975), who identified forbidden combinations of species among fruit-eating birds in New Guinea. Assembly rules was then defined as any ecological process selecting for or against a species from the regional species pool, thus defining the local community composition (Götzenberger et al., 2012). In some ways, microbial communities differ from macro-organisms communities, because these processes can act a bit differently (box 4) but in every case however, mechanisms underlying communities patterns are numerous and sorted in four distinct processes (Vellend, 2010):
• Drift, random changes in species abundances (i.e. births, deaths, offspring production).
• Dispersal, the movement of organisms across space, leading to migration.
• Speciation, the creation of new species (relying on mutation)
• Selection, a deterministic process where the most adapted species / individuals can survive, reproduce, and spread whereas the others decline.
Figure 8 displays how these processes articulate together to shape, as an example, the structure of a community of plants. Speciation and dispersal are the two forces which bring new organisms into communities, and drift and selection are the one affecting changes in the presence, absence, and abundance of species (Nemergut et al., 2013). Correlated to spatial distance (and connectivity), selection and drift increase dissimilarity between communities while dispersal decreases it. Speciation (mutations) increases dissimilarity regardless of spatial distance (Hanson et al., 2012).
Communities patterns are various (Vellend, 2010):
• Species-area relationships.
• Abundances and relative abundances.
• Composition-environment relationships.
• latitudinal gradients.
• Distance-decay of similarity.
• Diversity-productivity relationships.
• Diversity-disturbance relationships.
On the first hand, communities are then driven by stochastic factors. Indeed, drift, speciation and dispersal all rely on random (n.b. : by random, we mean observable but not predictable) events, and on the other hand, by deterministic processes gathered under the term “selection”, because all these processes imply a response of an organism/species to environmental abiotic and biotic factors. Abiotic factors can be environmental, like pH, temperature, humidity, nutrients availability (Nemergut et al., 2013). Biotic factors gather a wide set of ecological interactions between species, each with different outcomes (beneficial/neutral/detrimental, box 6) for each species.
The relative importance of deterministic and stochastic events
The relative contribution of stochastic and deterministic events is not clearly known, particularly for microbial communities (Morrison-Whittle and Goddard, 2015). Extreme views exist on the subject, from Hubbell’s theory to authors like Clark (Clark, 2009), for which stochasticity is only an attribute of models, artifacts of unknown or left-aside processes. Mostly, it is admitted that deterministic and stochastic processes combine to generate coexistence, with a primacy of selection, neutral processes being mostly underlying. The contribution of each process can also vary depending on time and spatial scales (Morrison-Whittle and Goddard, 2015; Zhou and Ning, 2017). However, it has to be mentioned that the effect of stochasticity is in general more difficult to measure because of a recurrent bias, which is sampling across environmental gradients or habitat types. This sampling emphasizes the strength of environmental selection, and therefore may artificially minimize the effect of drift (Hanson et al., 2012).
The niche theory, habitat filtering, and the neutral theory
The niche theory assumes that deterministic factors, both abiotic (pH, temperature, salinity…) and biotic (species traits, ecological interactions) shape community structure. It implies that species differ in their niche, the niche being a set of biotic and abiotic conditions defining the volume in which the species can persist (figure 10) (a more precise definition is in box 5). The idea that two species coexisting in the same place must occupy different niches, already present in Darwin’s work, and Hutchinson formalized the concept (Pocheville, 2015). A distinction is made between the
fundamental niche, i.e. the maximal, theoretical volume where a species can survive indefinitely , and the realized niche, i.e. a reduced volume, limited because of interactions with present competitors, where the species actually survives. Various frameworks exist under this theory, such as niche differentiation, which implies microbial communities less phylogenetically clustered than expected by chance, and habitat filtering, which relies on the opposite statement.
Niche differentiation (or niche segregation/partitioning/separation) is the process in which competing species use the environment differently to coexist, despite having similar ecological niches. Indeed, according to the competitive exclusion principle, two species with identical niches cannot coexist because of competition, leading one of them to exclusion. Differentiating their respective niche, such as occupying different spaces or consuming different foods (Caldwell and Vitt, 1999; Brochet et al., 2021), facilitate their coexistence. Other species differentiate in their competitive abilities based on varying environmental conditions : for example, some might be more efficient in dry season while others perform better in rainy seasons, such as plants in the Sonoran Desert (Angert et al., 2009). Another kind of partition can be caused by predators by maintenance of low enough densities of competing species (Grover, 1994). Niche differentiation is then a shift from potential to realized niche, which may cause evolutionary changes afterwards. Niche differentiation has been widely used to explain community patterns both in field and experimental studies, mostly on relatively small scales. Indeed, neutral processes are believed to be of a greater importance at very large scales, but niche differentiation might still be at play (Tang and Zhou, 2011). An examples of field study is by (Kang et al., 2020), where niche differentiation caused by grazing implied various responses from species. For experimental examples, mixtures of several plants led to relative differences of height and leaf surface between species compared to monocultures (Zuppinger-Dingley et al., 2014). Closer to microorganisms, (Burson et al., 2019) experimentally highlighted the coexistence of phytoplankton species thanks to differential use of the underwater light spectrum, which can result from niche partitioning.
Habitat filtering works as the opposite by selecting the ecological traits that confer the best tolerance to a site, leading to a convergence in traits distributions among species (no differentiation). Such process might lead to competition between species, however competition might in return also lead to trait convergence, because only strongly competitive and ecologically equivalent species would remain (Zhang et al., 2017). It is in general hard to predict who between niche differentiation or habitat filtering shapes a community. However, we dispose of hints on spatial and temporal scales. For example, the study of spatial distribution of functional traits in plants highlighted an higher effect of habitat filtering at small spatial scales, during early plant succession (Ulrich et al., 2017). Habitat filtering is likely to determine the niche occupancy, thus community structure of many plants communities worldwide, as demonstrated by (Li et al., 2018). Concerning microbial communities, (Yang et al., 2019) supposed habitat filtering in Chinese grassland in reason of numerous unique OTUs in different habitats, interpreted as habitats specialists.
In opposition to the niche theory, Hubell’s neutral theory of biodiversity is a null hypothesis assuming that all species are ecologically equivalent and have equal rates of birth, death, immigration, emigration. A community’s structure is then independent of species traits and only determined by stochastic processes (drift, dispersal, speciation) (Vellend et al., 2014). Both neutral and niche theories managed to explain community structure, depending on which pattern, time and spatial scales were considered.
Factors and events affecting microbial communities
Various processes can affect a microbial community structure. Among them, wind dispersal and deposit (via aerosols and currents), as well as rain, are known to be relevant vectors affecting microbial communities, with a capacity to connect different microbial ecosystems at a local scale. Microbes are however deferentially dispersed due to their life-strategies and morphological features (Womack et al., 2010; Griggs et al., 2021). As previously mentioned, physio-chemical properties of the biotope are determinant : climate, nutrients availability, pH, humidity, temperature (Zogg et al., 1997; Carrero-Colón et al., 2006; Drenovsky et al., 2010; Meron et al., 2012; Mello et al., 2016; Zhou et al., 2018; Chai et al., 2019; Cui et al., 2021). Landscape features, connectivity to other microbial sources, vectoring by animal hosts, neighboring plants and animals, are also multiple environmental factors shaping microbiota composition (Hacquard, 2016; Griggs et al., 2021). Stochastic events such as the order of arrival (also called assembly history or priority effects) of microbial species in the system may also be at play. For example, manipulation of early immigration history in wood decomposer communities revealed differences in fungal species richness and composition (resulting in different carbon dynamics), most likely associated with different magnitudes in species interactions (Fukami et al., 2010). Echoing these results, Diamond (1975) observed that community composition varied among sites which seemed similar in several ways, thus suggesting that assembly history could lead to multiple stable equilibria. A single equilibrium is more likely to be reached in systems which have a small species pool, a high connectivity, a low productivity and frequent disturbances. Multiple equilibria are more likely to exist in the opposite scheme (Chase, 2003). The same microbe might then have facilities or trouble to establish in a system according to its timing of arrival, which echoes the barrier against pathogens offered by a microbiota to its host, a supplemental filter after host selection.
Selection by the host
In plants, microbes associated with the embryo and endosperm are more likely to be transmitted vertically than those associated with the seed coat. Despite the existence of a large cohort of studies, knowledge of implied mechanisms is scarce (Vandenkoornhuyse et al., 2015). Limits are similar in humans where, despite a lack of evidence in humans, there is support for vertical transfer (from mother to offspring) in animal models, but the colonization mechanisms are unclear (Walker et al., 2017).
Horizontal transmission and pseudo-vertical transmission
In plants, root colonization by soil microbes is in part deterministic and a host-controlled process. Bulk soil constitutes a reservoir of microbes in which variations in root morphology and exudates allow plants to actively recruit their rhizosphere microbiota in the surrounding soil (Bulgarelli et al., 2013; Griggs et al., 2021). The process of recruitment lead to a significant reduction in microbial diversity from soil to rhizosphere and endosphere, and is influenced by plant age, which exert a continuous selective pressure over time (Griggs et al., 2021). Furthermore, the colonization of the internal compartments of the plant could be an attractive goal because of reduced competition to reduced microbial diversity (Hartmann et al., 2009). The effect of host plant is not limited to the roots but extend to other compartments, which exert distinct selective pressure. Indeed, compartments of cultivated agaves were found to have a convergent microbiota’s structure, independent to spatial distance (up to 2000km) (Coleman-Derr et al., 2016). Plants emit from their roots a large variety and quantities of organic exudates (carbohydrates, carboxylic acids, phenolic amino acids) and inorganic ions, according to plant physiological and developmental stage. These exudates have the capacity to condition the rhizospheric environment and therefore the rhizospheric microbial community composition. Furthermore, inhibitory (antimicrobial) or stimulatory (sugars) exudates modulate selective pressure (Hartmann et al., 2009). Because most of the seeds fall and germinate close to the mother plant, the plantlets come in contact with similar microorganisms than the mother. Thus associated to this short term seed dispersion has been hypothesized this pseudo-vertical transmission (Wilkinson and Sherratt, 2001).
The gut microbiota is known to be colonized during early life, although much more research has been conducted on bacteria than on archaea and eukaryotes. The initial microbiome colonization is crucial for the development of individuals. At birth, a subset of the maternal microbiota is supposed to be transferred, here again a pseudo-vertical transmission of microorganisms from the mother to the baby. Within a year (for humans), significant shifts in composition and abundances were highlighted, with strong fluctuations of the microeukaryotic community, as well as a diversification delay caused by cesarean-section and formula milk (Wampach et al., 2017).
Microbe-microbe interactions
In comparison to abiotic factors, much less is known about how microbial interactions shape microbial communities (Nemergut et al., 2013). Determining how species interact is challenging : in situ, the observed behavior of a species is the integration of all its interactions with the other community members. Disentangling the nature of individual interactions is then hard. In addition, when it comes to microbes, particularly host-associated ones, it becomes difficult to directly observe the community in situ (Logan et al., 2018). Ecological interactions between microbes are wide and range from mutualistic exchange of metabolic products to antagonistic secretion of antibiotics and direct predation (Coyte and Rakoff-Nahoum, 2019; Pacheco and Segrè, 2019).
The simplest form of an interaction is pairwised (two participants only) and is classified according to its outcome, which ranges from positive for both (mutualism, noted +/+) to negative for both (competition, -/-). Various outcomes lie in between, with combinations of positive, neutral, or negative outcomes such as ammensalism (-/0) or commensalism (+/0) (Figure 11, box 6, (Zélé et al., 2018). However, microbe-microbe interactions are far from being limited to pairwise-associations. Microbial communities incorporate numerous species of a high taxonomic diversity, characterized by a dense network of high-order (i.e. more than two species) interactions. It is also argued that ecological interactions is somehow incomplete and miss various and crucial nuances, and a enhancement of this framework is needed, including not only ecological outcome, but various attributes which are specificity, cost, contact dependencies, spatial and time dependencies, site, habitat, and compounds involved (Pacheco and Segrè, 2019).
Table of contents :
Chapter 1. General Introduction
Part I – Microbiomes and microbial ecology
1. The world of microorganisms
1.1 What and where are microorganisms ?
1.1.1 The biogeography of microorganisms
1.1.2 Temporal patterns in microbial communities
1.2 Microbiology and microbial ecology
1.3 Microbes are fundamental in ecosystems
1.3.1 Implications in biogeochemical cycles and ecological processes
1.3.2 Microbiotas as a host symbionts
1.3.2.1 The gut microbiome
1.3.2.2 The plant microbiome
1.3.3 Microbiome functional diversity and redundancy, core microbiome
2. Microbiomes’ assembly rules
2.1 Processes behind communities assembly
2.1.1 The relative importance of deterministic and stochastic events
2.2 The niche theory, habitat filtering, and the neutral theory
2.3 Factors and events affecting microbial communities
2.3.1 External or abiotic factors
2.3.2 Selection by the host
2.3.2.1 Vertical transmission
2.3.2.2 Horizontal transmission and pseudo-vertical transmission
2.3.3 Microbe-microbe interactions
2.3.3.1 An overview of ecological interactions
2.3.3.2 Time and spatial patterns affect existence, magnitude, and outcomes of ecological interactions
2.3.3.3 Interactions existence, magnitude and outcome are environment-dependent
3. Data and methods in Microbial ecology
3.1 The -omics revolution
3.2 Metabolic networks and system biology are the basis to study metabolic interactions
Part II. Metabolic cross-feeding with Microbial System Ecology to disentangle coexistence in microbiomes: A mini-review
1 Introduction
2 Metabolic cross-feeding as a major driver of microbiota assemblages
2.1 Evolution and stability of cross-feeding
2.2 The growing importance of metabolic cross-feeding compared to competition
3 Microbial Systems Ecology: a crossroads between system biology and community ecology
3.1 Microbial Systems Ecology approaches and framework
3.1.1 Metabolic network reconstruction
3.1.2 The microbial systems ecology framework calls for shifts between top-down and bottom-up approaches
4 Conclusion
Chapter 2 – Multi-genomes metabolic modelling predicts functional inter-dependencies in the Arabidopsis root microbiome
1 Introduction
2 Materials and Methods
2.1 Genomes data
2.2 Reference database
2.3 Metabolic networks (GEMs) reconstruction
2.4 Metrics from genomes and GEMs
2.5 Targeted Predicted Producible Metabolites (TPPM)
2.6 Nutritional constraint (growth media) modeling
2.7 Putative GEMs combinations for metabolic interactions
2.8 Quasi-Poisson GLMs
2.9 Other statistical analyses
2.10 Scripting
3 Results
3.1 A link between genome-predicted unconstrained metabolism and phylogeny
3.2 SynCom unconstrained PPM are greater than that of single strains and rapidly reaches saturation
3.3 PPM and TPPM number and composition depends on nutritional constraints
3.4 Simplest SynComs are predicted to produce TPPM through metabolic exchanges
4 Discussion
4.1 Fundamental ecological niche signature in GEMs
4.2 Phylogenetic distance, similarity and complementarity, antagonism and cooperation in SynComs
4.3 Metabolism is nutritional-constraint dependent
4.4 Metabolic dependencies are predicted to be major drivers of microbial communities structure.
4.4.1 Metabolic exchanges are nutritional-constraint dependent and compensate severe growth constraints
4.4.2 Minimal combinations of GEMs reflect functional redundancy for targeted compounds
4.4.3 Genomes size effects remains unclear
4.4.4 Few strains are enough to reach the community’s potential
5 Conclusions and prospects
Supplementals
Chapter 3. Cross-feeding and predicted metabolic diversity promote coexistence in A. thaliana root microbiota
1 Introduction
2 Material and methods
2.1 Culture collection and genome-scale metabolic models
2.2 Construction of the predicted produced metabolites gradient
2.3 Bacteria cultures
2.4 Cultures in split-system
2.5 DNA extraction, purification, and sequencing
2.6 Formatting sequencing data
2.7 Data analysis
3 Results
3.1 Reduced competition in SynComs with high PPM
3.2 Cases of cross-feeding identified among many competitive situations
3.3 An Achromobacter sp with an apparent benefit from cross-feeding in three SynComs.
3.4 The constant strain is a slow grower and a weak competitor that could benefit of non-contact secretions
4 Discussion
4.1 The prevalence of antagonistic interactions
4.2 Some strong cross-feeding patterns for Achromobater sp
4.3 Cross-feeding could exist in many cases but might be masked by antagonistic interactions.
4.4 The importance of spatial configuration and metabolites flow
5 Perspectives
Supplementals
Chapter 4. General discussion
1 Metabolic cross-feeding, coexistence, and community assembly : a summary
1.1 The prevalence and importance of cross-feeding
1.2 Cross-feeding selective drivers
1.3 Integration of cross-feeding with competition and environmental factors : from in silico predictions to experimental validation
1.3.1 The difficulty to find simple and appropriate metrics
1.3.2 Cross-feeding as a mostly underlying process compared to competition ?
1.3.3 A few competitive behaviors at the origin of many cross-feeding opportunities ?
1.3.4 Environmental constraints
2 Metabolic cross-feeding is not only about GEMs and various padlocks have to be considered.
2.1 Metabolites availability, secretion, transport, and uptake
2.2 Root exudates composition and their effect on microbiota
2.3 Spatial scale
2.4 Genomes annotation
2.5 Models complexity or simplicity ?
3 Conclusion and perspectives
Supplementary chapter 5 Phylogenetic and functional clustering in a representative bacterial consortia of the Arabidopsis thaliana’s root microbiota
1 Introduction
2 Material and methods
2.1 Data acquisition
2.2 Data filtering
2.3 Co-occurrences network
2.4 Bacterial strains metrics, and scale to the OTU level
2.5 Modules’ phylogenetic distances
2.6 Accumulative null models
3 Results
3.1 The co-occurrences network follows a scale-free topology
3.2 Phylogenetic clustering and functional redundancy in modules
4 Discussion
4.1 The network is a basis to hypothesize interactions, notably between keystone species and the host
4.2 Phylogenetic and functional clustering reveal habitat filtering
5 Conclusion and prospects
General Bibliography
