CHAPTER THREE SMN modelling of mixed microbial cultures: approaches and applications
This chapter presents a full review of applications and approaches used to characterize metabolic activity in environmental bioprocesses using Stoichiometric Metabolic Network (SMN) modelling. The review and discussion of these approaches is necessary as SMN modelling is a technique normally applied to studies of pure species cultures. Further, this review seeks to provide in some detail how SMN modelling can be adapted for metabolic characterization of mixed species microbial cultures (MMC). SMN modelling has successfully been applied to the design and optimization of antibiotic, alcohol and amino acid production processes. It has however been less than successfully applied in the analysis of engineered environmental processes, such as biological nutrient removal or methanogenic anaerobic digestion. Poor applicability occurs due to the fact that the catalytic activity in environmental processes is provided by an actual microbial community, which complicates the model development process considerably. SMN modelling of the environmental bioprocesses required to simulate microbial communities must include multiple species metabolism and their ecological relationships. This review chapter presents and discusses recent approaches to modelling microbial communities using SMN modelling and the applications of those models to Omic data interpretation, process optimization and biochemical pathway design.
Environmental biotechnology aims to develop, use and regulate biological systems for remediation of contaminated environments (land, air, water) and for sustainable and ―clean‖ manufacturing of goods (Kleerebezem and van Loosdrecht, 2007; Vallero, 2010). Moreover, environmental biotechnology aims to consolidate biobased economies. Our current society depends on many natural resources, and the availability of these natural resources (minerals, fossil fuels) is becoming more and more limited. The challenge of a sustainable and biobased economy is to develop innovative technologies to recover and reuse minerals and energy-rich compounds from waste streams and non-food agricultural crops. Environmental biotechnology combines traditional elements from environmental engineering in terms of cleaning of waste streams with process engineering aimed at product manufacturing maximization. Environmental engineering is a broad field, including both abiotic and biotic solutions to pollution and environmental problems. In this context, environmental biotechnology focusses on the development and use the biotic solutions applied by environmental engineering.
Mixed microbial cultures in environmental bioprocesses
The aim of environmental biotechnology is relevant to applications in the fields of wastewater treatment, soil bioremediation, marine bioremediation, bioplastic production, biofuel production and more (Agler et al., 2011; Marshall et al., 2013). These applications require the use of mixed microbial cultures (MMC) to deliver the required good or service (Miller et al., 2010). A mixed microbial culture consists of a microbial community stabilized by selecting the source of the natural inoculum and by controlling the bioprocess conditions so that a natural/ecological selection can be promoted in the microbial population (Kleerebezem and van Loosdrecht, 2007; Rodríguez et al., 2006). This stabilized microbial community has the required metabolic capacities to control the rate of chemical conversions in the mixed microbial culture (Miller et al., 2010). The microbial community is composed of groups of organisms that exploit the same class of environmental resources in a similar way. These groups of similar organisms are called microbial guilds or an ecological functional microbial group (EFMG). A particular mixed microbial culture can be dominated by one or many guilds (Begon et al., 2005). In the case of biological nitrogen removal systems, the microbial community is dominated by the guilds of ammonia and nitrite oxidizers or by denitrifiers. Table 3.1. contains a list of common bioprocesses performed by MMC for environmental applications.
Advantages of mixed microbial cultures
In the context of environmental biotechnology and engineering, bioprocesses based on MMC have four clear advantages over bioprocesses based on traditional pure cultures (Kleerebezem and van Loosdrecht, 2007; Marshall et al., 2013; Rodríguez et al., 2006). They are:
• no sterilization requirements and subsequent reduction of operational cost;
• the capacity to use mixed substrates;
• adaptive capacity owing to microbial diversity; and
• the possibility of a robust process (able to maintain high yields and productivities under on highly variable environmental conditions), e.g. wastewater biotreatment.
MMC are especially attractive for the production of bulk-chemicals, because they reduce the costs associated with culture contamination and work in strictly sterile conditions. For example, the high cost of substrate and equipment required for aseptic operation are the main factors responsible for the high selling price of polyhydroxyalkanoates (PHA), polyesters used for bio-plastic production. The use of open mixed cultures and waste materials as substrate can therefore substantially decrease the cost of PHA and increase their market potential (Dias et al., 2005).
MMC do not require expensive substrates compared to pure culture fermentations, which generally require the use of pure – and therefore more expensive – substrates. In MMC, the use of waste products and less-pure substrates is possible, and has subsequent cost and environmental implications (Rodríguez et al., 2006). The production of energy carriers or other valuable products by mixed culture fermentation finds a use for what were previously considered useless wastes or by-products; and also enables interesting downstream integrations (Rodríguez et al., 2006). Anaerobic digestion is a classic example of a process that combines the objectives of elimination of organic compounds from a waste stream with the generation of a valuable product in the form of methane-containing biogas (Kleerebezem and van Loosdrecht, 2007). Another example is the use of fermented agro-industrial wastes as low-cost substrates for PHA production (Pardelha et al., 2012).
When targeting industrial applications, bioprocess robustness and reproducibility are highly desirable. Bioprocesses based on MMC exhibit these attributes (Allison and Martiny, 2008; Werner et al., 2011). The physicochemical properties of bioreactor feed select the most efficient and effective microbial catalysts and even lead to the evolution of a more stable and productive microbial community (Marshall et al., 2013). Because mixed microbial cultures have a diverse microbial community with multiple metabolic capabilities, they can be resilient to adverse conditions and recover rapidly following an environmental upset. For instance, biological wastewater treatment by activated sludge can operate continuously for years. Similarly, it has been demonstrated that under adequate operational conditions, PHA can be produced by MMC continuously for two years in the same bioreactor, achieving high and stable production rates and yields (Dias et al., 2005).
Limitations of mixed microbial cultures
Despite the above-mentioned advantages MMC did not find wide application at industrial scale – except for waste and wastewater biotreatment applications – as this technology still presents significant limitations. The products formed by MMC vary in amount and composition (Agler et al., 2011). The control of the optimum balance among the microorganisms is not straightforward and requires a better understanding of microbial community behaviour (Agler et al., 2011). Many of the final products of MMC bioprocesses have a low market value, like anaerobic fermentation methane production from waste, which has limitations owing to the low price of natural gas (US$ 0.5 per kg) (Kleerebezem and van Loosdrecht, 2007). In some MMC processes the observed yields are much lower than the ones observed from pure cultures and/or expected from the theoretical process reaction stoichiometry. For example, in bio-hydrogen production from carbohydrates by anaerobic MMC, the measured hydrogen production per mole glucose is much lower (two moles) than the theoretical four mol-H/mol-Glucose yield expected from the bioprocess reaction stoichiometry (Li and Fang, 2007). Another disadvantage is that metabolic routes for waste degradation or product formation can be undefined, therefore complicating the definition of operation strategies. For example, the main limitation of the fermentative hydrogen production process is that no generally accepted selection criterion for the most favourable fermentative hydrogen production route is available (Li and Fang, 2007; Rodríguez et al., 2006). N2O production in wastewater BNR also faces the problem of the many pathways of N2O production due the diversity of microbial guilds and metabolic routes. The outcome of all of these limitations is that designing process operation strategies to avoid the emission of this green-house gas and ozone depleting substance is difficult.
SMN modelling of environmental bioprocesses
Bioprocesses modelling and metabolic modelling
Engineered environmental bioprocesses are complex systems that depend on external chemical and physical processes to achieve the desired goals. Due to complex bioprocess behaviour, environment variability, biological population diversity and operation strategy diversity, it is not always possible to estimate how changing any operational parameter will affect the desired outcome. The problem of complexity can be addressed with mathematical models that enable simulation of a process; and estimating the impact that changing parameters will have on its effectiveness in delivering the service or product (Makinia, 2010). Mathematical modelling of environmental bioprocesses is a common practice in environmental biotechnology and engineering. For instance, the Activated Sludge Models (ASMs) (Kaelin et al., 2009; Makinia, 2010) is a family of bioprocess models widely used by researchers and wastewater treatment facility operators. The main applications of ASM models are listed below (van Loosdrecht et al., 2008):
• to gain insight into process performance;
• to evaluate possible scenarios for upgrading;
• to evaluate new WWT plant design;
• to support management decision-making; and
• to develop new control schemes
The bioprocess model is only a part of full treatment or remediation technology. For example, a model of a full wastewater bio treatment technology based on activated sludge has the following hierarchy of sub models (Makinia, 2010):
Hydraulic process model (describes the hydraulics of each unit operation and their connections)
1. Influent wastewater characterization model
2. Sedimentation model
3. Reactor model (A mass balance equation is applied to each reactor) 3.1. Temperature model
3.2. Oxygen transfer model
3.3 Hydrodynamic mixing model, e.g. CSTR or plug flow. (Mixing and mass transfer characteristics and reactor mass balance equation)
3.4. Bioprocesses model (e.g. ASM)
3.4.1. Kinetic model
188.8.131.52. Metabolic model (e.g. metabolic control analysis (MCA))
3.4.2. Stoichiometric model
184.108.40.206. Metabolic model (e.g. SMN modelling)
The organogram above shows that metabolic models are extensions of bioprocess models developed and used when it is necessary to do more detailed research into microorganism physiology, or when overproduction of a useful or undesired metabolic intermediate is sought (Ishii et al., 2004a). The metabolic modelling approach relies on the concept of metabolic pathways as sequences of specific enzyme-catalysed reaction steps converting substrates into cell products. Despite the existence of accurate bioprocess models, such as ASM, this kind of model does not contain detailed information on cellular behaviour and metabolic pathways. The inclusion of metabolic information is essential for deeper bioprocess understanding and operation improvement. Metabolic models commonly have the following applications (Oehmen et al., 2010):
• as an analytical tool to generate mechanistic hypothesis from experimental observations;
• to improve process efficiency by providing quantitative basis for process design, control and optimisation;
• as a numeric method to estimate the activity of a specific microbial guild; and
• as a tool to investigate the involvement of a specific metabolic pathway in observed process.
Metabolic models have a bright future as both state-of-the-art research tools and for practical applications through linkages with bioprocess models such as ASM (Oehmen et al., 2010). In this way metabolic models can serve as a bridge between molecular/biochemical research and environmental engineering practice, functioning as a tool that can better link the work of microbiologist and engineer in optimising a particular environmental bioprocess (Oehmen et al., 2010).
Stoichiometric metabolic network modelling
SMN modelling is a cutting edge metabolic modelling method used to quantify metabolic reaction rates and in this way describe the metabolic state of cells (Ishii et al., 2004b; Kitano, 2002; Palsson, 2009; Varma and Palsson, 1994). Metabolic network modelling is a data analysis technique of the new branches of biological sciences: Bioinformatics and Systems Biology. These disciplines are involved particularly in computational modelling of cells with a view to investigating and understanding the systematic relationships between genes, molecules and organisms (Endler et al., 2009; Kell, 2006; Kitano, 2002; Park et al., 2008).
SMN models have become an important tool for characterizing the metabolic activity of cells in biotechnological process and have huge potential to assist in the analysis and understanding of MMC (Lovley, 2003; Zengler and Palsson, 2012). The explosion in the number of new SMN models for up to 200 different organisms over the last few years highlights the increasing popularity of this approach in the pharmaceutical, chemical, and environmental industries (Kim et al., 2012; Park et al., 2008). This usefulness relies on their application as a computational tool to address questions that cannot be easily addressed experimentally.
Approaches to modelling MMC with SMN modelling
While several aspects of microbial metabolism can be fruitfully addressed by studying pure cultures of individual microbial species, many environmental bioprocesses require an understanding of how microbes interact with each other (Klitgord and Segrè, 2010; Lovley, 2003; Zengler and Palsson, 2012). Lack of information about environmental factors controlling the growth and metabolism of microorganisms in polluted environments often limits the implementation of biodegradation strategies (Lovley, 2003). Only a detailed understanding of the functioning and interactions within microbial populations will allow a rational manipulation for the purpose of optimizing bioremediation efforts (Vilchez-Vargas et al., 2010). Within this context, SMN modelling may be especially relevant to the analysis of environmental and industrial bioprocesses based on MMC (Miller et al., 2010).
Environmental bioprocesses have the peculiarity of being open systems where the catalytic activity is provided by microbial communities instead of single species populations (Grady et al, 1999; Comeau, 2008). Interaction between organisms may be especially important in microbial communities where multiple species are involved in degrading substrates available on the environment (Stolyar et al., 2007). The development of more sophisticated metabolic network modelling methods for interacting species will enable increasingly realistic prediction of communal phenotypes (Stolyar et al, 2007; Oberhardt et al, 2009). These developments will allow SMN models that include metabolic information from different species to quantify rates of exchange of compounds between different species populations (Lovley, 2003; Stolyar et al, 2007). The model can be applied to simulate cellular metabolism of a homogenous mixture of bacterial cells in suspension (completely mixed system in stirred tank reactor). However, it can be extended to simulate cellular metabolism in a biofilm or flocculated system by implementing reaction–diffusion equations (Rodríguez et al., 2006).
The stoichiometric metabolic modelling approach has been used since 1999 to understand the behaviour of biological systems in complex environments and to model organisms relevant to engineering environmental bioprocesses, when Pramanik and co-workers (Pramanik et al., 1999) developed a SMN model of phosphate accumulating organisms. This study was the first attempt to adapt SMN to model a MMC. Professor Derek R. Lovley from University of Massachusetts was one of the first researchers to present a coherent framework to use omics techniques, computational biology and metabolic network modelling to study engineered environmental processes (Lovley, 2003). However, as shown in Table 2.2., literature reviewed up until 2014 indicates that SMN modelling has been applied to quantify metabolic rates in engineered environmental bioprocesses in only a few studies, including modelling studies by (Poughon et al., 2001) on nitrification.
Table 2.2. also shows that four approaches have been developed to model microbial communities in mixed cultures using SMNs. These SMN modelling approaches are described above and in the following sections:
• Lumped network
• Compartment per guild network (also known as multi-compartment)
• Dynamic-SMN (also known as hybrid) and
• Bi-level simulation
In this approach the community is modelled as a single entity in which all metabolic reactions and metabolites from the guilds are combined into a single set of reactions. A metabolic network of the whole mixed culture is built up by simply inventorying the most common catabolic reactions, i.e. electron transport chain, glycolysis, tricarboxylic acid cycle (TCA) and amino acid synthesis; and later lumping multiple subsequent reactions of specific pathways into a single reaction that represents the overall pathway (Rodríguez et al., 2006). Reactions catalysed by more than one guild are only considered once. The method captures the metabolic constraints of the overall matter and energy transformations without the need for detailed knowledge of every organism in the community (Taffs et al., 2009).
This modelling approach is based on the assumption that all the organisms in the community have reactions in common. It treats the MMC or microbial community as a single virtual microorganism catalysing the most common pathways (Rodríguez et al., 2006). The product spectrum is obtained by maximizing the biomass growth yield which is limited by catabolic energy production. The virtual microorganism proposed here should be regarded as a representation of the different microbial strains involved in the bioprocess. Microbial diversity and the dynamics of the process are neglected at this stage (Rodríguez et al., 2006). Ignoring microbial diversity and assuming a virtual microorganism able to carry out the most common fermentative conversions is acceptable in steady state conditions (Rodríguez et al., 2006). thus, simplifying the processes of model development and calibration.
The lumped network approach is ideally suited for investigating the metabolic potential of a community based solely on metagenomic data as the assignment of each reaction to a constituent guild is unnecessary. The approach is quite flexible and can be scaled to different levels of detail. An additional advantage of the lumped approach is the reduction of computational burden. With these advantages, the method is uniquely suited for initial and exploratory analyses of diverse or poorly understood communities (Taffs et al., 2009). A weakness in the lump reactions is that the model‘s output does not specify which guilds employ a particular enzyme or produce biomass and maintenance ATP. Instead, the results describe potential performance of the microbial community or MMC. The method also neglects the logistics associated with transferring metabolites between organisms, including conversion of the relevant metabolite into one for which transporters are available (Taffs et al., 2009)
Compartment per guild approach (multi-compartment)
In a compartment per guild network, each organism or guild is modelled as a distinct compartment and exchangeable metabolites are transferred through an extra compartment representing the extracellular environment (Klitgord and Segrè, 2010; Stolyar et al., 2007; Taffs et al., 2009). The approach is implemented by assigning reactions and metabolites to a network representing each guild, with suffixes on metabolite identifiers preventing sharing of compounds common to the metabolism of multiple guilds. Explicit transport reactions accounting for the exchange of metabolites between guild members and the extracellular space are defined (Taffs et al., 2009). This approach introduces a fictitious compartment that represents the extracellular environment shared by the microbial species in addition to the original extracellular spaces for individual models (Klitgord and Segrè, 2010). The compartments represented by different microbial species are separated spatially by the extracellular medium. Consequently, the presence or absence of a single transporter in one species may greatly affect the behaviour of other species in the system (Stolyar et al., 2007).
The compartment per guild modelling approach has the advantage of conceptual tractability. Dividing the community into guild-level compartments linked by transferred metabolites, e.g. oxygen, is an intuitive way to represent interactions within a community. It is also an ideal method for understanding which guild performs a particular metabolic transformation. For example, it is easy to estimate the fraction of total biomass (carbon moles) or total maintenance ATP (used to account for energy-dependent cellular processes other than growth) produced by each guild (Taffs et al., 2009). Using a multi-compartment approach, (Klitgord and Segrè, 2010) developed the Search for Exchanged Metabolites (SEM) algorithm to verify potential interactions between a pair of organisms and produce a list of putatively exchanged metabolites. Selected carbon and nitrogen sources are then combined in all possible ways to give rise to a set of putative media that can sustain growth of the join pair model (Klitgord and Segrè, 2010). In this way, The Search for Interaction-Inducing Media (SIM) algorithm identifies the set of media that support growth of multi-species co-cultures and predicts the class of interaction they induce (Klitgord and Segrè, 2010)
One drawback of this approach is that the size of the resulting network can lead to a ‗combinatorial explosion‘ of new pathways composed by reactions from different guilds (Klamt and Stelling, 2002). To address this limitation, the models for each guild member can be constructed to only capture the necessary metabolic capabilities while maintaining computational tractability (Taffs et al., 2009). A second drawback of this approach is the requirement for significant a priori information or assumptions, as reactions must be assigned to each individual guild (Stolyar et al., 2007).
This approach couples the rate predictions of SMN models with differential equations that capture the dynamic response of the biological process with respect to substrate concentrations, temperatures or pH. Differential equations have been coupled to single species SMN (Çalik et al., 2011; Hjersted et al., 2005; Mahadevan et al., 2002) or to various SMNs to yield a multispecies model (Scheibe et al., 2009; Zhuang et al., 2011). The main attribute of hybrid models is that they can predict reaction rates and compound (metabolite) concentrations across a time interval.
Dynamic-SMN captures both metabolic complexity and metabolic dynamism. Because the majority of environmental bioprocesses are in fact dynamic systems, this approach has the potential to truly capture the behaviour of these systems. The constraint-based modelling approach may be particularly well suited to modelling microorganisms in heterogeneous environments, as it does not assume constant yield coefficients and has been shown to account for the changes in the metabolic network in response to nutrient limitations (Schuetz et al., 2007).
Models that can accurately predict microbial growth and activity are particularly important when dealing with the dynamic conditions expected in wastewater biotreatment and soil bioremediation. For instance, in (Scheibe et al., 2009), a genome-scale SMN model of the metabolism of Geobacter sulfurreducens was coupled to a soil reactive transport model (HYDROGEOCHEM). The objective was to model operations of in situ bioremediation of uranium spills in soil.. By discretizing an aquifer as a numeric grid, the hybrid model simulates hydrologic, geochemical and metabolic processes in the spill are. At each time step, the SMN model estimates microbial mediated metabolic reaction rates of the simulated geochemical conditions in each grid element at that time. These fluxes then feed back to the reactive transport model as reaction rates for the current time step. The reactive transport model is stepped forward one time interval, after which the process is repeated. The step that involves referencing the metabolic model from the reactive transport model can take one of two forms:
(i) concurrent execution of the metabolic model through a direct subroutine call from the reactive transport model; or (ii) selection of metabolic fluxes at each time step of the reactive transport simulation from a large set of pre-calculated metabolic model solutions covering the expected range of environmental conditions. In this study, researchers chose an example of in situ subsurface bioremediation application that is essentially catalysed by one microbial genus in an initial attempt to apply SMN modelling to bioremediation. The coupled SMN and reactive transport model predicted acetate concentrations and U(VI) reduction rates in a field trial of in situ uranium bioremediation that were comparable to the predictions of a calibrated conventional model, but without the need for empirical calibration other than specifying the initial biomass of Geobacter. The results from this study suggest that coupling SMN models with reactive transport models may be a worthwhile approach to developing models that can be truly predictive (Scheibe et al., 2009).
Table of contents
Table of Contents
List of Figures
List of Tables
Notation and Symbols
Chapter One. Introduction and research scope
1.1. Relevance of nitrous oxide (N2O) in our environment
1.2. Microbes as the major source of N2O
1.3. N2O emission from biological nitrogen removal processes
1.4. Linking N2O production metabolism with BNR operational parameters through SMN modelling and metabolomics
1.5. Why bring Systems Biology tools to Environmental Engineering?
1.6. Research aim and objectives
1.7. Thesis outline
Chapter Two. Development of SMN models of biological nitrogen removal microbes
2.1. Research approach overview
2.2. Used software
2.3. Concept of SMN modelling
2.4. Stoichiometric metabolic network reconstruction
2.5. Conversion of reconstructed SMN into a mathematical model
2.6. Model simulation
2.7 Simulation software
2.8. Model calibration
Chapter Three. SMN modelling of mixed microbial cultures: approaches and applications
3.1. Environmental biotechnology
3.2. Mixed microbial cultures in environmental bioprocess
3.3. SMN modelling of mixed microbial cultures
3.4. Approaches to model MMC with SMN
3.5. Applications of SMN modelling of MMC
Chapter Four. Effect of aerobic and anoxic conditions on N2O and NO production pathways in ammonia oxidizing bacteria
4.2. Materials and Methods
Chapter Five. N2O production in nitrifying mixed cultures: Effect of NO turnovers, oxygen and ammonium concentrations and microbial community structure
5.2. Materials and Methods
Chapter Six. N2O accumulation in denitrifying mixed cultures: Influence of medium COD/N ratio and high nitrite accumulation
6.2. Materials and Methods
6.3. Results and Discussion
Chapter Seven. Discussion and conclusions
7.1. Linking BNR operation parameters to microbial N2O production
7.2. Importance of gene regulation and enzyme kinetics
7.3. SMN models and metabolomics of microbial communities
7.4. Future research
GET THE COMPLETE PROJECT