Biology is full of complexities, and the more we learn, the more we realize how much remains unknown. A major debate in microbiology is whether DNA alone dictates an organism’s function or if metabolism and energy flows play an equally fundamental role. This question is particularly relevant for microbes in dynamic environments, where survival depends on metabolic adaptability.

This thesis focuses on “Candidatus Accumulibacter”, a key microorganism in wastewater treatment that removes excess phosphorus from water. These bacteria endure feast-famine cycles by storing and utilizing energy reserves as conditions change. While extensively studied, much remains unknown about their metabolic strategies and how environmental factors shape their function. This research combines computational models, laboratory cultivation, and multi-omics analysis to explore how “Ca. Accumulibacter” optimizes its metabolism.

Chapter 1 introduces the central debate: Is DNA the sole blueprint for microbial function, or do metabolism and energy constraints shape microbial behavior? It traces the shift from biochemical models to genome-centric approaches and highlights the potential of a metabolism-first perspective. It also contextualizes “Ca. Accumulibacter” within existing research, outlining its role in biological phosphorus removal and summarizing past findings.

Chapter 2 investigates extracellular polymeric substances (EPS) produced by “Ca. Accumulibacter”, revealing novel glycans and glycoproteins that challenge genome-based predictions. These biomolecules are crucial for biofilm formation and microbial interactions, emphasizing the need for direct biochemical analysis alongside genetic data.

Chapter 3 uses elementary flux mode analysis (EFMA) to map the metabolic potential of “Ca. Accumulibacter”. While genome annotations suggest flexibility, thermodynamic constraints limit feasible metabolic strategies, highlighting the role of energy availability in shaping microbial function.

Chapter 4 introduces the development of the Conditional Flux Balance Analysis (cFBA) Toolbox, an open-source Python framework for modeling metabolism in fluctuating environments. Unlike conventional models that assume steady-state conditions, cFBA enables dynamic predictions of resource allocation over time.

Chapter 5 explores the impact of temperature on “Ca. Accumulibacter” metabolism using cFBA. The findings confirm that biomass synthesis is mainly aerobic but also uncover metabolic shifts at lower temperatures that influence phosphorus removal efficiency and microbial competition.

Chapter 6 examines how “Ca. Accumulibacter” metabolizes multiple substrates simultaneously, revealing unexpected synergies that enhance survival in microbial communities. Combining experimental enrichment cultures with cFBA, this study identifies key metabolic trade-offs and resource optimization strategies.

Finally, Chapter 7 synthesizes the thesis findings, advocating for a shift beyond genome-based interpretations toward a metabolism-centric understanding of microbial function. It discusses broader implications for microbial ecology, wastewater engineering, and metabolic modeling, emphasizing the need for multi-omics approaches and potential applications in synthetic biology.
By integrating experimental and computational approaches, this research deepens our understanding of how “Ca. Accumulibacter” thrives in fluctuating environments. More broadly, it highlights the importance of metabolism and energy flows in shaping microbial function, offering insights that extend beyond wastewater treatment to microbial ecology and engineered bioprocesses. ... Read more

Production of cultured red blood cells (cRBCs) hold the promise of being a potentially unlimited source of cells that could cover the increasing demand of RBCs for transfusion purposes, while having more control on the quality and safety of the cells compared to the current donor-dependent system. cRBCs could also be used for novel therapies in which cells are used as carriers of therapeutic molecules. Scaling up cRBC manufacture is essential to produce the large number of cells needed for such applications. However, scaling up the current static culture systems for the production of erythroblasts (RBC precursor cells) would be prohibitively labor-intensive, requiring large volumes of medium and a high footprint. The work presented in this thesis aims to develop solutions to some of the key challenges in the scaling up of cRBC manufacture.

Stirred tank bioreactors (STRs) are the standard for the large-scale production of biopharma therapeutics, including monoclonal antibodies and vaccines. Agitation in this type of reactors can reduce the concentration gradients of essential nutrients compared to static culture systems such as culture dishes. STRs also offer active control of critical operating parameters in the culture, such as dissolved oxygen concentration, pH and temperature. We therefore developed a culture protocol for the proliferation and differentiation of erythroblasts in STRs (Chapter 2). To define the operating conditions that sustain erythroblast proliferation in STRs, the effect of agitation, aeration strategy, and dissolved oxygen concentration was evaluated using 0.5 L STRs. Using this knowledge, the cultivation process could then be scaled up to 3 L bioreactors.

Erythroblasts lose their replication capacity when transitioning from proliferation to differentiation culture conditions. Thus, efficient proliferation of erythroblasts is essential to produce the large number of cells required for cRBC manufacture. Growing erythroblasts under proliferative conditions is typically performed following a repeated-batch cultivation strategy, in which the culture is diluted every 24 hours with fresh medium to a fixed lower cell concentration. To reduce culture volumes, it is desirable to use higher cell concentrations. However, at increasing cell densities we observe a decrease in growth (Chapter 3). The observed growth limitations of erythroblast cultures at high cell densities appeared to be caused by depletion of low molecular weight nutrients (molecular mass <3 kDa) in the spent medium. We quantified consumption rates of amino acids, major contributors to biomass synthesis in proliferating mammalian cell cultures. Although the concentration of some amino acids decreases considerably over time, supplementation with additional amino acids did not improve growth. Following an untargeted metabolomics approach, we identified multiple pathways that indicate an excess of oxidative stress in erythroblast proliferation cultures.

Perfusion proved to be a successful alternative cultivation strategy to overcome growth limitations due to depletion of nutrient components (Chapter 3). Increasing the maximum cell concentration in erythroblast cultures leads to an increase in the volumetric productivity (number of cells produced per reactor volume per culture time), which decreases the reactor volume needed to produce the same amount of cRBCs. However, large volumes of medium would still be required to sustain those cultures. Currently, the cost of culture medium for erythroid cultures makes cRBC manufacture economically unfeasible. Growth factors and proteins added to the medium are major contributors to the cost of the medium. Holotransferrin, an iron-carrying protein, is the main cost driver in erythroblast differentiation medium. We show that holotransferrin in erythroblast cultures can be replaced by a GMP-compatible iron chelator (deferiprone; Def), bound to ferric ion (Def₃⋅Fe³⁺; Chapter 4) . Addition of Def₃⋅Fe³⁺ to the culture medium resulted in similar final cRBC yields of cRBCs during proliferation and differentiation of erythroblast cultures compared to optimal holotransferrin concentrations. During differentiation, Def₃⋅Fe³⁺ fully supported enucleation and hemoglobinization. We did not observe toxic effects of Def₃⋅Fe³⁺.

Finally, the main conclusions of this thesis are discussed, providing also an overview of the next developments that are required to make the production of cRBCs at large scale technically and economically feasible (Chapter 5). A multidisciplinary approach is needed to further reduce media cost, optimize medium composition to improve cell yields, and to improve the bioreactor culture system developed in this work.
... Read more

Saccharomyces cerevisiae, also known as baker’s yeast, is a robust microorganism frequently used in industrial biotechnology. The scale of its applications ranges from several millilitres for research and process development in the lab to hundreds of cubic meters for cultivation in industrial production processes. In large-scale reactors mixing limitations inherently lead to physiochemical gradients in substrate and oxygen concentrations, pH or temperature. Such inhomogeneous environment in production processes can cause a reduced yield or titer compared to the small-scale development processes. Such scale performance differences can lead to significant worse process economics and increase costs and development time.
The scope of this thesis is to study and understand the regulation of Saccharomyces cerevisiae metabolism under dynamic substrate conditions, using both experimental and modelling approaches. ... Read more

Phosphate accumulating organisms (PAOs) perform a storage polymer metabolism within an anaerobic-aerobic cycle. Anaerobically, PAOs take up volatile fatty acids (VFA) and store them as poly-ẞ-hydroxyalkanoates (PHA). The energy (mainly ATP) necessary for the transport and storage of VFA (and general maintenance) is obtained through the cleavage of polyphosphate. While the reducing equivalents (e.g. NADH) for VFA storage are obtained through the cleavage of glycogen and/or from the anaerobic operation of the TCA cycle. Aerobically, PAOs replenish their reserves of polyphosphate and glycogen, resulting in P uptake, whilst degrading PHA to obtain a carbon and energy supply for growth.
PAOs have the metabolic flexibility to adapt the synthesis of each polymer according to the resources available in the environment, and thus affecting the growth of the organism. Hence, within a PAOs metabolism, there is a trade-off between the use of glycogen and polyphosphate. This trade-off is dependent on the cell's requirements to obtain ATP and NADH for PHA storage. In turn, ATP and NADH amounts can be obtained in different ratios depending on the active metabolic routes. This thesis aims to determine what is the mechanism controlling this trade-off and if there are limits to this relationship.
To investigate this trade-off, a metabolic model for PAOs was created and simulated through a conditional flux balance analysis (cFBA) approach. The resulting amounts of the metabolites simulated with this approach were comparable to those obtained experimentally (figure 6). Additionally, this model was simulated to different sets of starting amounts of glycogen and polyphosphate at a constant acetate feed of 3.84 mCmol/gdw, and the resulting growth was compared between each simulation. This led to an optimal range of initial polyphosphate amounts [2.1-23.5 mPmol/gdw] and initial glycogen amounts [0.3-1 mCmol/gdw]. In reality, these glycogen amounts were never observed experimentally and to the extent of our knowledge never have been reported in PAOs literature. This suggests a glycogen minimal limit amount (e.g. 1 mCmol/gdw), that might reveal a robustness mechanism employed by PAOs to guarantee survivability in uncertain environments. In parallel, a thermodynamic analysis was performed on the malate dehydrogenase reaction, which led to the conclusion that this reaction is not feasible in an anaerobic environment, potentially highlighting a control mechanism. ... Read more