PHA Production in Aerobic Mixed Microbial Cultures

Abstract

Polyhydroxyalkanoate (PHA) is a common intracellular energy and carbon storage material in bacteria, which is considered as a bioplastic due to its plastic like properties. PHAs are versatile materials which are biodegradable and made from renewable resources. Commercial production of PHAs is currently based on pure culture processes employing either natural PHA producers or genetically modified bacteria. Pure culture processes use generally pure sterile substrates and axenic reactors, leading to high production costs and thus relatively expensive products. An alternative approach for the production of PHAs is the use of mixed culture biotechnology, using non-sterile waste streams as a substrate and open reactors. The use of cheaper substrates, less energy (no sterilization of substrate or reactors) and cheaper equipment could reduce the production costs compared to pure culture processes. However, the mixed culture PHA production process requires optimization for higher cellular PHA contents to be competitive with pure culture processes. The research described in this thesis aimed at improving the cellular PHA contents that can be achieved in open mixed cultures. A two-step process consisting of (i) a culture enrichment and growth step and (ii) a PHA production step was used. For the enrichment of a mixed culture with PHA producing bacteria a selective pressure in the form of alternating periods of short presence of the carbon substrate (feast phase) and long absence of the carbon substrate (famine phase) under fully aerobic conditions was employed. PHA storing bacteria generally outcompete other bacteria in such a feast-famine system due to their very high substrate uptake rate (which is not limited by the growth rate) and due to the ability to grow in a more balanced way throughout feast and famine phase. A sequencing batch reactor (SBR) was used to establish the feast-famine regime. The cultures enriched in the first step under different operational conditions were tested for their ability to produce PHA in the second step, the PHA production step. For this purpose the cultures were supplied with an excess of carbon source (fed-batch reactor) while withholding a suitable nitrogen source in order to avoid growth and direct as much carbon as possible into PHA storage. To simplify the system for the optimization studies a mineral medium with acetate as the sole carbon substrate was used in all experiments rather than real wastewater. Acetate yielded pure polyhydroxybutyrate (PHB) as the storage polymer. In order to compare different operational conditions, specific reaction rates and observed yields had to be calculated for the key compounds acetate, biomass, PHB, carbon dioxide, oxygen and the nitrogen source ammonia from measurements performed during a stable SBR cycle or fed-batch experiment. Both SBR and fed-batch reactor were highly dynamic systems with changing reaction rates and liquid volumes, making the evaluation of experimental data a complex task. A very detailed data analysis was carried out for each SBR cycle measurement and fed-batch experiment. The data analysis included for example the correction of measurements for sampling effects and liquid volume changes, the computation of oxygen consumption and carbon dioxide evolution, and the calculation of the best estimates for all reaction rates and total conversions at each time point with the help of a metabolic model (Chapter 2). The metabolic model was used in order to be able to describe the dynamics of the system and in order to ensure that material balances would close. The metabolic model described the measurements generally very well. The reaction rates computed with the metabolic model showed clearer trends than those calculated without the help of the model. Different operational conditions were tested for the biomass enrichment step (SBR). The first two process parameters investigated were low sludge residence times (SRTs) of 4 d, 1 d and 0.5 d and the impact of different degrees of nitrogen versus carbon limitation (Chapter 3). Low SRTs are required for a high biomass productivity in the first step. The impact of nitrogen limitation was investigated, because many waste streams that are suitable substrates for mixed culture PHA production are nutrient limited. Enrichment of a PHA storing community was successful at 4 d and 1 d SRT, but less successful at 0.5 d SRT. Nitrogen limitation in the SBR generally led to competition for nitrogen and consequently to a selective pressure for high growth rates. Carbon limitation in the SBR led to a PHB storage strategy (high acetate uptake rate) and usually to higher PHB contents (about 70 wt%) in subsequent fed-batch experiments compared to cultures enriched under nitrogen limitation. Carbon limitation in the SBR allowed PHB storing bacteria to benefit more from their ability to store PHB by being able to grow throughout the famine phase. Carbon limitation and SRTs higher than 0.5 d were identified as favourable conditions for the biomass enrichment step in the SBR. Nutrient limited wastewaters may require supplementation with nutrients for this step. Another parameter that was investigated was the reactor temperature (Chapter 4). The reactor temperature will influence the reaction rates, but also the selective pressure in the SBR. The influence on the reaction rates can be investigated by applying short-term temperature changes (i.e. one SBR cycle) while the combined effect on reaction rates and selective pressure can be studied in long-term temperature change experiments. In short-term temperature change experiments the reactor temperature of a stable SBR operated at 20°C was changed for one cycle to 15, 25, 30 or 35°C. It was found that reaction rate changes in the famine phase could be described over the whole temperature range with the Arrhenius equation with one temperature coefficient. For the feast phase different temperature coefficients were estimated for acetate uptake, PHB production and growth. These were only valid for temperatures 5°C higher or lower than the steady state temperature. After long-term changes to either 15 or 30°C the reactor performance changed considerably: At lower temperatures the feast phase was long and a growth strategy prevailed. This culture had a very low PHB storage capacity (about 35 wt%). At 30°C the feast phase was short and a PHB storage strategy dominated. This culture was able to store 84 wt% PHB. Higher SBR temperatures appear to be a good strategy to support the enrichment of PHB storing bacteria. In Chapter 5 we report the most successful operating strategy applied during this thesis. A SBR culture was enriched that was able to store 89 wt% PHB within only 7.6 h in a fed-batch experiment. This culture had been enriched with a longer cycle length of 12 h as compared to our previous studies (4 h cycle length), at 1 d SRT, 30°C and carbon limitation. Another key to the high PHB content was the long operating time under these conditions of over a year. The maximum PHB storage capacity of this culture had improved with time. The long cycle length combined with a low SRT was found to favour growth of bacteria that can store a high amount of PHB at a high rate, since this is needed in order to continue to grow throughout the much longer famine phase. After the operating conditions in the SBR had been optimized, also the PHA production step in the fed-batch reactor was investigated. The temperature in fed-batch experiments did not influence the maximum PHB storage capacity, but only the reaction rates (Chapter 4). Fed-batch experiments were typically conducted using fed-batch systems without nitrogen source in the feed. With the aim of using waste streams as a substrate for PHA production, nutrient limitation or starvation may not always be feasible. We therefore investigated the influence of nitrogen starvation, nitrogen limitation and nitrogen excess on the maximum PHB content obtained in fed-batch experiments (Chapter 6). Under nitrogen starvation conditions the biomass reached a maximum PHB content of 89 wt%, under nitrogen limitation 77 wt% and under nitrogen excess 69 wt%. In the latter two experiments PHB contents decreased after these maxima were reached, because growth led to a dilution of the PHB pool. Nutrient starvation seems thus to be the best strategy for maximal PHB production in the fed-batch step. Chapter 7 summarizes and integrates the findings from all individual studies. In this chapter also some remaining issues are discussed and recommendations for future research are provided. With the aim of using real waste streams in the future and producing other PHAs apart from PHB, the next steps would be the use of more diverse carbon source mixtures and eventually a scale-up of the system. In conclusion, mixed culture PHB production has been successfully optimized in this thesis. A mixed culture was established with the capacity to produce PHB levels as high as in pure culture production processes, and at very high PHB production rates. Cultivation conditions have been identified that lead to a selection of a stable mixed microbial culture with a superior PHA production capacity. Compared to previous work with mixed cultures, a more than four times higher cellular PHB content was obtained. Herewith a highly competitive process has been established that may contribute to the development of a more sustainable and renewable biopolymer production in a future bio-based economy.