(643b) Experimental and Model-Based Design of the Microbial High-Yield Production of PHB in Alcaligenes Latus

Various bacterial species can accumulate intracellularly polyhydroxyalkanoate (PHA) granules as energy and carbon reserves (Yamane et al., 1996). Unlike petrochemical-derived polymers, PHAs are made from renewable sources. Moreover, they have excellent biodegradability, biocompatibility and zero-toxicity characteristics. A major problem in commercializing PHAs is their high production cost. These bioplastics are currently far more expensive than petrochemically derived plastics and are therefore used mostly in applications that conventional plastics cannot perform (e.g., medical, pharmaceutical applications, etc.). Current microbial fermentation PHA production processes cannot challenge present polymer manufacturing technologies. Thus, research work leading to the economical production of PHAs is of utmost significance (Nath et al., 2008).

In this study, the fermentative production of poly-β-hydroxybutyrate (PHB) is investigated. Optimal culture conditions and optimal fermentation strategies, aiming at the maximization of the biomass concentration, the accumulated polymer mass and the total productivity, are examined both experimentally and theoretically. An additional objective is to identify the process parameters that control the molecular weight and end-use properties of the biopolymer. Furthermore, the problem of efficient polymer recovery from the cells and polymer characterization are also dealt. The above objectives were pursued in relation to the microbial production of PHB in Alcaligenes latus bacteria. The bacteria were cultivated under different nutrient conditions in flasks and in a lab-scale bioreactor. The effects of different operating and environmental process parameters (i.e., nutritional elements, dissolved oxygen concentration, residence time in the stationary phase, feeding policy, etc.) on the overall PHB productivity were examined. Also, the effects of the same process parameters on the molecular properties of PHB (e.g., average molecular weights and molecular weight distribution) were experimentally investigated.

For model-based optimisation of the microbial fermentation process with respect to biopolymer productivity and polymer quality, an advanced mathematical model, capable of describing the fermentation process dynamics is developed. In particular, a multi-scale modeling framework is developed for the description of the PHB microbial production in Alcaligenes latus. This framework comprises individual models at different length/time scales namely, an unstructured metabolic model and a polymerization kinetic model (Kawaguchi and Doi, 1992).

The microbial production of PHB was studied in both flasks and a lab-scale fermentor (i.e., glass vessel bioreactor of 2 lt. working volume - New Brunswick Scientific, BioFlo 110) under batch or fed-batch operating conditions. The fermentation of A. latus was monitored using a number of on-line measurements including, temperature, pH, DO and composition of the gas outflow stream. Temperature and pH were controlled at 30 ± 0.01 °C and 7.0 ± 0.05, respectively. On the other hand, the biomass and polymer concentrations, the substrate concentrations (e.g., carbon source, nitrogen source) in the fermentation medium and the molecular properties of the polymer were measured off-line. A gel permeation chromatograph (GPC), coupled with an on-line viscometer (Viscotek model 210) and a refractive-index (RI) detector was used to measure the molecular weight distribution of the PHB. For the extraction and recovery of the intracellularly accumulated PHB, a separation/purification protocol was adopted. Accordingly, the harvested cells were mechanically disrupted via sonication and then treated with CHCl₃ and cold methanol for the extraction and precipitation of PHB.

Within the experimental framework described above, a PHB accumulation of the order of 55-60% wt. of the cellular dry mass (CDM) was measured under batch operating conditions. A maximum value for the PHB concentration and intracellular content were observed for an initial carbon to nitrogen ratio (C/N) equal to 10 and a residence time at the stationary phase equal to 1 hr. It was found that the average molecular weight of the PHB exhibited a monotonous dependence on the (C/N) ratio. Thus, for low values of the (C/N) ratio, the M_w was relatively low (116,800 g/mol). However, when the value of the (C/N) ratio increased, the value of M_w increased too (i.e., up to 2,576,000 g/mol). The nutrients assimilation rate, the biomass growth rate and the PHB accumulation rate were highly dependant on the amount of oxygen and its time-addition policy during cell cultivation. Notice that long residence times during the stationary phase resulted in a significant decrease of M_w since degradation of the PHB polymer commences right after the onset of the stationary phase. It was found that the PHB content increased up to 90-95% w.t. of CDM when a single pulse of the culture medium was introduced during the exponential phase. For batch fermentation, a CDM concentration of 9 g/l was obtained, while the accumulated PHB concentration was 8.5 g/l. Finally, the above described PHB separation/purification protocol was assessed with respect to the polymer extraction efficiency and product molecular properties. Long sonication times clearly improved the polymer extraction efficiency, however, at the cost of lower polymer molecular weights.

Based on the above experimental results, a multi-scale mathematical model was developed for the description of the cell metabolism, in terms of substrate utilization, biomass growth, PHB accumulation and the prediction of the molecular weight distribution of the PHB in Alcaligenes latus bacteria. The pathway of the central aerobic carbon metabolism, that leads to the production of PHB and biomass growth in this bacterium, consists of a large network of biochemical reactions. In the present study, the complex metabolic network was simplified since we were only interested in the prediction of specific metabolites. Thus, through the integration of the metabolic model with the polymerization kinetic model, the monomer production rate (i.e., the metabolite of interest for the polymerization model) could be calculated in a computationally efficient way. The accuracy of the integrated mathematical model was tested against our experimental measurements. Subsequently, the model validity was assessed against experimental measurements derived under different fermentation policies. The present model is capable of predicting the consumption rates of various substrate species (i.e., carbon, nitrogen and oxygen sources), the residual and total biomass growth rates, the PHB production rate and the time evolution of the molecular properties (i.e., molecular weight distribution, average molecular weights) of PHB. The developed mathematical model is used for the calculation of the time-optimal operating policies (e.g., initial conditions and time-optimal feeding profiles) to optimize the operation of our fed-batch fermentation system.

References

? Kawaguchi Y., Doi Y. (1992) ?Kinetics and mechanism of synthesis and degradation of Poly(3-hydroxyburate) in Alcaligenes eutrophus?, Macromolecules, vol. 25, pp. 2324-2329.

? Nath A., Dixit M., Bandiya A., Chavada S., Dedai A.J. (2008) ?Enhanced PHB production and scale up studies using cheese whey in fed batch culture of Methylobacterium sp.?, ZP24? Bioresource Technology, vol. 99, pp. 5749-5755.

? Yamane T., Fukunaga M., Lee Y. W. (1996) ?Increased PHB productivity by high-cell-density fed-batch culture of Alcaligenes latus, a growth-associated PHB producer? Biotechnology and Bioengineering, vol. 50, pp. 197-202.