Monitoring, characterizing, and influencing the poly(γ-glutamic acid) production by Bacillus licheniformis

  • Monitoring, Charakterisierung und Beeinflussung der γ-Polyglutaminsäureproduktion von Bacillus licheniformis

Meißner, Lena Elisabeth; Büchs, Jochen (Thesis advisor); Schwaneberg, Ulrich (Thesis advisor)

Aachen (2016)
Dissertation / PhD Thesis

Dissertation, Rheinisch-Westfälische Technische Hochschule Aachen, 2016

Abstract

Bacillus spp. are used for the production of industrial enzymes, such as proteases for the detergent industry, but are also known to be capable of producing biopolymers such as poly(γ-glutamic acid) (γ-PGA). Biopolymers increase the viscosity of the fermentation broth, thereby impairing mixing, gas/liquid mass transfer, and heat transfer in any bioreactor system. Undesired biopolymer formation has a significant impact on fermentation and downstream processing performance, respectively. The first part of this thesis shows how undesirable γ-PGA formation of an industrial protease producing Bacillus licheniformis strain (A1) was prevented by switching the nitrogen source from ammonium to nitrate. The maximum viscosity at a shear rate of 300 s-1 was reduced by over 90 %, while enzyme production was not affected. Protease activities of 38 and 46 U mL-1 were obtained for ammonium and nitrate, respectively. The applicability was shown for shake flask and lab scale stirred tank bioreactor cultivations. With the presented results, protease production with industrial Bacillus strains is now possible without undesired γ-PGA formation and its negative impact on shake flask and stirred tank bioreactor fermentations.On the other hand, Bacillus species are also used for the commercial production of the biopolymer γ-PGA. The second part of the thesis presents the characterization of the typical γ-PGA producing Bacillus licheniformis strain ATCC 9945 in shake flask cultivations under standard cultivation conditions from the literature. The cultivations were strongly oxygen limited, and after a certain time, the oxygen transfer rates of the parallel running cultures diverged. The respiration activity of some cultures ceased while others continued for several days. An early termination of respiration activity could always be attributed to a low pH value. Sample viscosity, which is an indirect measure of γ-PGA formation, was maximal after 24 to 48 h, thus preceding the parallel cultures’ diverging. The observed non-reproducible growth, therefore, did not interfere with γ-PGA production. After its maximum was reached, the effective viscosity decreased severely, indicating the start of the enzymatic degradation of γ-PGA. The maximum of effective viscosity coincided with a depletion of citric acid and phosphate. Different medium compositions and filling volumes were investigated for their influence on γ-PGA production. A higher initial citric acid concentration increased the maximum effective viscosity significantly while a higher initial phosphate concentration had no impact under standard conditions. More citric acid in combination with more phosphate was beneficial for the γ-PGA production. However, the increase of citric acid and ammonium concentrations at the same time led to low pH values and early termination of the respiration activity. By using different filling volumes, different levels of oxygen transfer capacities were induced, resulting in oxygen unlimited or limited conditions. The different oxygen transfer capacities were shown to exhibit a strong influence on growth and γ-PGA formation. It could be demonstrated that a certain level of oxygen limitation is useful for the γ-PGA production process. In the third part of this thesis, online measurement of the volumetric power input is applied to follow γ-PGA formation and degradation continuously during a running cultivation. The online measured power input corresponded well to the offline measured effective viscosity. The determination of molecular weight and concentration of γ-PGA showed that viscosity and power input depended more strongly on changes of the molecular weight than of the concentration. The increase in viscosity due to a higher citric acid concentration or the supplementation of trace elements was also detectable with the online measured power input signal. Moreover, the power input measurement was useful to reveal different production patterns of γ-PGA producing strains, as was demonstrated by a comparison of the strains ATCC 9945 and A1. Finally, a correlation between offline determined viscosity and online measured power input was established which enabled the calculation of a continuous viscosity signal from the online measured power input.