Industrial fermentation processes comprise the potential sources of heterogeneity since the microbial cells exhibit an intrinsic cell-to-cell variability that is both affected by and influence the fermentation process itself (reactor conditions and medium composition) over time. However, population-averaged data, assuming uniform cellular behavior in the bacterial cultures, restrict our understandings of how changing environments affect population heterogeneity or how different sub-populations contribute to the fermentation process. In this thesis, the focus was given to study and quantify single cell responses to glucose perturbations in the bioreactor using Escherichia coli as model systems.
In the introduction chapters, the concept of microbial population heterogeneity is briefly introduced and hereafter used as a way to study the heterogeneous fermentation process. Followed by a short experimental outline, a short description of reporter strain construction is given. The reporter system using a low copy plasmid under the regulation of the rrnB promoter or fis promoter was constructed, the expression of which may provide complementary signal in responding to nutrient availability since they have similar growth related expression pattern. Then characterization of the reporter strains was performed in batch, fed batch and continuous cultivations with different environmental perturbations to simulate largescale conditions and population heterogeneity was analyzed by flow cytometry.
Oxygen and glucose oscillation, typical for large-scale cultivation, was studied in Manuscript 1 with the two-compartment scale-down reactor consisting of a wellstirred compartment connected to a plug-flow compartment, which allows reproduction of mixing imperfections. Similar expressions of both rrnB and fis promoters were related to the cultivation mode with different glucose limitation. The sudden glucose relief elevated the ribosomal RNA synthesis as well as its transcriptional activator Fis. Heterogeneities induced by scale-down may affect protein leakage into the broth medium because of the increased membrane permeability and thus resulted in the loss of cellular fluorescence. Such an observation was further investigated in Manuscript 2 with an in-house flow cytometer and PI staining. Growth and cell permeability were monitored in real-time during the process at the single cell level and the performance of the modified on-line analysis was demonstrated by showing similar result of glucose perturbations carried out in chemostat cultivation as shown in Manuscript 1. A “mean to median ratio” of recorded parameters was introduced to detect segregation in the population. At the mean time, the observed suggested that bacterial cells are robust cell factories that they could self-repair membrane integrity and thus contribute to the redistribution of subpopulations.
The study of response adaptation was extended in Manuscript 3. Single glucose pulse experiments were performed in aerobic glucose-limited Saccharomyces cerevisiae and Escherichia coli chemostat cultures. The physiology of S. cerevisiae and E. coli at low growth rates deserves more attention as slow-growing cells were more robust towards stress than fast-growing cells. A high GFP expression and more dynamic changes were observed in slow growing cells than fast growing cells during the glucose oscillation. Population heterogeneity was quantified by introducing new parameters to describe fluorescence distribution data.
As a complement to using reporter strains for detection of heterogeneity in the fermentation context, the use of different fluorescent stains targeting particular physiological features such as membrane integrity and metabolic activity was studied as these will provide tools that easier may be applied in an industrial setting where one maybe do not want to introduce specific reporter strains. In Manuscript 4, characterization and application of the fluorescent dyes such as CTC, SYBR Green, PI, DiBAC4(3) and RSG was used in batch cultivation to assess cellular viability of bacterial cells as regards respiration activity and membrane potential. Cellular heterogeneity was reflected at the transcription level of the ribosomal genes and specific physiologic and metabolic characteristics of the bacteria pictured by the reporter strain and the fluorescence stains.