Interleukin-3 is a cytokine, which acts on many target cells within the haemopoietic system, often in synergy with the other cytokines. Streptomyces lividans NCIMB 11416/IL3 p002 secreting human interleukin-3 was used as the host organism in this study of improving target protein production. Streptomyces also produces several proteases including extracellular endoprotease that truncate the N-terminus of the recombinant protein. Federal guidelines and regulations banning animal-derived medium components necessitate the refinement or redevelopment of industrial medium formulations. The development of a defined medium without animal products is most desirable for the production of pure and safe biological products. The objective of the proposed research was the development and application of engineering methodology for the development of a defined medium and the analysis and optimization of a bacterial bioprocess for recombinant protein production. The underlying hypothesis is that a significant improvement of target protein productivity is achievable by using appropriate optimization techniques. During the first phase of this study the task was to develop a systematic procedure for the design and optimization of a chemically defined medium. The study aimed at replacing casein peptone in conventional medium for S. lividans with essential amino acids and determining the optimum proportion of the amino acids. To accomplish this, starvation trials with growth limiting amino acids were performed to establish the baseline for the nutritional requirement. The starvation trials revealed that essential amino acids for growth and product formation are amongst the following eight amino acids: Arg, Asn, Asp, Glu, Leu, Met, Phe, and Thr. Following these preliminary experiments, a statistically based experimental method called mixture experiments along with distance-based multivariate analysis revealed that Asp, Leu, Met, and Phe were the essential amino acids. Then, another mixture experiment design known as simplex lattice design was performed and artificial neural networks were employed to obtain the optimum proportions of the essential amino acids. The optimal medium was found to be composed of 56% Asp, 5% Met, and 39% Phe. It was found in previous studies that in complex media, several types of protease are produced during fermentation. Using the defined medium no proteolytic activity was detected in the fermentation broth. The second optimization method was based on metabolic flux analysis. A comprehensive metabolic network was developed for S. lividans. The metabolic network included carbohyderate and amino acid metabolism in both anabolic and catabolic reactions. According to the experimental results, the time course of the fermentation was divided into two phases, Phase E1 and Phase E2. In the first phase amino acids were used as a nitrogen source and in the second phase ammonia was the nitrogen source for growth and product formation. The metabolic network was used to form a set of linear algebraic equations based on the stoichiometry of the reactions by assuming pseudo-steady state for intracellular metabolites. The metabolic flux model consisted of 62 intracellular metabolites and 91 biochemical reactions. Two different objective functions were considered for optimization: maximizing the specific growth rate and minimizing the redox equivalent. A linear programming approach was used for optimizing the objective functions. The proposed model was able to predict the specific growth rate very accurately with a maximum error of 10%. The oxygen uptake rate and carbon dioxide evolution rate were evaluated with maximum error of 27% and 35%, respectively. Sensitivity analysis revealed that amino acid uptake was the growth limiting flux during the Phase E1 of the fermentation. During Phase E2 the uptake rate of ammonia had a significant effect on the specific growth rate. Sensitivity analysis of the specific growth rate and redox potential with respect to the biomass components showed that any additional supply of biomass building blocks (amino acids, nucleotides) would not significantly affect the specific growth rate and redox potential production as well as the calculated flux pattern.

Aquatic organisms in response to toxic insults from environmental pollutants activate defence systems including the aryl hydrocarbon receptor (AhR) in an attempt to metabolize and excrete these toxicants and their metabolites. These detoxification mechanisms however may come with certain energetic costs. I hypothesize that the activation of the AhR by β-Naphthoflavone (β-NF), a model AhR agonist, results in increased energetic costs requiring metabolic reorganization in rainbow trout hepatocytes. While the results obtained suggest that there are no significant energetic costs of AhR activation, analysis of enzyme activities suggests possible metabolic reorganization. This study also showed significant changes in cellular processes in hepatocytes over the incubation periods which previously were not reported. Furthermore, for the first time in fish hepatocytes, metabolic flux analysis (MFA) was used to examine intra-cellular metabolism, the applicability of which is discussed.

In this study, firstly, bioprocess characteristics for Serine Alkaline Protease (SAP) production, using recombinant Bacillus subtilis carrying pHV1431::subC, were examined. The cell concentration, substrate concentration, SAP activity and SAP synthesis rate profiles demonstrated that the system reaches to a steady state in terms of cell growth and SAP synthesis between t=15-25 h, therefore, this time interval is appropriate to employ both metabolic flux analysis and metabolic control analysis, which apply strictly to steady state systems. After that, three separate perturbations were introduced by addition of aspartate to the production medium at a certain time of the bioprocess. The response of the cells were observed and / by comparing the changes in intracellular reactions of aspartate pathway, Asn, Thr and Ile productions were determined to be the bottlenecks in aspartate pathway and the branchpoints splitting from Asp and AspSa were identified to be weakly rigid branchpoints. Lastly, metabolic control analysis principles were applied to determine the elasticity and flux control coefficients of the simplified aspartate pathway. Aspartate formation reaction and Lys, Thr, Ile, Met producing group share the control of asparagine synthesis. The results revealed that lysine producing branch flux dominates the other branch fluxes, therefore to eliminate bottlenecks and increase SAP production, the activity of the branches leading to the formation of Asn, Thr and Ile should be increased while decreasing the activity of lysine synthesizing branch. This could be achieved either by genetic manipulation or by addition of specific inhibitors or activators to the system.

In this study, the effects of oxygen transfer (OT) on beta-lactamase production and on intracellular reaction rates were investigated with Bacillus licheniformis ATCC 2597. In order to clarify the oxygen transfer effects on the production of beta-lactamase, firstly a glucose based defined medium was designed and using this medium, the effects of bioreactor operation parameters, i.e., pH and temperature, on beta-lactamase activity and cell formation were investigated in laboratory scale batch-bioreactors using shake bioreactors having V=33 ml working volumes. Among the investigated bioprocess conditions, the highest beta-lactamase activity was obtained as A=115 U cm-3, in the medium with 7.0 kg m-3 glucose, 7.1 kg m-3 (NH4)2HPO4 and the salt solution, at pH0=7.5, T=37C, N=200 min-1. At the optimum conditions found in laboratory scale the effects of OT on cell generation, substrate consumption, product (beta-lactamase) and by-products formations were investigated at three different air inlet (Q0/ VR = 0.2, 0.5 and 1 vvm) and at three agitation rates (N=250, 500, 750 min-1) in V = 3.0 dm3 batch bioreactors consisting of temperature, pH, foam, stirring rate and dissolved oxygen controls. Along with the fermentation, cell, substrate and by-product concentrations, beta-lactamase activity, yield coefficients, specific rates, oxygen uptake rates and the liquid phase mass transfer coefficient values were determined. The highest beta-lactamase activity was obtained at 0.5 vvm 500 min-1 and at 0.2 vvm 500 min-1 conditions as ca. A=90 U cm-3 while the highest cell concentration was obtained as Cx=0.67 kg m-3 at 0.5 vvm 750 min-1 and at 0.2 vvm 750 min-1 conditions. KLa, increased with the increase in the agitation and aeration rates and its values varied between 0.007-0.044 s-1 and oxygen uptake rate varied between 0.4-1.6 mol m-3 s-1. Finally, the influence of OT conditions on the intracellular reaction rates was investigated using metabolic flux analysis to evaluate the effects of oxygen on the metabolism. Keywords: beta-lactamase, production, Bacillus, oxygen transfer, metabolic flux analysis

Hyaluronic acid (HA) is a high molecular weight polysaccharide that is mainly produced by animals and certain bacteria. This polymer is biocompatible and possesses desirable rheological properties that are accentuated by high molecular weight. Diverse therapeutic applications have developed which harness these features. Pharmaceutical grade HA is mostly extracted from animal tissue. The HA derived from this source is suitable for most pharmaceutical preparations but there is growing pressure to avoid animal tissue products. This has provided the incentive to expand microbial-based HA manufacturing. However, the inherent low molecular weight of the polymer derived via this route has hampered widespread acceptance of microbial HA. This thesis examined the ramifications of improving the cellular economy of the HA-producing, gram-positive bacterium, Streptococcus zooepidemicus. Improved cellular economy is believed to be a prerequisite for achieving superior HA yields and molecular weights in this microorganism. This work examined the metabolic variation that accompanied the shift to more efficient modes of carbon utilization. In particular the effect of different sugar sources, uncoupling growth and polymer formation, and changes to the cellular oxidoreduction capacity were studied in more detail. This study utilized different sugar sources to enhance the recovery of energy. Fermenting glucose, fructose and maltose produced contrasting patterns of growth and HA formation. Culturing the organism in maltose caused a shift towards energy-efficient heterofermentative metabolism. Maltose-cultured cells displayed a biphasic pattern of metabolism. The first stage corresponded to a growth phase in which biomass synthesis profited from the increased energy yield. The second stage corresponded to an arginine-deficient stationary phase where the majority of the HA was formed. The fermentation rate was slower during stationary phase but continued to support HA biosynthesis. This bisphasic metabolism proved to be beneficial. A protracted stationary phase led to higher molecular weight HA. Fructose was unable to sustain a comparable polymer yield or molecular weight as glucose or maltose. There was evidence that the arginine deiminase pathway was responsible for the premature depletion of arginine in maltose-fermenting cultures. The accumulation of biomass exhibited a concentration-dependent response to the amount of glutamine in the medium. A second arginine transporter possessing a low affinity for glutamine could explain this phenomenon. Arginine consumption was slower when the glutamine level was elevated. This may indicate competition for a common transmembrane carrier. An elevated energetic yield and ATP formation rate were features of aerobic maltose metabolism. The relative improvement in biomass and HA yields were substantially greater for cultures fermenting maltose compared to glucose. However, no improvement in molecular weight compared to glucose was observed. A major factor contributing to the success of aerobic maltose fermentation was the particularly high NADH oxidase flux. This enzyme reoxidizes reduction equivalents in a reaction that is physically decoupled from the production of reduced metabolic products. Less lactate and ethanol accumulated in the presence of high NADH oxidase levels but acetate production was stimulated leading to an improved energetic yield. This result prompted an investigation into the effect of elevating the NADH oxidase level. The native NADH oxidase gene was sequenced and cloned into an inducible expression plasmid and introduced into S. zooepidemicus. Overproduction of this enzyme led to the desired improvement in ATP yield. A significant improvement in biomass yield was demonstrated. HA yield and molecular weight were not affected. Lactate and acetate were the main fermentation products. At high induction levels the quantity of lactate and acetate approached limiting levels and pyruvate overflow was more pronounced. This was attributed to insufficient flux capacity of the pyruvate dehydrogenase enzyme complex. The application of metabolic engineering to S. zooepidemicus has provided some insight into the regulation of energy metabolism in this microorganism and its relationship to HA synthesis. This study has observed that the specific rate of HA synthesis is correlated to the sugar uptake rate but is unaffected by the ATP yield. Under present conditions the formation of HA is not limited by the availability of energy. Nonetheless, microbial HA production will benefit from maximizing energetic yield. It was demonstrated that less catabolic carbon was expended to support biomass formation if the energetic yield was high. Therefore more residual carbon was available for HA synthesis.

A 13C-Análise de Fluxos Metabólicos (13C-MFA) tornou-se uma técnica de alta precisão para estimar fluxos metabólicos e obter informações importantes sobre o metabolismo. Este método consiste em procedimentos experimentais, técnicas de medição e em cálculos para análise de dados. Neste contexto, os grupos de pesquisa de engenharia metabólica necessitam de ferramentas computacionais precisas e adequadas aos seus objetos de estudo. No presente trabalho, foi construída uma ferramenta computacional na plataforma MATLAB que executa cálculos de 13C-MFA, com balanços de metabólitos e cumômeros. Além disso, um módulo para estimar os fluxos metabólicos e um módulo para quantificar as incertezas das estimativas também foram implementados. O programa foi validado com dados presentes na literatura e aplicado a estudos de caso. Na estimação de fluxos de Pseudomonas sp. LFM046, identificou-se que esse micro-organismo possivelmente utiliza a Via das Pentoses em conjunto com a Via Entner-Doudoroff para a biossíntese de Polihidroxialcanoato (PHA). No design ótimo de experimentos para uma rede genérica de Pseudomonas, identificou-se a glicose marcada no átomo cinco como um substrato que permitirá determinar o fluxo na Via das Pentoses com menor incerteza. / 13C-Metabolic Flux Analysis (13C-MFA) has become a high-precision technique to estimate metabolic fluxes and get insights into metabolism. This method consists of experimental procedures, measurement techniques and data analysis calculations. In this context, metabolic engineering research groups demand accurate and suitable computational tools to perform the calculations. A computational tool was implemented in MATLAB platform that performs 13C-MFA calculation, using metabolite and cumomer balances, as well as a module to estimate the fluxes and a module to quantify their uncertainty. The program was validated with some classical cases from literature. From the flux estimates of Pseudomonas sp. LFM046, it was identified that the microorganism possibly uses the Pentose Phosphate Pathway along with the Entner-Doudoroff Pathway for Polyhydroxyalkanoate (PHA) biosynthesis. From the optimal experimental design for a generic Pseudomonas network, it was possible to conclude that glucose labeled at atom five is the best option to determine the flux in the Pentose Phosphate Pathway with smaller uncertainty.

Aquatic organisms in response to toxic insults from environmental pollutants activate defence systems including the aryl hydrocarbon receptor (AhR) in an attempt to metabolize and excrete these toxicants and their metabolites. These detoxification mechanisms however may come with certain energetic costs. I hypothesize that the activation of the AhR by β-Naphthoflavone (β-NF), a model AhR agonist, results in increased energetic costs requiring metabolic reorganization in rainbow trout hepatocytes. While the results obtained suggest that there are no significant energetic costs of AhR activation, analysis of enzyme activities suggests possible metabolic reorganization. This study also showed significant changes in cellular processes in hepatocytes over the incubation periods which previously were not reported. Furthermore, for the first time in fish hepatocytes, metabolic flux analysis (MFA) was used to examine intra-cellular metabolism, the applicability of which is discussed.

Continuous fermentations were performed in a novel external-recycle, biofilm reactor using D-glucose and CO2 as carbon substrates. Corn steep liquor (CSL) and yeast extract (YE) served as nitrogen sources. In anaerobic fermentations using medium containing CSL and YE, succinic acid (SA) yields were found to be an increasing function of glucose consumption. The ratio of SA to the major by-product, acetic acid (YAASA), increased from 2.4 g g-1 at a glucose consumption of 15 g L-1, to 5.7 g g-1 at a glucose consumption of 46 g L-1. For medium containing no CSL, YAASA remained near 1.97 g g-1, exceeding this for cases where biofilm grown on CSL-containing medium was present. The ratio of formic acid to acetic acid (YAAFA), for CSL-containing medium, decreased from an equimolar value (0.77 g g-1) at a glucose consumption of 10 g L-1 to zero at 46 g L-1 glucose consumed. In contrast, YAAFA for YE-only medium remained at 0.77 g g-1. Therefore, pyruvate was metabolised solely by pyruvate-formate lyase when no CSL was present. The highest SA yield obtained on glucose, SA titre and SA productivity were 0.91 g g-1, 48.5 g L-1 and 9.4 g L-1 h-1, respectively, all for medium containing CSL. Medium that included CSL significantly outperformed medium that excluded CSL, achieving 64%, 21% and 203% greater SA titres, yields on glucose and productivities respectively. Metabolic flux analyses based on the established C3 and C4 metabolic pathways of Actinobacillus succinogenes revealed that the increase in YAASA, for CSL-containing fermentations, could not be attributed to the decrease in formate and biomass formation, and that an additional source of reducing power was present. The fraction of reducing power (NADH) unaccounted for increased with glucose consumption, suggesting that the maintenance or non-growth metabolism encountered at higher SA titres differs from the growth metabolism. It is postulated that the additional reducing power originates from an active pentose phosphate pathway in non-growing cells or from an undetected component(s) in the fermentation medium. No major metabolic flux variations were found in fermentations that excluded CSL. / Dissertation (MEng)--University of Pretoria, 2013. / gm2014 / Chemical Engineering / unrestricted

<p>The fast advancing of high-throughput technology has reinforced the biomedical research ecosystem with highly scaled and commercialized data acquisition standards, which provide us with unprecedented opportunity to interrogate biology in novel and creative ways. However, unraveling the high dimensional data in practice is difficult due to the following challenges: 1) how to handle outlier and data contaminations; 2) how to address the curse of dimensionality; 3) how to utilize occasionally provided auxiliary information such as an external phenotype observation or spatial coordinate; 4) how to derive the unknown non-linear relationship between observed data and underlying mechanisms in complex biological system such as human metabolic network. </p> <p><br></p> <p>In sight of the above challenges, this thesis majorly focused on two research directions, for which we have proposed a series of statistical learning and AI-empowered systems biology models. This thesis separates into two parts. The first part focuses on identifying latent low dimensional subspace in high dimensional biomedical data. Firstly, we proposed CAT method which is a robust mixture regression method to detect outliers and estimate parameter simultaneously. Then, we proposed CSMR method in studying the heterogeneous relationship between high dimensional genetic features and a phenotype with penalized mixture regression. At last, we proposed SRMR which investigate mixture linear relationship over spatial domain. The second part focuses on studying the non-linear relationship for human metabolic flux estimation in complex biological system.  We proposed the first method in this domain that can robustly estimate flux distribution of a metabolic network at the resolution of individual cells.</p>

<p>The data in this dissertation describes changes in metabolic flux in bovine cell culture and bovine hepatic models. </p>