Project

Development of a generic screening platform for metabolic engineering applied to the selection of an optimal sialic acid producer

Code
178BW0613
Duration
01 January 2013 → 31 December 2016
Funding
Regional and community funding: IWT/VLAIO
Research disciplines
  • Natural sciences
    • Computational biomodelling and machine learning
    • Synthetic biology
  • Engineering and technology
    • Industrial microbiology
Keywords
sialic acid producer metabolic engineering
 
Project description

Spurred by environmental motives and the limited supply of fossil fuels, chemical industry increasingly relies on microbial engineering to create economically feasible production processes from renewable sources. Several successful metabolic engineering efforts were driven by enabling technologies from synthetic biology, which allow the development of new microbial cell factories for the production of a molecule of interest. One such interesting molecule is N-acetylneuraminic acid (Neu5Ac), a sugar moiety with a vital role in various physiological processes, such as, tumor progression, bacterial infection, infant brain development and, immune responses, which results in various applications in pharmaceutical and food industry. However, the development of these applications is hindered by the limited availability of Neu5Ac due to the lack of sufficient production technologies. This shortage could be solved by metabolic engineering which allow ad hoc rewiring the metabolism of microbes to obtain an optimized biosynthetic pathway for the with maximal productivity. This is a daunting task that requires various tools to modulate gene expression, build genetic circuitery, specifically detect molecules throughout the cell. Therefore, the main objective of this PhD research was the development of advanced enabling technologies for metabolic engineering, allowing the construction of genetic circuitry and detection of small molecules. Recently, the programmable nature of RNA spurred the development of various novel tools based on RNA regulators, which are increasingly employed for various metabolic engineering strategies to maximize productivity. One interesting type of RNA devices to build complex biological systems are riboregulators, which allow rapid control of translation without the need of coexpressed burdensome proteins. However, these devices are limited by the lack of clear design principles, hindering their applicability in metabolic engineering. To address this problem, so called translation inhibiting RNAs (tiRNAs) were developed, which are riboregulators that allow programmable control of protein expression on a post-transcriptional level. These tiRNA devices were created by exploring possibly important features using a design of experiments (DOE). The de novo developed riboregulators repressed translation up to 6% of the original protein expression levels, outperforming the dynamic range of previously described riboregulators. Moreover, compared to previous efforts, the tiRNA regulators created here are designed from scratch and do not require any naturally occurring chassis to function. To link the properties of tiRNA riboregulators to its performance, a partial least squares (PLS) regression model was constructed, further increasing the programmability of the riboregulator. Besides gene expression modulation, RNA technology also allows controlling gene expression based on the presence of small molecules. Recently, various ligand responsive RNA devices, more specifically riboswitches, were previously used in various metabolic engineering strategies as an attractive alternative to their traditional protein counterparts. However, the creation of these riboswitches from in vitro selected aptamers typically involves laborious high-throughput screening efforts. To remove this hurdle in metabolic engineering, a computer-aided design approach was developed, allowing the in silico screening of riboswitches. To quantify the riboswitch capacities of a specific untranslated region (UTR), an objective function was defined based on previously described riboswitches. Using this objective function, 29 potential riboswitches were computationally designed using a simulated annealing algorithm. Subsequently, these riboswitches were evaluated in vivo, yielding functional riboswitches out of the box with 12 out of the 29 created riboswitches activating gene expression more than five fold. However, despite the high probability of yielding functional riboswitches, linking performance to structural or thermodynamic properties remains challenging. Overall, the developed algorithm can help reducing the development times of translational riboswitches, improving their applicability. The natural complex regulation of the microbial metabolism spurred the development of various metabolic engineering strategies, which often require intracellular detection of small molecules. To this end, various biosensors were created based on naturally occurring transcription factors (TFs), typically having limited possibilities to engineer the desired response curve. As a proof of concept, novel biosensors were created that respond to Neu5Ac based on native and engineered promoters that interact with the TF NanR, which were evaluated using a engineered Neu5Ac producing strain. To allow modular biosensor optimization, a NanR binding site was inserted in a constitutive promoter, which resulted in biosensors composed of defined parts. This enables more reliable engineering of the response curve, further expanding the applicability in metabolic engineering. The increased engineering capabilities by the modular design of biosensors was shown by modulating the response of one of the created biosensors by solely changing the ribosome binding site (RBS) used for NanR expression. Also, when exposed to varying Neu5Ac production levels (up to 1.4 ¡À 0.4 g/L extracellular Neu5Ac produced) three biosensors emit fluorescence proportional to amount of Neu5Ac produced. This indicates the broad operating range of these biosensors, a critical property of biosensors for various applications in metabolic engineering. Overall, the range of biosensors was further expanded with various functional biosensors capable of detecting Neu5Ac, which can be applied in various metabolic engineering approaches to produce Neu5Ac with maximal productivity. Overall, various tools were developed in this doctoral research that enable the reliable optimization of microbial cell factories. Specifically, the forward engineering capacity of translation inhibiting riboregulators and translational riboswitches was improved, which further improves the applicability of the various tools originating from the field of RNA synthetic biology. Additionaly, a various techniques were used to create modular biosensors composed of defined parts, which allows reliable response curve engineering. Moreover, various biosensors were creating to detect Neu5Ac in vivo, which was previously impossible and paves the way for various novel metabolic engineering strategies.