Skip to main content

Two-Step-Strategy towards an Orthogonal Bio-System

Final Report Summary - MAZPROTEC (Two-Step-Strategy towards an Orthogonal Bio-System)

The synthesis of chemical compounds for industrial sectors as different as pharma, construction, or car making is one of the central pillars of Europe’s industrial landscape. Increasingly, biological catalysts – enzymes - play an important role here, due to their unparalleled selectivity and due to the fact that they, unlike most of their chemical counterparts, can all work under a common set of reaction conditions and hence can be used together to carry out multiple reactions concomitantly or in sequence in one reactor.
However, preparing enzymes is costly – they need to be prepared with the help of microorganisms. Setting up reaction cascades – sequences of reactions catalyzed by a multitude of reactions – can thus become prohibitively expensive. We are exploring novel ways how to provide such complex reaction cascades in a highly cost effective manner, using the tools made available by nature and chemistry.
One of the central problems in biocatalyst production is that microorganisms use (costly) nutrients primarily to make biomass instead of the enzymes we are interested in. Consequently, channeling resources into only one specific activity – such as enzyme production – is a widely pursued goal in biotechnology research. This is where the Marie Curie - Intra European Fellowship project MazProTec (number 298326, “Two-Step-Strategy towards an Orthogonal Bio-System”) tried to introduce a novel principle into biocatalyst production, exploiting the ability of regular cells of the bacterium Escherichia coli to decide on a cell-wide scale which proteins to make and which not. E. coli uses for this the MazEF toxin-anititoxin system. MazF is the toxin, an RNA-hydrolyzing enzyme that digests mRNA molecules containing triplets of the sequence ACA, which is normally held in check by MazE. In the appropriate situation, MazF becomes active and prevents most mRNA molecules from becoming translated, while ACA-free mRNAs continue to be read out into proteins. In MazProTec, the goal was to use this mechanism to force the cell at one specific point in time to dedicate its resources entirely to the production of desired proteins, thus leading to cells (highly) enriched with specific enzymes. For application, we chose, as an industrially relevant model system, the 12-step synthesis of N-acetyl neuraminic acid (Neu5Ac), a precursor of antiviral drugs such as Zanamivir®, from cheap starting materials such as glucose (Glc) and N-acetyl glucosamine (GlcNAc) (see figure 1).
The project tried to address three specific questions: [i] Is it possible to re-code the genes that are of interest to us in an ACA-free genetic code and still have them overexpressed in an active form? [ii] What is the best strategy to assemble the recombinant genes for a 12-step pathway into a functional and powerful genetic system? And finally: [iii] Are any observable effects sufficiently large to indeed produce a system of superior performance?
Regarding the first question, the answer is a definite yes. Of the four proteins invested in some detail (green fluorescent protein (GFP), red fluorescent protein (RFP), Neu5Ac synthetase NeuB, Neu5Ac synthetase SiaC), 3 showed identical specific activities when expressed from the re-coded gene as when expressed from an ACA-free gene, and one (RFP) showed a slightly reduced specific activity. Furthermore, we have conclusive evidence of 10 more functional enzymes after re-coding of their genes, even if we did not test all of them for specific activity.

In order to assemble the system, different considerations hat to be taken into account. In order to manage the workload to assemble such a pathway, we applied in vitro (Gibson assembly) and in vivo DNA assembly methods (yeast homologous recombination). Next, of course the expression of the gene for MazF has to be tightly controlled to prevent premature activity. The E. coli rhamnose-inducible promoter (Prha) is virtually silent in the absence of rhamnose, and served well as the regulatory structure for the MazF gene. Next, 12 ACA- free genes, originating from 4 different microorganisms, were assembled into three different operons, encoding the entire 12-step pathway. Next to an ACA-free reconstruction of the glycolysis based on standard E. coli genes, the system includes an NADH oxidase from Streptococcus pneumonia (to permanently regenerate NAD+ from the NADH produced during glycolysis), a Neu5Ac epimerase from Synechocystis PCC6803, and a Neu5Ac synthase from Neisseria meningitides. The overall system architecture had to be modular, to allow eventual operon modifications such as gene replacements or translation and transcriptional signal engineering. Finally, the host that was used to receive the recombinant genes was deleted of activities that are known to act as sinks for starting materials, intermediates, or products. In particular, the following six genes nanATEK (Neu5Ac consumption), pykF (pyruvate consumption), and nagK (GlcNAc consumption) were knocked out.

Regarding the third question, we performed first a test with reduced system, specifically a system based on the gene-deleted E. coli strain and the enzymes required for the conversion of GlcNAc and phosphoenolpyruvate (PEP). Here, it could be clearly demonstrated that (a) after over expression of MazF, protein synthesis activities were redirected to a larger extent into the synthesis of recombinant proteins from ACA-free genes than from regular genes. Importantly, this corresponded with a substantial increase (an increase in protein specific productivity by a factor of 4) in Neu5Ac production in the ACA-free system when compared to a similar system but using ACA-containing genes.

We are currently using a proteomics approach to investigate whether a similar (or even more pronounced) increase in Neu5Ac productivity can be obtained when the entire 12-step system is composed of proteins synthesized from ACA-free genes.

In summary, we could show that a massive re-organization of protein synthesis as a consequence of the expression of MazF as a toxin is a perfectly suitable strategy to increase the productivity in multi-step biocatalytic systems. We expect that this method will be an important element in the future construction of biocatalytic reaction cascades, primarily in the fine chemical industry, and might in the short-term contribute to a more sustainable synthesis route to antiviral precursors built on Neu5Ac derivatives. For more information, please contact sven.panke@bsse.ethz.ch