13TSB_SynBio: Synthetic biology to improve antibiotic production

Bacterial infections are a major cause of death world wide and antibiotics provide one key resource for controlling them.
Mupirocin is a successful antibiotic used against Gram positive bacteria, particularly MRSA which is associated with both
hospital- and community-acquired infections and is resistant to most currently available antibiotics. It is also a standard
treatment to remove MRSA from the skin and nose of healthcare workers. The market for the antibiotic is growing in China
and other parts of the developing world and GSK wish to increase production without expanding production-plant/fermenter
capacity or running costs. The Thomas group have carried out extensive research on the mupirocin biosynthetic cluster in
the soil bacterium Pseudomonas fluorescens and have studied how the genes are switched on and controlled. As a result,
we have clear strategies for increasing production by manipulating the gene cluster.

However, the set of genes coding for the protein factory that makes mupirocin is complex and occupies a segment of about
75,000 base pairs of DNA. There are more than 30 genes in the cluster so that it is difficult to manipulate it. Synthetic
biology (building the genes from chemcially made DNA to our own design) should provide a convenient way to do this and
this project gives an opportunity to validate that idea. By rebuilding the genes we can change the code so that it is
optimised for fast and increased protein synthesis and at the same time we can split the DNA into convenient "Biobricks"
(the building blocks for Synthetic Biology) which can be assembled in different orders and supplemented with additional
DNA sequences that increase the extent to which they are switched on.

Confidence that the current gene cluster is not the only efficient way to configure the genes comes from our discovery that
the genes that make a related plasmid called thiomarinol are arranged in a different order. On top of that we have found
that increased production of the activator MupR, in the existing genetic organisation, can increase production up to 20-fold.
We will therefore first introduce mutations that produce more MupR and we will then systematically insert DNA that
promotes expression of these genes in a MupR-dependent way, to increase the productivity of each bacterium. We will use
state of the art techniques to assess the effect of these changes and see how it affects antibiotic production in shake flasks
and then on a larger scale in fermenters. If successful, these changes will be incorporated into the design of the new
genes.

Another feature of the gene cluster is that the order of genes is not very logical - often genes in a cluster are lined up in the
way they work in the biochemical pathway. We will therefore shuffle the mupirocin gene (Biobrick) order to increase
pathway efficiency and we will screen derivatives for increased production. We will do this using an enzyme (Int) that
deliberately shuffles genes in bacteria. We will insert DNA that allows Int to work between Biobricks and then transiently
express Int to shuffle the genes to produce many permutations of gene order. This approach has been validated by others
and shown to improve the efficiency of the E. coli tryptophan biosynthetic operon - a well studied model system. Bacteria
will be assessed for increased production in the lab and in fermenters as above.
Finally, to explore how the genes can be further improved we will add extra functional units to key modules of the pathway
to increase throughput capacity and to fuse genes to create new multifunctional genes. Gene fusions may increase
efficiency by ensuring that protein partners fold together and subsequently catalyse successive enzymic steps more
efficiently. Examples of both of these sorts of changes are found in other biosynthetic factories.

The project will use synthetic biology to re-engineer the mupirocin biosynthetic gene cluster for increased biosynthetic output. It will involve a combination of directed and partially random changes followed by selection for increased antibiotic production. The directed changes will: introduce new gene expression signals at strategic points in the gene cluster to increase expression; introduce additional acyl carrier proteins to increase the carrying capacity of each component of the biosynthetic factory; create fusions of partner proteins that work together. The partially random changes will be achieved by flanking the Biobricks of the cluster with sequences that are target sites for the site-specific integrase enzyme. This will allow random reassortment of the assembled modules when integrase is supplied in trans. Gene clusters with novel gene orders will be screened using a novel screening procedure that selects bacteria producing antibiotic or elevated levels of antibiotic.

The most obvious way in which this work will have impact is that it aims to improve the production of an antibiotic that is
already used clinically and provides an important line of defence against the super-bug MRSA. If it is possible to achieve
an improvement in yield then this should allow the antibiotic to be produced more cheaply, increasing the competitiveness
of GSK production compared to competitors who have entered the field since the antibiotic went off patent. Cheaper
antibiotic would mean that it could reach a larger market in as yet relatively poor countries, with a concomitant increase in
the quality of healthcare.

A side effect of increased yield may be that it is easier to purify the antibiotic, making it possible to market the GSK brand of
mupirocin on the basis that it is purer than products of their rivals. This should provide further evidence to GSK that
working with UK researchers is productive, which is appropriate since mupirocin as an antibiotic was first developed in the
UK in the 1970s and indeed was awarded a Queen's Award to Industry in 1992. Improvement in productivity should also
ensure the long term economic viability of mupirocin production in the UK. This will also then have further impact on the
UK economy through the supply chain of both production and marketing of mupirocin.

A secondary impact will relate to the knowledge the work will generate if it is successful. We plan to publish as much of
this work as possible so long as it does not jeopardise the commercial exploitation of key improvements in production and
this will involve the broader implications of our results so that the impact of the work will go far beyond just those interested
in mupirocin production. There is growing interest in the expression of biosynthetic clusters because of the need to activate
clusters discovered by genome sequencing projects and found to be essentially cryptic (inactive). Optimising expression to
increase product yield by genetic manipulation/synthetic biology will be increasingly used in this endeavour. The link
between gene order and cluster activity is only poorly understood and so if we find a significant improvement this would
provide the basis of further studies to understand why. In addition, the cassette consisting of the mupR, mupX and mupI
genes may constitute a convenient and innovative tool which could be spliced into newly discovered clusters to boost
expression.

Finally, the work will generate a lot of experience in synthetic biology in both the University of Birmingham and in GSK.
This will have knock on effects on the sorts of projects undertaken in future and the approaches that will be used. It will
create a body of expertise that will ripple outwards to colleagues within our organisations and elsewhere. Because we
believe that it will be a valued application of synthetic biology it will increase the positive perception of this technology and
increase the chance that it will seen as a suitable technology for achieving other goals.