The mechanism of phiC31 integrase; a unidirectional recombinase for genome engineering

Lead Research Organisation: University of Aberdeen
Department Name: School of Medical Sciences


It is really difficult to cure people with genetic diseases. These diseases are caused by defective genes. The best cure would be to give them a gene that works. Ideally this treatment would need to be given only once in their lifetime because genes, when they are part of the chromosome, are passed faithfully from one cell to the next and so the cure would perpetuate. Although this sounds simple, in practice its very hard. This project concerns a possible way of getting the right gene into a sick person's chromosome. Some viruses that infect bacteria (called bacteriophages or phages) have a way of getting their own genes into the chromosome of their bacterial hosts. This process involves proteins called integrases, because they integrate two pieces of DNA into one. Most integrases use a particular site, an attB site, in the bacterial chromosome preferentially over all others and the phage DNA goes into that site. There is also a preferred site in the phage DNA, the attP site. In order to introduce correct genes into people as a cure for disease, we need to engineer the integrase so that it can find a safe and suitable site for integrating DNA in that person's chromosome. Consequently this project is about understanding how these integrases work so that we can alter them in a rational way. We would like to know, for instance, how integrase is controlled. A feature of integrase is that it is irreversible in the absence of any other phage proteins. This means that once the correct gene is inserted it is there forever, hence the need for only one treatment. In fact what we mean by irreversible is that it can only use the attP and attB sites in the integration reaction. During this process the two halves of attP get split and join up with the two halves of attB site to form hybrid sites attL and attR. Of these four sites integrase only reacts with attP and attB. Integrase detects the presence of attP and attB very early on in the reaction pathway, a stage that brings the two sites together. There is a kind of lock and key interaction between integrases bound to the two sites attP and attB that activates the rest of the pathway to complete integration. Without the right lock and key interaction, such as when integrase is bound to two attP sites, an attP and an attL site or the attL and attR sites, the pathway is blocked. We have recently discovered a small part of integrase, a 'control module', that is part of a mechanism that senses which type of site it is bound to and communicates this information to generate a 'lock' or a 'key' type structure. We have discovered that a truncated integrase (the C-terminal domain or CTD) that lacks the part of integrase necessary for the chemistry of the reaction, but still contains the control module and DNA binding activity, can do the lock and key reaction on its own. We have broken down integrase still further to just the control module which we showed can bind to itself. Is this the lock and key interaction that occurs in full length integrase? We will use more mutants to test ideas about how the control module interacts with itself and whether this is the lock and key interaction or whether it is part of a sensing mechanism that discriminates between attP and attB on DNA binding. We will also try to identify the part of integrase that directly recognises the attP and attB sites and how this interacts with the control module. Finally we will look at the very beginning of the integration reaction, i.e. the process of DNA binding, and use chemicals and thermodynamics to look at how the footprints made by integrase on the attP and attB sites are different.

Technical Summary

The integrase encoded by the Streptomyces temperate phage, phiC31, mediates both integration and excision. In the absence of accessory factors, phiC31 integrase is unidirectional i.e. it only recombines attP x attB in an integration reaction to form the products, attL and attR. As phiC31 integrase is a serine recombinase it has the same (remarkable) mechanism of DNA cleavage and strand exchange as other serine recombinases. Our work is aimed at understanding the control of integrase activity, which we propose involves subtle differences in protein conformation as a result of protein-DNA interactions with different recombination sites. Integrase has a large C-terminal domain (CTD; ~450 aa) that is required for DNA binding, synapsis of attP and attB and prevents integrase from performing the excision reaction in the absence of accessory factors. Native integrase can only synapse attP and attB. We propose that integrase adopts different conformations on attP and attB that enable synapsis between these sites only and prevents synapsis between other pairs of sites. The problem is therefore how does integrase recognise its different recombination sites and how does this recognition translate to the synaptic interface? We have isolated mutants that have lost the ability to prevent excision, i.e. bidirectional or hyperactive mutants. These map to a motif in the CTD that appears to have a coiled coil structure. Preliminary data suggests that this coiled coil motif is central to the recognition of attP and attB and transmitting this information to a synaptic interface. The objectives are to test the predictions on the structure of the putative coiled coil motif and how it interacts with other integrase functions. Information will also be sought on how integrase binds to and discriminates between attP and attB.


10 25 50
publication icon
Bent AF (2013) Structure of PatF from Prochloron didemni. in Acta crystallographica. Section F, Structural biology and crystallization communications

publication icon
Houssen WE (2012) The discovery of new cyanobactins from Cyanothece PCC 7425 defines a new signature for processing of patellamides. in Chembiochem : a European journal of chemical biology

publication icon
Koehnke J (2013) An enzymatic route to selenazolines. in Chembiochem : a European journal of chemical biology

publication icon
Olorunniji FJ (2012) Gated rotation mechanism of site-specific recombination by ?C31 integrase. in Proceedings of the National Academy of Sciences of the United States of America

Description The most significant discovery we made during this research is that integrase actually uses two motifs to recognize its DNA substrates; a zinc finger motif that is used to recognize all the substrates and a motif in a conserved so-called 'recombinase' domain that appears to discriminate between integration and excision substrates. The DNA binding motifs are likely to co-operate in overall DNA binding and recognition. Excitingly for us, we have obtained, using X-ray crystallography, models of the 3-dimensional structure of integrase fragments that include the recombinase domain, so we have started to build hypotheses as to how DNA recognition might be occurring. These findings are an important first step in being able to engineer integrase to target specific DNA sequences for use in integrating new genes into chromosomes.

We also made significant progress in understanding how the RDF, gp3, works with integrase. We know that gp3 binds to integrase to switch its directionality from an integrating enzyme to an excising enzyme. We discovered that gp3 does this by binding to the C-terminal 200 amino acids of integrase, a region that also includes the coiled-coil motif and in which mutations can lead to loss of directional control in integrase. We discovered that gp3 can control an integrase with a quite different amino acid sequence, suggesting that gp3 binds to a structural motif rather than a specific sequence of amino acids. We obtained evidence that gp3 acts by modifying the conformation of integrase when bound to its excision substrates.
Overall we have made real progress in understanding how integrase functions. With further structural information it may become possible to rationally design novel integrases for integration of DNA at desired sequences in chromosomes to extend their use in biotechnology applications.
Exploitation Route Our findings have helped to interpret the first major structural insights on the serine integrases from Greg van Duyne's laboratory at the University of Pennsylvania in 2013.

The RDF that we discovered in combination with phiC31 integrase has been used by others in synthetic biology applications, notable the rapid assembly of DNA to generate novel metabolic pathways and in the development of biocomputing devices.
Sectors Manufacturing, including Industrial Biotechology

Description Resources generated. 1. Plasmids encoding the recombination directionality factor (RDF) gp3 from phiC31. These have been sent to several groups in the UK and world wide. 2. New arabinose inducible integrase expression plasmids have been sent to other laboratories. 3. Mutant integrases with novel properties have been generated, specifically a mutant that is defective in integration but as the wild type in excision. Such a mutant has potential applications in constructing a counting device (Bonnet et al, PNAS 109:8884). 4. Structural models of integrase fragments that will be deposited once the work has been published
First Year Of Impact 2011
Description NPRONET-NIBB 
Organisation University of Manchester
Country United Kingdom 
Sector Academic/University 
PI Contribution Management board member, scoring pump priming applications.
Collaborator Contribution NPRONET provide a network of contacts from academia and industry for our research
Impact NPRONET workshops and pump priming have stimulated networking and research on natural products.
Start Year 2013