An extended DNA recombinase toolkit for mammalian systems

Identifying, screening and optimising a novel recombinase toolkit for mammalian cells. 

The Idea

DNA recombinases perform excision, integration and inversion events on recognition of pairs of cognate DNA recombinase recognition sites. The outcome of the recombination event is directed by the orientation of the two recognition sites. These functions make valuable tools for a wide variety of applications including DNA assembly, control of gene expression, mutation, information processing (e.g. logic gates) and gene delivery/therapy. The field of synthetic biology has used these enzymes to build a variety of interesting devices, including memory modules, cellular counters and the full gamut of logic gates. However, the more sophisticated recombinase-based devices have so far been limited to use in prokaryotic organisms. This is in part due to the limited number of recombinases with high efficiency in other (e.g. mammalian) systems. 

We are particularly interested in the application of recombinases to investigating the neural circuits that subserve specific brain functions and associated disorders. A principal means of investigating this is to express reporter or effector genes in specific pathways at specific times either by delivery of synthetic genetic constructs via viruses or by generating transgenic animal models. This relies on a genetic reporter construct containing recombinase recognition sites designed such that reporter gene expression occurs only in the presence ofrecombinase activity (e.g. excision or inversion). The combination of injection of a retrograde virus carrying a recombinase expression cassette into brain structure A and injection of a second virus containing the reporter construct into brain structure B results in specific labelling of neurons in structure B that project to A. This technique has yielded interesting data on neuroanatomy. However the complexity of brain networks is such that structures typically receive inputs from several other loci. In these circumstances, multiple recombinases are required in order to simultaneously map these multiple circuits and their associated function. Moreover, the efficiency of theserecombinases will become increasingly important where outputs depend on the spatial intersection of their activity. 

We are currently limited in our designs by the small number of high-activity recombinases available. There are currently just two tyrosinerecombinase - Cre and Flp - with high activity across common cell types. Many recombinases have been identified, for example in microbial and phage genomes, however, their activity in mammallian cells is likely to be poor without sequence optimisation. We therefore propose to identify, screen and optimise (e.g. GC content, codon usage) a set recombinases whose activity in mammalian cells have not yet been reported. 

We proposed to: 
1) Identify candidate recombinases (principally from literature) 
2) Optimise the DNA sequences of the recombinsases in silico (GC content, codon usage etc). 
3) Have the optimised DNA sequences synthesised and clone into a repoter construct. 
4) Test the activity and orthogonality (i.e. test for specific for each recombinase for its cognate target sequence vs those of otherrecombinases) in fibroblasts using a transient transfection assay. 
5) Test the most promising recombinases in other mammalian cell lines. 

The recombinase activity data would be made public through publication online and the physical DNA sequences would be made available via Addgene. 

We think the availability of a larger set of high-efficiency recombinases would extend the utility of these versatile tools to a new range of varied applications including those application mentioned. The proposal also has the potential to benefit applications in eukaryotic systems beyond mammals, since the process of optimisation of recombinases from prokaryotic systems would likely involve similar hurdles (e.g. GC content). 

 The Team

Peter Davenport
Research Assistant, Department of Pathology

Maxime Fouyssac
PhD student, Department of Pharmacology

Haydn King 
Graduate student, Department of Pathology


Project Outputs

Project Report
Summary of the project's achievements and future plans

Project Proposal
Original proposal and application
Download PDF >> 

Outcomes, Outputs and Progress

Serine-type site-specific DNA recombinases are valuable tools for conditional expression of genetic constructs. We have a particular interest in mammalian neurobiology, where the currently small number of efficient serine recombinases limits the role these enzymes can play in elucidating (for example) neural connectivity. The core aim of this project is to identify and optimise serine recombinases for function in mammalian cells.

We have 1) constructed plasmid-based fluorescence reporters assays for recombinase activity, 2) demonstrated the function of these reporter assays in human fibroblast cells, 3) designed codon-optimised open-reading frames for a set of 6 recombinases whose functions have not previously been tested in mammalian cell lines. Having validated our reporter system, we plan to have these codon-optimised genes synthesised and to insert them into our validated vector backbones for characterisation.

Genetic constructs

Figure 1. Design of recombinase reporter and expression systems. Co-transfection of a recombinase expression plasmid (e.g. pCAG-Cre) with its cognate reporter plasmid (e.g. pCAG-loxP-STOP-loxP-ZsGreen) should result in recombination of the reporter and hence expression of ZsGreen fluorescent protein.

In order to assess recombinase efficiency we tested two designs (see figure 1):

1)      A terminator-excision system consisting of a transcriptional termination region flanked by recombinase recognition sites: removal of the transcriptional terminator facilitates RNA polymerase progression through the reporter ORF.

2)      An ORF-inversion system consisting of an inverted reporter ORF (ZsGreen) flanked by recombinase recognition sites.

We have assembled and tested a terminator-excision system for Cre (a tyrosine recombinase) and an ORF-inversion system for Bxb1 (a serine recombinase system). We chose Cre and Bxb1 for our test systems because these recombinases are among the most efficient DNA recombinases for excision and inversion reactions, respectively. We are using these systems to validate our strategy and will use them as references for the activities of other recombinases (see below). We in the process of constructing a terminator-excision system for Bxb1 for comparison.

Validation of reporter constructs

Figure 2. Validation of reporter and recombinase expression constructs. A) Co-transfection of Cre reporter (pCAG-loxP-STOP-loxP-ZsGreen) and recombinase constructs (pCAG-Cre). Green cells indicate successful recombination events. B) Relative recombination efficiencies of Cre and Dre reporter/expression systems.

The above plasmids were transfected into NIH 3T3 human fibroblasts to test for recombinase activity. Images of cells were captured using epifluorescence microscopy at 24 hours post-transfection. Transfection with reporter plasmids alone resulted in low background fluorescence that could easily be distinguished from constitutively fluorescent cells. As hoped, co-transfection of recombinase and reporter plasmids resulted in populations of brightly fluorescence cells, illustrated in figure 2A. Green cells in figure 2A indicate successful recombination of plasmid pCAG-loxP-STOP-loxP-ZsGreen by Cre recombinase expressed from plasmid pCAG-Cre. Similar results were obtained for the Bxb1 ORF-inversion system, though with lower efficiency (see figure 2B).

Ongoing work

  • Characterize the reporter constructs using flow cytometry.
  • Construct optimised recombinase expression cassettes.
  • Characterize the full set of expression cassettes by microscopy and flow cytometry.
  • Share reporter and recombinase plasmids via Addgene as well as associated online methods for open access to the scientific community.