The Caltech iGEM 2007 team is composed of four undergraduates from Caltech and one undergraduate from MIT. Team members are current juniors and seniors in biology, chemistry, chemical engineering, and biological engineering. The team was advised by three graduate students and three faculty mentors.
Our project tries to answer the following question: can viruses be engineered to selectively kill and/or integrate into specific subpopulations of target cells, based on their RNA or protein expression profiles? This addresses an important issue in gene therapy, where viruses engineered for fine target discrimination would selectively kill only those cells over or underexpressing certain disease or cancer associated genes. An even more ambitious goal would be to rewire target cells, by integrating a small gene cassette which would modify the target cell's expression profile, possibly fixing a disease state.
While this is clearly an ambitious goal, we managed to choose a simple model system for this problem suitable for undergraduates working over a summer. The bacteriophage λ is a classic, well studied virus capable of infecting E. coli, another classic model genetic sytem. We therefore seek to engineer a λ strain targeted to lyse specific subpopulations of E. coli based on their transcriptional profiles. Together, λ and E. coli provide a tractable genetic system for this larger problem, while hopefully providing lessons applicable to more ambitious, future projects.
Briefly, our project relies on controlling key viral developmental processes in a target-cell specific manner. In our design, the engineered viruses are capable of entering all cells. The viruses are engineered to lack the native copy of a key developmental gene, while containing a second, regulated, copy which is only expressed when the virus infects specific target cells. Thus, viruses infecting non-target cells stall early in their development and are quickly destroyed by the host. Viruses infecting target cells, however, manage to express these essential genes and successfully complete their infection cycle.
We will first use recombineering techniques to insert amber mutations into three key developmental genes in λ-Zap. Next, a second copy of these genes, controlled by a cis-repressing riboregulator, will be cloned into the phage genome at the ribosome binding site upstream of each of the three critical genes, thus blocking the expression of key viral developmental proteins. As depicted in the diagram below, the expression of trans-activating RNA in the target bacterial host will relieve the repression by opening up the ribosome binding site, enable the translation of the viral developmental gene and allow lysis of the host cell. Hosts which do not contain this RNA will remain intact.
Working with lambda phage has been new and fun for all of us -- please browse our team pages to learn about our results!
Currently, E. coli strains have been constructed that contain a low-copy plasmid construct where one of three key developmental viral genes - coding for the Cro, N, or Q proteins - is regulated by a tetracycline-dependent promoter. The addition of anhydrotetracycline (aTc) inactivates the tetracycline repressor and leads to the production of the respective viral protein in the E. coli cells. This allows us to control the concentration of viral protein produced in the cells by adding varying amounts of aTc to the bacterial growth media. Heterologous N and Q have been shown to complement phages with amber mutations in the respective genes. Adding a cis-repressor to the Q construct lowered production of Q and prevented complementation. We were unable to express sufficient cro from a plasmid to rescue a cro mutant phage.
Multiple riboregulator designs are being tested (for both activation and repression levels), and successful designs will be cloned into the plasmid constructs. Phages resulting from the recombineering process are also being screened for successful N and Q amber mutants.