Tumor suppressor protein p53 is a transcription activation factor that is found mutated in more that 50 percent of human cancers. Despite its pathological significance, there is no robust, in vivo, bacterial screen to select for functional mutants of p53. We have developed a transcription interference screen for p53 core domain based on an artificial p53-responsive lac operon, controlling the expression of GFPuv in the host plasmid pGFPuv. The operator region of the lac promoter was replaced with a library of p53 binding elements based on the consensus sequence (RRRCWWGYYY)2. Wild type or wt-like p53 binds to this site, blocking the polymerase leading to a non-fluorescent phenotype. P53 Quad was expressed under the control of another promoter. The host plasmid pACBAD-p53 can be co-maintained in the cell with pGFPuv. Known hotspot mutants of p53, V143A, R175H, R249S and R273H were constructed by overlap PCR and expressed from the pACBAD-p53 host plasmid. Our results show a marked decrease in the fluorescence of pGFPuv when co-transformed with p53-Quad (wt like) and higher fluorescence when co-transformed with the hotspot mutants of p53. We have successfully designed a screen which discriminates between p53 which shows DNA binding activity and variants of p53 that cannot bind or weakly bind to DNA. Our screen provides a simple and quick method to screen for stable and functional variants of p53 and function correlated to stability. The validity of the screen was further proved by biophysical characterizations of stable and functional variants resulting from the screen. Further, the screen can potentially be used in combination with small molecule libraries, for identifying lead compounds that may stabilize p53.
BRCA1 is another tumor suppressor protein for which the mutations are implicated in familial cancers, and its function is governed by its ability to form complexes with its various interacting partners. Its ability to interact with BARD1 through its RING finger domain to form heterodmeric complex is particularly important for its function. In vivo analysis of a comprehensive list of cancer-associated mutations of BRCA1 using split GFP complementation assay indicates that the interface between BRCA1 and BARD1 is surprisingly robust to mutations. We further studied this interface in vitro, choosing a few mutants of BRCA1, V11A, M18K, I21V, L52F and R71G in the absence of GFP fusion. V11, M18 and I21 are at the interface while L52 and R71 are away from the interface. Among these only M18K and V11A significantly affected the compex formation. Since these proteins are soluble only as a complex we utilized this to analyze the interaction in vitro. Co-expression of the various mutants of BRCA1 with wild type BARD1 using compatible plasmids with orthogonal tags showed that at constant expression levels of BRCA1, the amount of the complex that can be purified from the soluble fraction obtained is a function of the interaction between BRCA1 and BARD1. Using in vitro analysis to complement the in vivo results, we were able to decisively prove that this interface is indeed quite robust to mutations.