Post-translational inhibition or inactivation or removal of a selected intracellular protein offers an alternative to the use of mutated cells or antisense constructs or cells from knockout animals. Moreover, if an intracellular targeted protein can be deactivated for a given time and subsequently allowed to reactivate such that the cell reverts to its normal state, then the specific function of a protein can be elegantly probed. Within the current project, a strategy was devised to inactivate specific targets within a cell by creating libraries of binding proteins through recombinant randomization of the hypervariable CD3 domain of a heavy chain antibody gene. Once the initial randomized gene library was engineered, then foreshortened by eliminating leader sequence and Fc domain, the remaining nucleotide sequence which represents a peptidal sequence that is monocistronically expressed was provided with a start codon and necessary transcriptional cues. The created sequence was fused to the P8 gene of M13 for display and discovery. The phage array was used to search for binding proteins to an arbitrarily defined target, G6PD. The best binding protein was determined, its corresponding gene was placed into an induceable mammalian expression shuttle vector and the functionality of the construct tested. After inducing expression of the binding protein intracellular G6PD was diminished and ultimately not detectable. Likewise, the reversibility of the process was proved by removal of the inducing stimulus. Interestingly, G6PD activity loss was not a consequence of simple binding to G6PD and subsequent inhibition, rather the binding of G6PD marked it for proteosomal destruction. We concluded that the intracellular deactivation strategy is feasible but much too demanding for ready translation into general use. The design strategy was changed. Namely, P3 and not P8 fusion protein constructs are used to lessen non-specific selection, and the antigen-binding protein avidity/association assay is far too time consuming and also imposes the restriction that quantities of pure antigen must be isolated and fixed to a solid matrix. The assay step was therefore eliminated and a more efficient means of antigen exposure and binding protein affinity testing was devised. The result of which is that when the binding protein (attached through P3 fusion to the M13 virus) affixes to the antigen/P3 fusion, the phage displaying the binding protein is gains entry into a bacterium. Since the phage that enters into the bacterium is unique and carries a antibiotic selection marker then the approach provides in one step both selection and amplification. The new strategy eliminates isolation of antigen, speeds the search for effective binding, and provides a better template for discovery. We completed the antigen/P3 fusion cassette such that the nucleotide sequence corresponding to the first 60 amino acides from the N-terminus is fused to the bacterial docking segment (135 aa) of P3. Likewise, we have all but completed the M13-P3/binding protein cassette such that the phage docking portion of P3 is fused to a binding protein library. The P3/binding protein library is constructed and being ligated into a P3 deleted M13 cassette vector. The antigen/P3 fusion consists of 180 of the 5 nucleotides of the P70 Ku complex fused to 405 of the 5 nucleotides of P3. We have engineered and proved the functionality of the mammalian shuttle vector such that the binding protein, once expressed, will be directed to the cytoplasm or, as in the case of the Ku binding protein, to the nucleus of the cell: the site of Ku activity (this was done by placing a nuclear traffiking sequence within the cassette and proved by expressing green fluourescent protein from the vector and noting exclusive localization into the nucleus of a mamallian cell). Ku inactivation in the context of the aftermath of ionizing radiation exposure will be studied. The results of our studies will be compared directly to that of experiments done with K70 null cells that were developed through knockout technology. Once the new, more efficient and ?user friendly? method that is now being constructed has been completed, the binding library should have immediate application to numerous biologically relevant questions.
