Proteolytic antibodies are a novel therapeutic modality that has the potential to revolutionize antibody-based therapeutics. The overall goal of this project is to develop potent proteolytic human monoclonal antibodies (mAb) against a new target for treatment of psoriasis. Several recent studies demonstrate a key role for the keratinocyte growth factor amphiregulin (AR) in the pathogenesis of psoriasis. AR-neutralizing antibodies given at high doses alleviate some of the characteristic lesions in human psoriatic skin transplanted onto immunocompromised mice. However, grams of passively administered antibodies will be required for chronic therapy, a serious limitation of the traditional monoclonal antibody approach. We hypothesize that a proteolytic anti-AR antibody can provide comparable benefit at a dose 3-4 logs below that of conventional antibodies. Proteolytic antibodies have been observed naturally, and certain disease states provoke the formation of antibodies with physiologically significant proteolytic activities. For example, Factor VIII hydrolysis by proteolytic antibodies accounts for Factor VIII resistance in many hemophiliacs. Such effects suggest the potential efficacy of proteolytic antibodies in the clinic. It has been estimated that a proteolytic antibody with a modest turnover rate of 0.1/min could inactivate over 2000 times much antigen as an antibody that binds stoichiometrically. A third of the human antibody kappa light chain variable domain (V() germline repertoire contain canonical serine protease-like catalytic triads in their complementarity-determining regions (CDR), and some have detectable peptidolytic activity. The tools of antibody engineering allow the construction of large libraries based on these human V: structures, which may contain effective proteolytic activities against a broad spectrum of therapeutic targets. What is lacking is a robust system for direct selection of desired activities from such libraries. To address this need, we have developed a proprietary cell-based system for the selection of target-specific proteolytic activities with potentially unprecedented catalytic efficiencies from large libraries of antibody Fab fragments containing proteolytic V( domains. The selection system comprises a proteolytic activity sensor in which an auto-inhibited reporter is linked to the target in such a way that cleavage of the target leads to activation of the reporter. When the sensor is expressed in E. coli cells along with a proteolytic V(-containing antibody library, cells expressing antibodies with target-cleaving V( domains can be identified by the selectable phenotype conferred on the cells by the activated reporter. The system will be used to select and optimize human antibody Fab fragments containing V( domains with high proteolytic activity against human amphiregulin. The most active proteolytic Fab(s) will be characterized in vitro with respect to enzyme kinetics, specificity, and the ability to neutralize AR bioactivity on cultured cells. In Phase II, the most promising molecules will be subjected to pre-clinical evaluation in animals, pursuant to submission of an IND application for clinical trials.
Monoclonal antibodies have shown great promise for the treatment of psoriasis, but frequent administration of prohibitively large doses would be required. Proteolytic antibodies are a novel therapeutic modality that has the potential to provide a comparable or better benefit with much smaller doses. In this project, a system which has been developed for engineering proteolytic activity into antibodies will be applied to developing proteolytic antibodies against amphiregulin, a promising new target for the treatment of psoriasis.
