Abstract
AbstractClostridioides difficileis a major cause of secondary disease in hospitals. During infection,C. difficiletoxin B drives disease pathology. Here we use deep learning and Rosetta-based approaches to de novo design small proteins that block the entry of TcdB into cells. These molecules have binding affinities and neutralization IC50’s in the pM range and are compelling candidates for further clinical development. By directly targeting the toxin rather than the pathogen, these molecules have the advantage of immediate cessation of disease and lower selective pressure for escape compared to conventional antibiotics. AsC. difficileinfects the colon, the protease and pH resistance of the designed proteins opens the door to oral delivery of engineered biologics.Significance statementC. difficileinfection (CDI) is a major public health concern with over half a million cases in the United States annually resulting in 30,000 deaths. Current therapies are inadequate and frequently result in cycles of recurrent infection (rCDI). Progress has been made in the development of anti-toxin mAb therapies that can reduce the rate of rCDI, but these remain unaffordable and out of reach for many patients. Using de novo protein design, we developed small protein inhibitors targeting two independent receptor binding sites on the toxin that drives pathology during CDI. These molecules are high affinity, potently neutralizing and stable in simulated intestinal fluid, making them strong candidates for the clinical development of new CDI therapies.
Publisher
Cold Spring Harbor Laboratory