Natural products remain a valuable and important source of molecular diversity for biomedical research, yet the frequency of rediscovery of known scaffolds has reduced the efficiency of the drug discovery process. To address this issue, our laboratory is interested in the development of novel screening technologies that provide detailed characterization of the biological attributes of active molecules early in the discovery process, with the aim of discovering compounds with unique biological properties with respect to known drugs. Current screens in this area include:
Natural products have traditionally played an essential role in antibiotic development, with over 80% of FDA-approved antibiotics being of natural product origin. However, the extensive investigation of natural product libraries for novel antibiotics in previous decades means that rates of rediscovery using traditional growth inhibition assays are now very high. The BioMAP screen profiles antibiotic potency across a panel of 15 clinically relevant bacterial pathogens to identify extracts containing antibiotics with unique spectra of activity, while simultaneously eliminating the large number of extracts containing well known antibiotic scaffolds with little development potential.
An estimated 80% of clinical bacterial pathogens have the capacity to form biofilms, meaning they can settle on surfaces to form biofilm colonies that are extremely difficult to eliminate using standard antibiotics. This is a particular concern for medical procedures involving medical implant devices (catheters, stents, joints, pacemakers etc) where the newly introduced device provides an ideal non-native surface for biofilm colonization. Using the image-based microscope in the UCSC Chemical Screening Center we have developed new screening tools for the discovery of biofilm inhibitors against a range of clinical pathogens, and have used these screens in conjunction with our natural products library to develop new classes of biofilm inhibitors.
Although traditional live/dead screening has been a rich source of discovery for anti-cancer drugs, this screening approach provides no information about the mechanism by which compounds exert their effects on cancer cells. To address this, we have employed an image-based profiling method, termed cytological profiling, to directly visualize the effect of test compounds on the development of cancer cells. Using this technology, we can compare the effect of unknown natural products to those of a library of known drugs, and cluster these phenotypes by mode of action to either discovery new molecules that work by defined targets, or to identify compounds that function by unique modes of action
Antibiotic Mode of Action Profiling
In addition to image-based screening using mammalian cells, we have recently demonstrated that it is possible to image bacterial cells, and to use size and shape parameters to predict to mode of action of unknown antibiotics. We have developed a new image analysis software platform, in conjunction with new image-based 384-well screening tools, to profile natural product libraries and directly predict compound modes of action from primary imaging data.