Translation of Hot Loops into Macrocyclic Inhibitors of Protein-Protein Interactions
Macrocyclic natural products include many of our most potent drugs and most useful biological probes, but
there are no reliable design tools for developing macrocycle drugs. The Kritzer lab uses synthesis, biophysics,
and cell biology to discover and apply macrocyclic compounds. Recently, we reported LoopFinder, an
algorithm that analyzes protein-protein interactions (PPIs) and identifies critical “hot loops”. We hypothesize
that hot loops can be translated into inhibitors of their associated PPIs by stabilizing their bioactive
conformation. This hypothesis is supported by preliminary results, but there remains a critical gap in
knowledge: how can we stabilize a desired loop conformation within a macrocycle? In this project, we use
complementary synthetic, computational and biochemical approaches to address this question. In one
approach, we apply a new “diversity-oriented stapling” strategy that prepares and tests large libraries of
macrocycles with varied 3D shapes. In a complementary approach, we are collaborating with world-leading
computational scientists to apply new Rosetta algorithms that search for highly structured macrocycles, and to
refine predictions using explicit-solvent MD. For each PPI inhibitor we discover, we will analyze its solution
structure and quantify its target affinity, selectivity, metabolic stability and cell penetration. This represents an
unprecedented data set for correlating macrocycle sequence, structure, and function. These data will also
provide continual feedback for the synthetic and computational design approaches.
As targets for macrocycle design, we have identified loop-mediated PPIs that control vesicle trafficking
pathways that are involved in cancer and infectious disease. These include: Eps15, which is required for
endocytosis; EHD1, which is involved in receptor recycling; the ESCRT-II complex, which mediates endosomal
sorting; and the COP-I complex, which controls endosome budding from the Golgi. There are no selective
inhibitors for any of these pathways, making our macrocycles immediately useful to biology and medicine.
This project develops innovative computational and synthetic tools that are specific to the development of
macrocycle drugs. We cannot currently design the next cyclosporine, but we anticipate being able to do so
using this a new generation of methods, tools and molecules. This project represents a timely opportunity to
bring macrocycle drug design closer to reality.