Site-Directed Drug Discovery for Challenging Targets

Site-directed drug discovery for challenging targets is one focus of our research.

The challenge of protein biology

Much of biology is regulated by protein-protein interactions (PPIs), or protein allostery (PA). PPIs have been deemed undruggable since they use much larger flat interfaces than traditional enzyme active sites. PA is also challenging to manipulate with drugs because of difficulties in both identifying sites and studying their mechanisms.

Early successes

While in the Protein Engineering Department at Genentech (1982–1998), we showed that PPIs have hot-spots, where only a small subset of residues drive the interaction. We further showed that these sites can flex to adapt to mutations and to engage many other binding partners, including naïve peptides selected by phage display.

At Sunesis Pharmaceuticals, Inc. (1998–2005), we developed a site-directed fragment discovery approach called Tethering^®, or disulfide trapping. Here, a synthetic library of disulfide-containing fragments are allowed to undergo reversible thiol-disulfide exchange in the presence of a redox buffer with engineered or existing cysteines such that only those fragments with some non-covalent affinity are trapped at binding sites. This approach has advantages over non-covalent fragment approaches in that 1:1 labeling efficiency can be readily obtained to saturate the site for structural and functional analysis.

With the lab groups of Michelle Arkin, PhD (UCSF); Andrew Braisted, PhD; and Brian Raimundo, PhD, we used Tethering to discover and structurally define the first high affinity molecules (Kd = 60nM) that bound a PPI, that of the alpha-receptor binding site in IL-2. Over 25 different drug targets, as diverse as the phosphatase PTP1B and the C5a receptor a GPCR, were subjected to Tethering at Sunesis, many in collaboration with pharmaceutical partners Merck, Janssen, and Biogen.

Advancing drug discovery at UCSF

This work continues at UCSF. In collaboration with Adam Renslo, PhD (UCSF), a new small molecule library generated by single pot-synthesis was constructed and has been used extensively in both our lab and Arkin Lab (UCSF) as well as those of our many collaborators mentioned below. For example, we applied Tethering to define a general allosteric network that operates the zymogen activation mechanism for caspases-1, -3, -5, and -7. This inspired small molecule screening to identify caspase activators that spontaneously self-assemble into remarkable nano-fibrils, which induce cell death. This also motivated the selection of conformationally selective antibodies that activate or inhibit caspase-1, in collaboration with Sachdev Sidhu, PhD (University of Toronto), and that activate the zymogen of caspase-3.

More recently, in collaboration with Matt Jacobson, PhD (UCSF), we have used Tethering to identify an allosteric site in the master kinase, PDK-1, and show how some compounds can activate or inhibit the kinase by pushing on the famed C-helix in opposite directions. We further identified and structurally defined non-covalent and cell active compounds by fluorescent polarization in collaboration with Arkin, and by virtual screening in collaboration with Brian Shoichet, PhD (UCSF).

In additional to small molecule research we have used mutagenesis to define the SAR for thiocillins, a class of ribosomally encoded natural products, and found potency was extremely dependent of ridged ring entropy in collaboration with Chris Walsh, PhD (Stanford University) and Jacobson.

Cryptic and allosteric sites in proteins are poorly annotated and offer great alternatives to actives sites. There are a number of other Tethering projects in this area driven by collaborators, including Anna Mapp, PhD (University of Michigan), on the transcription factor Kix, and with colleague Kevan Shokat, PhD (UCSF), on RAS. We are excited to pursue collaborative projects like these.

Tethering is a registered trademark of Sunesis Pharmaceuticals, Inc.