Protein-Protein Interactions

Over the last two decades, a number of protein-protein interactions (PPIs) have been targeted by the pharmaceutical industry. Pharma as a whole has historically considered PPIs to be undruggable or at the very least high-risk targets, and the relative lack of success in modulating PPIs with small molecules has done little to change this prevailing view. However, many compounds are now in clinical trials, and the experiences of the last 20 years have at the very least led to improved understanding of how to approach these challenging targets. This chapter discusses some of the issues that PPIs present as targets for small molecule modulation, with emphasis on the structural characteristics of PPIs in general, and also of classes of PPIs that share specific attributes. Grouping PPIs by structural class produces a clearer picture of both the characteristics of optimized small molecules, and the relative merits and drawbacks of various PPIs as drug targets. Within this framework, much of the past work in the PPI area is summarized through capsule descriptions of efforts directed against individual targets. Some contributors to individual successes and failures, and some insights gained from the many avenues of research followed within the PPI field are put forward. Themes of the importance of understanding the structural basis of mechanism of action and of structural support for drug discovery emerge, and guidelines for future study are offered.

The p53 tumor suppressor protein is a transcriptional factor that plays a key role in regulation of several cellular processes, including the cell cycle, apoptosis, DNA repair, and angiogenesis. The murine double minute 2 (MDM2) protein is the primary cellular inhibitor of p53, functioning through direct interaction with p53. Design of non-peptide, small-molecule inhibitors that block the MDM2-p53 interaction has been sought as an attractive strategy to activate p53 for the treatment of cancer and other human diseases. In recent years, major advances have been made in the design of small-molecule inhibitors of the MDM2-p53 interaction in recent years, and several compounds have moved into advanced preclinical development or clinical trials. In this chapter, we will highlight these advances in the design and development of MDM2 inhibitors, and discuss lessons learned from these efforts.

Apoptosis is a genetically regulated process of cell death that is critical for cellular homeostasis. Dysregulation of apoptosis can lead to the absence of normal cell death and contribute to cancer development and progression. The Inhibitor of Apoptosis (IAP) proteins contain up to three baculovirus IAP repeat (BIR) domains that interact with members of the caspase family of proteases, thereby blocking apoptosis. Select BIR domains of IAP proteins contain a defined protein-protein interaction region with the N-terminal amino acid residues of the natural IAP antagonist protein, SMAC. Antagonizing this interaction formed the basis of all drug discovery efforts and has yielded promising drug candidates. One member of the IAP family, XIAP, can inhibit caspases 3 and 7 through its BIR2 domain and caspase-9 with its BIR3 domain. The c-IAPs play a key role in the regulation of the NF-κB pathways, which is manipulated by the IAP antagonists to yield single agent activity. SMAC mimicking IAP antagonists can be monovalent, representing an equivalent of four amino acid residues from the processed SMAC, or bivalent – having two units connected through a chemical linker. In addition, antagonists specific for a particular IAP protein or a group of IAPs have been reported. So far, six molecules have entered clinical trials with early results showing no dose limiting toxicities and suggesting that IAP proteins can be targeted by small molecules.

Efforts to interfere with four key protein-protein interactions in the HIV-1 lifecycle with the goal of achieving clinically-relevant, orally administered HIV-1 therapies are reviewed. These four targets: the HIV-1 gp120/human CD4 interaction, the HIV-1 gp41 six-helix bundle formation, the human LEDGF/p75-integrase interaction, and HIV-1 protease dimerization each present unique challenges to the discovery of viable small molecule inhibitors. Background information from the literature is provided. A class of inhibitors which target gp120 from which an orally dosed member has been advanced into Phase II clinical studies as well as other small molecule approaches to disrupt the gp120/CD4 interaction are discussed. The unrealized efforts to find a small-molecule inhibitor of gp41 six-helix bundle formation that is suitable for clinical studies are described, including a summary of the work on effective, peptidic inhibitors that lack the properties needed for oral use. An overview of the progress to identify small molecule inhibitors of the LEDGF/HIV-1 p75-integrase interaction and the dimerization of the HIV-1 protease enzyme describes the preclinical compounds of greatest interest and discusses the rationale behind their design/activity.

The assembly of the N-terminus heptad repeats of the respiratory syncytial virus (RSV) F protein into a trimeric complex that associates with the C-terminus heptad repeats to form a six-helix bundle is a critical step in the process of virus-host fusion and represents an intramolecular protein-protein interaction. Screening campaigns using replicating virus assays have identified several structurally distinct but mechanistically similar chemotypes that interfere with RSV fusion by disrupting the function of the F protein six-helix bundle. This chapter summarizes structure-activity relationships and mechanistic insights associated with the most prominent RSV fusion inhibitors and the key issues in the development of potential clinical candidates.

Cellular function depends on highly specific interactions between biomolecules (proteins, RNA, DNA, and carbohydrates). A basic limitation of drug development is the inability of traditional “small-molecule” pharmaceuticals to specifically target large protein interfaces, many of which are desirable drug targets. α-Helices, ubiquitous elements of protein structures, play fundamental roles in many protein-protein interactions. Stable mimics of α-helices that can predictably disrupt these interactions would be invaluable as tools in molecular biology, and as leads in drug discovery. The past decade has seen exciting progress in the molecular design of these protein domain mimetics and their remarkable potential to inhibit challenging interactions. Key challenges in the field include identification of suitable targets and bioavailability of medium-sized molecules, which do not conform to empirical rules followed in traditional drug design. Stabilized α-helices bypass some of the strict limitations that have been placed on drug discovery. When designing potential drug candidates, medicinal chemists often adhere to the Lipinski rules, which stipulate that the molecular mass of a drug should not exceed 500 Da. Recent findings suggest that large synthetic α-helices can traffic into the cell and efficiently compete with cellular protein-protein interactions, contrary to predictions based on the Lipinski rules. Although these molecules have undoubtedly proven their value as probes for decoding biological complexity, the next big question is whether these molecules can become therapeutics. This chapter discusses the properties of protein-protein interactions, emerging rules for identifying protein targets and design criteria guiding construction of helix mimetics.

A case history of the Abbott Oncology Bcl-2/Bcl-xL inhibitors program is presented. The target proteins interact with other members of the Bcl family through surfaces that are very large and hydrophobic even compared to other PPIs that have been targeted by pharma. Resulting inhibitors are correspondingly large and hydrophobic and thus tend to be highly protein bound and possess low oral bioavailability. The Abbott drug discovery effort began with the creation of a soluble, stable version of Bcl-xL, and a solution structure of Bcl-xL bound to a native ligand. Structural support facilitated efforts to find chemical matter, which was accomplished through fragment screening. Structure-based drug design was also employed throughout the project, with the discovery process characterized by separate optimization of two widely separated hydrophobic hot spots. ABT-737, an IV-only compound, was initially selected as a development candidate. Later, efforts to derive an orally bioavailable compound from the same chemical series produced navitoclax (ABT-263), an extremely potent Bcl-2/Bcl-xL inhibitor, currently in Phase II clinical trials for cancer.