Biological processes are controlled by protein-protein interactions. The specificity of a protein-protein interaction permits highly selective pharmacological interference with a defined cellular process. Due to the vast diversity of protein-protein interactions they represent a large class of potential drug targets that offer great opportunities for therapeutic intervention. However, targeting of protein-protein interactions is a difficult task (Wells and Clendon, 2007, Nature 450:1001).

Peptides mimicking binding domains disrupt protein-protein interactions with high selectivity by competitively binding to one of the interacting partners. Their use has provided insight into the function of a plethora of protein-protein interactions. However, therapeutic applications of peptides are limited as they need to be administered parentarally, possess a short half-life and may evoke immune responses. Such drawbacks may be overcome through the development of stabilized peptides and peptidomimetics (for review see: Yin and Hamilton, 2005, Angew Chem Int Ed 44:2).

Interfaces for protein-protein interactions do not possess evolutionarily conserved hormone-, substrate-, or cofactor-binding pockets. On the contrary, proteinprotein interactions often rely on extended, relatively flat interaction surfaces (750-1500 Ä2) which cannot easily be blocked by small molecules (Freund et al., this volume). However, the exchange of a few amino acids within a defined part of the contact area ('hot spot') may completely abolish a protein-protein interaction. For example, key residues contribute a large fraction of the free energy of binding between the growth hormone and its cognate receptor (Clackson and Wells, 1995, Science 267:383). Thus, it is not necessary for a small molecule to cover the entire protein-binding surface for blocking a protein-protein interaction. The concept of 'hot spots' as areas crucial for the affinity of the interaction has been confirmed for numerous protein-protein interactions (for review see: Wells and Clendon, 2007, Nature 450:1001; Arkin and Wells, 2004, Nat Rev Drug Disc 3:301; Yin and Hamilton, 2005, Angew Chem Int Ed 44:2). Besides targeting 'hot spots', small molecules may also bind to allosteric sites distant from the interface. The conforma-tional change induced by the binding prevents a protein interaction. In addition, disrupting the homooligomeric assembly of an enzyme complex by the use of a small molecule can result in a highly isotype-specific allosteric inhibition, e.g. of inducible nitric oxide synthetase (iNOS; McMillan et al., 2000, Proc Natl Acad Sci USA 97:1506).

Most if not all diseases are associated with or are caused by dysregulation of signal transduction processes. Generally, signalling proteins, including protein kinases, protein phosphatases and phosphodiesterase, are ubiquitously expressed. However, it is increasingly recognised that subsets of signalling proteins are encompassed in cell type-specific multi protein complexes that are tethered to defined cellular compartments in close proximity to their substrates (e.g. Torgersen et al., Kreienkamp, Freund et al., Costa and Cesarini, McCahill et al., all in this volume). This compartmentalization is mediated by scaffolding proteins including arrestins (Gurevich et al, this volume), A kinase anchoring proteins (AKAPs; Dodge-Kafka et al., and Hundsrucker et al., this volume), receptors of activated C kinases (RACKs; Walker, this volume) and caveolin (Patel et al., this volume). Scaffolding proteins temporally and spatially coordinate signal processing, facilitate the assorted flow of information through a cell and thereby participate in converting an exogenous stimulus into a specific cellular response. Displacing selected proteins from complexes assembled by scaffolding proteins is likely to cause subtle local changes in signal processing that affect defined cell functions.

Disruption of protein kinase A (PKA) anchoring by AKAPs has, for example, been achieved using 17-25 amino acid residues-long peptides derived from the PKA-binding domains of different AKAPs. The PKA anchoring disruptor peptides prevent the interaction by binding to the AKAP interaction sites on regulatory subu-nits of PKA (Hundsrucker et al., this volume). In cardiac myocytes AKAP18a tethers PKA to L-type Ca2+ channels by direct interaction of the AKAP with the channel. The PKA anchoring disruptor peptides abolish P-adrenoreceptor-induced, AKAP18a-dependent PKA phosphorylation of L-type Ca2+ channels and the consequent increase of channel open probability by displacing PKA from AKAP18a. This in turn prevents the enhanced entry of Ca2+ that increases contractility (Hulme et al., Proc. Natl. Acad. Sci. USA 100, 13093, 2003; Proc. Natl. Acad. Sci. USA 103, 16574, 2006). This effect resembles that of P-blockers (Dodge-Kafka et al. and Chudasama et al., this volume). Also similar to the effect of P-blockers is that of a peptide disrupting the direct interaction of AKAP185 and phospholamban (PLN) in cardiac myocytes. This peptide, derived from the PLN interaction site for AKAP185, reduces the velocity of Ca2+ reuptake into the sarcoplasmic reticulum (Lygren et al., EMBO Rep. 8, 1061, 2007; Hundsrucker et al., this volume).

This book discusses therapeutically relevant protein-protein interactions with a major focus on scaffolding proteins tethering signal transduction processes to defined cellular compartments by direct protein-protein interactions. Recent advances in the development of peptides and small molecules as pharmacological agents interfering specifically with defined protein-protein interactions are reviewed and potential therapeutic applications of the agents are highlighted.

Enno Klussmann and Walter Rosenthal

0 0

Post a comment