Molecular Biology of Protein-Protein Interactions for Computer Scientists

Background: Ppi Types

PPI as any other molecular interaction must meet at least two conditions: sufficient concentration of interactants and a favorable difference of free energy upon interaction. Both parameters are dynamic and depend on the cellular environment. The concentration of a protein may depend on the level of gene expression, translation efficacy, its stability or turnover, and availability in the same cellular compartment as the potential interacting partner. The level of free of energy difference depends on the sum free energy differences derived from atomic level interactions such as hydrogen bond formation, van der Waals interactions and ionic bond formation. Corresponding to the difference of total free energy the resulting interactions are broadly classified into stable, strong transient, or weak transient interactions. Stable interactions are often formed by homodimers, for example RNA polymerase II. Many heterodimeric interactions, for instance the binding of cell division cycle 42 to Wiskott-Aldrich syndrome protein (Rudolph et al., 1998) are transient, but relatively strong. Interactions of cell surface receptors or cargo proteins are weak and transient (e.g., carbohydrate mediated binding of lectin mannose binding 1 to cathepsin C (Nyfeler et al., 2005)).

The detection of PPIs depends on the intrinsic properties and sensitivities of the assays which are reviewed in detail by Berggård (2007) and Lalonde (2008). Briefly, medium throughput surface plasmon resonance assays (e.g., BIACore^TM) are among the most sensitive in vitro assay systems that capture also the kinetics of PPIs. The popular in vivo yeast two-hybrid (Y2H) assay is high throughput, but not quantitative and biased towards the detection of stronger binary interactions between protein domains.