The term ‘conditional peripheral membrane proteins’(CPMPs) was coined by Mark Lemmon, who used it to describe a class of globular domains that associate with cellular membranes reversibly in response to specific ligands and/or membrane features. 1 CPMPs occur as individual modules in larger host proteins. These proteins play critical roles in signal transduction and membrane trafficking by participating in extracellular stimulus response, kinase regulation at the membrane surfaces, and modulation of membrane curvature. 2, 3 At least 11 types of CPMPs have been identified, based on their three-dimensional fold and ligand preferences (reviewed in Ref. 4). Of those, 7 are highly ligand-specific and have welldefined binding pockets for membrane-associated activators, such as different types of phosphoinositides (PH, PX, FYVE, ENTH, and tubby), diacylglycerol (DAG)(C1), and phosphatidylserine (C2). Other CPMPs are more promiscuous and respond to general membrane features such as negative charge (KA15) and membrane curvature (BAR6). CPMPs are capable of coincidence detection or response to more than one type of spatially proximal lipid ligands. This can be accomplished through having distinct binding sites within one domain and/or combining several domain types within a host protein, thereby significantly expanding the possibilities for the fine-tuning of the signaling response in the cell. Research efforts toward highthroughput characterization of protein–lipid interactions7 and the discovery of new CPMPs8 are ongoing.

Given the pivotal role of CPMPs in the regulatory events at the membrane surface, there is a need to understand the molecular mechanism of their action. Ideally, this requires atomic-level information about CPMPs in two phases: aqueous and membrane associated. In addition, the neighboring domains and flexible linker regions can have significant influence on the accessibility of lipid-binding sites and geometry of CPMP–membrane interactions. In a broader context, the structural and dynamical information about the ligand-binding and allosteric sites within CPMPs will inform the design of selective modulators of the host protein function. From the structural biology perspective, the two properties of CPMPs that tailor them to the membrane-targeting modular function–amphiphilicity and the ability to fold independently into a fully functional form–present both remarkable opportunities and challenges. The ability to express and isolate some CPMPs in soluble form with high yield enables one to obtain atomic-level structural information using either X-ray crystallography or NMR. This ‘divide-and-conquer’approach is free of the problems that would otherwise be encountered in the full-length host proteins: large molecular mass that results in unfavorable rotational diffusion properties and poor spectral resolution in solution NMR spectra; and flexible linker regions that impede crystallization efforts. The hydrophobic character of CPMPs fully reveals itself upon their interactions with membrane targets. Having suitable membrane mimics in place is essential, because even interactions with a short-chain soluble ligand analog may result in a sharp decrease in solubility. Cocrystallization of