Introduction: Membrane proteins are notoriously more difficult to work with than soluble proteins, yet there is a huge interest in determining structures from samples of membrane proteins. Traditionally, membrane protein crystals were grown from an aqueous mixture of protein and detergent in a manner similar to the crystallisation of soluble proteins.
The species that crystallises is not just the protein, but it is the protein/detergent complex. One of the confounding aspects of membrane protein crystallisation is to find the right detergent for crystallisation. This may not be the same as the detergent most appropriate for the extraction and purification steps. Crystals grown in this way tend to have crystal contacts formed through the extra-membrane parts of the protein. These are called type II crystals:
Type I and II crystals from a review by Eric Gouaux in the Journal Structure (1998)
In 1996, Landau and Rosenbusch published a paper in PNAS which suggested the use of Lipidic cubic phases (LCP) for the crystallisation of membrane proteins. In this, a lipid/water system is used which spontaneously forms a three-dimensional continuous bilayer. Protein molecules sit in the bilayer of this meso-phase and can move through the bilayer and contribute to the growth of crystals (figure 1b). Crystals grown from the cubic phase tend to have close contacts through the membrane spanning region, they look rather like stacks of 2-D crystals. These are called type I crystals, and in general diffract more robustly than type II crystals.
Left: Phase diagram of Monoolein/water (from Caffrey and Cherezov Nature Protocols, (2009) Right: Monoolein – or MAG9.9 This is an 18 carbon fatty acid attached to a glycerol molecule, with a single (cis) double bond between carbons 9 and 10 of the fatty acid.
The cubic phase was not widely adopted in the early days, as forming the phase was very time consuming. Furthermore, handling (e.g. dispensing) of the cubic phase is challenging, as it is a very viscous gel – rather similar to toothpaste in consistency. However, the development of the syringe mixing method (1998), the development of robust commercial automation for dispensing the phase and the slew of very high profile structure papers that have used this method have made this method much more popular.
In this method, solublised membrane protein solution is used as the aqueous phase in the Lipid/water mixture to create a mesophase with proteins incorporated into the 3D-lipid bilayer. Generally a monoacylglycerol (MAG) is mixed with an aqueous phase in a ratio that will generate the desired phase. The phase that results from mixing the two components depends on the ratio of the two and the temperature. The phase diagram is shown above for the most commonly used MAG in protein crystallisation (MAG 9.9, or mono-olein).
NEXT: Preparing the LCP Mix
RELATED: Booking an LCP Experiment