Membrane Protein Crystallisation: Current Trends and Future Perspectives
Alpha helical membrane proteins are the targets for many pharmaceutical drugs and play important roles in physiology and disease processes. In recent years, substantial progress has been made in determining their atomic structure using X-ray crystallography. However, a major bottleneck still remains; the identification of conditions that give crystals that are suitable for structure determination. Over the past 10 years we have been analysing the crystallisation conditions reported for alpha helical membrane proteins with the aim to facilitate a rational approach to the design and implementation of successful crystallisation screens. The result has been the development of MemGold, MemGold2 and the additive screen MemAdvantage. The associated analysis, summarised and updated in this chapter, has revealed a number of surprisingly successfully strategies for crystallisation and detergent selection.
KeywordsMembrane protein Crystallisation Screen development Detergent selection
Protein crystallisation is often described as a ‘black box’ process, full of mystery and superstition. In fact crystallisation itself is a well documented process following well understood physical chemistry laws and involving the supersaturation of the protein of interest to coax the molecules into a regular three-dimensional crystal (McPherson and Gavira 2014; Chayen and Saridakis 2008). Although not the topic of this chapter, many excellent sources of information are available on how to set up these conditions, using either vapor diffusion (Delmar et al. 2015), microbatch (Chayen 1998) or free-interface diffusion (Segelke 2005). The mystery begins when we try to consider which conditions will coax the proteins to assemble in a regular form and produce diffraction quality crystals. For many years, the standard experimental set up has involved screening your purified, homogenous protein sample against commercial, sparse-matrix style ‘crystallisation screens’ (Luft et al. 2014). The idea behind these screens was to sample as much ‘crystallisation space’ as possible with the minimal protein amount (Newman et al. 2007). Of course, many of these commercial screens were based on the currently available information regarding soluble protein crystallisation. This included the success of large molecular weight (MW) polymers, particularly polyethylene glycol (PEG). High concentrations of salt were also used, as these would naturally help to ‘salt out’ the protein and hopefully grow protein crystals (Page and Stevens 2004).
Membrane proteins however are different. The requirement to extract these proteins from the membrane using detergents, whilst simultaneously keeping them folded and stable in solution creates a new set of unknown variables (Iwata 2003). Added to this the fact that in a crystallisation experiment involving detergent solubilised membrane proteins, the actual entity being crystallised is the detergent-protein complex and not simply the protein alone (Kunji et al. 2008; Bill et al. 2011). It was against this backdrop that in 2008 a comprehensive analysis of membrane protein crystallization conditions was published (Newstead et al. 2008a). The idea behind this analysis was simple, the number of crystal structures had just reached 121 and our aim was to analyse these conditions and draw conclusions as to which chemicals were successful in growing membrane protein crystals. Could any trends be observed and could this information be used to improve success rates in current projects? The result of this analysis was the release of the first rationally designed sparse matrix style membrane protein crystallisation screen, MemGold (Newstead et al. 2008a).
In the following years the pace of membrane protein structure determination has increased exponentially (White 2007). This increase is due to progress being made in tackling many of the hurdles faced in determining the crystal structure of membrane proteins (Bill et al. 2011; Ghosh et al. 2015). This includes advances in protein production using recombinant systems (Tate et al. 2003; Chen et al. 2013), methods for screening stability (Drew et al. 2008; Kawate and Gouaux 2006; Sonoda et al. 2011) and in X-ray data collection using microfocus beamlines, fast read out detectors and modifications to sample application (Nogly et al. 2005). More recent progress has been made in protein engineering, resulting in either increases in protein stability (Tate and Schertler 2009) or the introduction of additional crystallisation scaffolds, such as T4 lysozyme or BRIL (Chun et al. 2012). However, growing well-ordered three-dimensional crystals still represents a significant hurdle. In 2012, we followed up our first analysis with another review of the current trends in crystallisation, this time based on 254 examples from the Protein Data Bank (PDB) (Parker and Newstead 2012). Our results showed that the initial trends described in 2008 had broadly held, but revealed intriguing new developments such as an increase in the number of cases where additional or mixed detergents had been required and changes in the types of membrane protein being crystallised. The new information enabled the development a sister crystallization screen, MemGold 2, to complement the original MemGold screen released 4 years earlier. In addition to our analysis of crystallisation conditions, an in depth analysis of additives was now possible. The use of additional chemicals to optimise initial crystals to improve diffraction quality is well documented and many commercial kits are available (Chayen and Saridakis 2008; Cudney et al. 1994). An additive screen targeted specifically for membrane proteins however, had so far remained absent from the commercial market. A specific membrane protein additive screen was therefore suggested to facilitate crystal optimization and released along with MemGold 2, called MemAdvantage.
As of August 2015, the number of crystallisation examples in our database is more than 500 and in this chapter we present an updated analysis from these conditions. Here we compare the results of these past analyses with each other and with those focused on soluble proteins (Fazio et al. 2015). The aim of this chapter is to equip the protein crystallographer with the knowledge to design their own screens using information that is up to date and relevant to membrane protein samples.
5.2 Current Trends in the Number and Types of Alpha Helical Membrane Protein Structures
5.3 Detergent Selection
An important development in membrane protein crystallisation over the past 3 years has been the increased use of the lipidic cubic phase (LCP) as a medium for crystal growth (Caffrey and Porter 2010; Caffrey 2009). This technological development has had an enormous impact on the GPCR field and is one of the main reasons for the increase in the number of structures from this group in the past few years (Ghosh et al. 2015). This methodology is sure to increase in use in the coming years. To date we have recorded 17 structures out of a total of 91, compared with 49 GPCR examples. As highlighted in Fig. 5.3, the mean resolution for structures determined in LCP is 2.5 Å, almost half an ångström lower than for the alkyl maltopyranoside detergents and very close to the mean resolution obtained for n-octyl-glucopyranoside (OG), which in many cases is too harsh for alpha helical membrane proteins. This data adds further support to the early adoption of lipidic mesophase crystallisation in any structure project. More information on detergents can be found in Chap. 2 of this book.
5.4 Precipitants – How Do They Differ Between Membrane Proteins and Their Soluble Counterparts?
Our 2008 analysis of precipitants revealed a striking success for small MW PEGs in the crystallisation of channels and transporters, with larger MW PEGs being more successful for respiratory complexes and membrane proteins with large hydrophilic domains (Newstead et al. 2008a). These trends have remained in the updated data set, with the notable appearance of small MW PEGs in the crystallization of the eukaryotic GPCR family. The successful concentration ranges have also been maintained, with small MW PEGs being successful at concentrations between 20 and 40 % v/v, and larger MW PEGs being used at lower concentrations, between 5 and 20 %. The successful use of organic molecules, such as MPD is still low, further confirming their unsuitability in general crystallisation conditions for alpha helical MPs, a situation that is dramatically different for outer membrane proteins where organic molecules are clearly more successful (Newstead et al. 2008b). Of note is the absence of high salt conditions in our database. This contrasts with a recent analysis of crystallisation space reported for the entire PDB in 2014 (Fazio et al. 2015). This analysis clearly demonstrates the most successful crystallization reagents are PEG 3350 and ammonium sulphate, which only make up 4.0 and 3.5 % of our database, respectively. This contrasts with PEG 400, which accounts for 33 % of the reported membrane protein conditions, but doesn’t appear in the ten most abundant chemicals reported in a non-redundant analysis of successful crystallisation conditions.
5.5 MemGold and MemGold2 – What’s the Difference?
5.6 Buffers, pH and Salts
We have also observed a significant increase in the number of different polyvalent cations and anions reported. Therefore MemGold2 contains a different set of these chemicals, which can be essential to enable proteins to interact and pack into a crystal (PepTSt) (Solcan et al. 2012). It is interesting to note that one of the most successful commercial crystallisation screens is the Hampton PEG/Ion screen, which also involves screening many different polyvalent and monovalent salts against the most successful precipitant for soluble proteins, PEG 3350 (Fazio et al. 2015). This is possibly something that should be replicated for membrane proteins.
5.7 MemAdvantage – An Alpha Helical Membrane Protein Additive Screen
5.8 MemMeso – A Systematically Designed in meso Crystallisation Screen
Our database contains 91 unique examples of membrane proteins crystallised using the in mesoLCP method. The crystallisation conditions from these examples have been analysed (Fig. 5.7c). As expected, the conditions are dominated by PEG 400. However, this result is heavily biased by the GPCR examples. Interestingly we observe a number of conditions using small organic molecules, which we had previously observed were largely unsuccessful for vapour diffusion crystallisation of membrane proteins (Newstead et al. 2008a). For example the Ca2+/H+ antiporter (PDB: 4KPP) was crystallised using pentaerythritol propoxylate and sensory rhodopsin I (PDB: 1XIO) was crystallised using MPD. The remaining examples in the organics are Jeffamine-M600, which is similar in chemical composition to polyethylene glycol. Interestingly we also observe a significant number of high salt conditions, contributed by Bacteriorhodopsin, Halorhodosin and sensory rhodopsin II. Although the number of examples in our analysis are small, it suggests that crystallisation in the lipidic cubic phase may be influenced differently to that in solution.
Membrane proteins represent important pharmaceutical targets and interesting subjects of study with respect to cellular biology and protein biochemistry. However, they still represent challenging targets to crystallise and study. To date our database of 569 unique structures compares to > 110,000 structures in the entire PDB, representing < 1 % of known crystal structures. The field of membrane protein structural biology is still developing at a rapid pace. The introduction of serial injection systems for crystals at synchrotron radiation and free electron sources (Conrad et al. 2015) and the development of in situ diffraction data collection methodology (Huang et al. 2015) suggest that what structural biologists need from a crystallisation experiment is likely to change in the coming years. The final chapter on the topic of crystal screen design and optimization is far from being written. As more information is gathered it seems likely that new trends will be discovered and new crystallisation methods invented or traditional methods refined to meet the growing need to understand these important and fascinating proteins at atomic resolution. The information contained in this chapter represents the current snapshot of ‘crystallisation space’ for alpha helical membrane proteins. It is our wish that this information will encourage the efficient use of the MemGold family of screens but also enable the design of more tailored crystallisation screens for particular projects of interest to you.
This works was funded through the Wellcome Trust Investigator Award 102890/Z/13/Z to SN.
- Caffrey M, Porter C (2010) Crystallizing membrane proteins for structure determination using lipidic mesophases. J Vis Exp 45Google Scholar
- Iwata S (ed) (2003) Methods and results in crystallization of membrane proteins, vol 4. Internat’l University. Line La Jolla, USAGoogle Scholar
- White S (2007) Membrane proteins of known 3D structure. http://blanco.biomol.uci.edu/mpstruc/
Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.
The images or other third party material in this chapter are included in the chapter’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.