Our research centers on the interactions between transmembrane helices and the different factors that can modify these interactions. Such interactions are at the center of the structure, assembly dynamics and function of helical transmembrane proteins. Certain of our projects will we hope however have a more wider applicability than helical membrane proteins extending to b-barrel membrane proteins such as OmpF and other non trans-membrane integral membrane proteins such as prostaglandin synthase. Our interest in helical membrane proteins is prompted by their structural simplicity, which we beleive should make the prediction and understanding of structure more simple than for soluble proteins, and the lack of physico-chemical understanding of their assembly and dynamics which severly restricts such predicitive efforts. Our current projects are all organised round a series of thematics all aiming to increase our understanding of the biochemistry of membrane proteins.
The prediction of membrane protein structure is conceptually simpler than the corresponding process for soluble proteins and the process can be separated into a series of distict steps: membrane insertion, packing of transmembrane helices, and constrained folding of soluble loops. Furthermore, current methods allow the accurate prediction of transmembrane helices, a prediction which corresponds to the first step. However at present there is very little information on how to predict the packing of transmembrane helices and what are the important parameteres that govern this step in the assembly of a membrane protein. In view of the increasing number of membrane proteins of known structure we have started a programme to use these structures to understand and predict transmembrane helix packing, we hope to be able to test any predictions we make using the various expermental systems at our disposal. As a first step we have developped a method to allow the analysis of a-helix structure, relative to the local helix axis, this method allows us to develop helical projections to better visulaise and understand the properties of the helical surface it also provides a tool for analysing the distorsions found in transmembrane helices. The first, visualisation, aspect has given rise to a graphical programme pTUNA (Arce-Loopera et al. In preparation) which we use to visualise helical surfaces and map properties to these surfaces. The programme is now being expanded to analyse the known transmembrane helical protein structures in terms of helix distorsions and helix-helix interactions.
For soluble proteins the unfavourable interaction of the unfolded polypeptide-chain with the aqueous environment gives rise to the hydrophobic effect which is of paramount importance in the folding of these proteins. However for membrane proteins the importance of the hydrophiobic effect is for membrane integration rather than folding. We have therefor undertaken a series of investigations to try and understand the role of the amphiphilic environment in the stabilisation of membrane proteins.
Historically this investigation has grown out of an early article in which we examined the effects of octyl-glucoside on the assembly of the Rhodospirillum rubrum light-harvesting antenna complex (Sturgis and Robert, 1994). We extended these measurements to the transmembrane helix of the protein Glycophrin a, in collaboration with Professor Engelman's Laboratory we have developed and validated with a fluorescence method for measuring the equilibrium constant for this dimerisation under a wide variety of biologically relevant conditions (Fisher et al. 1999).
Using this assay we have investigated the effects of diverse detergents on the stability of this homodimer. More recently, we have extended this study by investigating the effects of many detergents at widely varying concentrations (Fisher et al. Submitted). This latter study has allowed us to propose a simple thermodynamic model (see right) that describes the behaviour of the Glycophorin a dimerisation on the concentration and nature of the detergent. In this model we consider the solvent as a two phase system (aqueous and micellar phases) with the protein soluble in both phases. We have recently verified one of the predictions of this model which is peptide solubility and the posibility of measuring dimerisation equilibria at low detergent concentrations (Duneau et al. in preparation). We are currently extending this study to a series of different proteins, (KcsA, HER-2 and TolQ) to test the generality of our conclusions obtained with Glycophorin a in these different systems. We are also extending our investigations to examine the role of solvent dimensionality and membrane thickness using the methods we have developped in bicelle and lipid-bilayer systems.
Using a variety of different model systems we are investigating the dynamics and assembly of membrane protein complexes composed of multiple subunits.
The purple bacterial light harvesting complexes function to collect the light energy used for photosynthesis with remarkable efficiency. A large number of such complexes have been isolated and characterised from a variety of different bacteria. They are localised within a specialised photosynthetic membrane system, and in a number of cases are known to form para-crystaline arrays in vivo.
These light-harvesting complexes are composed of cyclic oligomers of a basic subunit (an example is shown here). This basic subunit contains two polypeptides (alpha and beta) and a number of pigment molecules that varies depending on the source. The cyclic oligomers are composed of between 8 and 16 copies of this basic subunit depending on the source.
In certain cases it has proved possible to reversibly dis-assemble and re-assemble these complexes in a stepwise manner. Thus strarting with isolated pigments and polypeptides it is possible to assemble forms containing monomeric polypeptide bound pigments, a heterodimeric form containing the alpha-beta hetero-dimer with its associated pigments, the native ring structure and subsequently form two dimensional crystals of these rings, all in a controlled and reversible manner. Furthermore these different steps can be followed spectroscopically.
We are currently investigating a number of different stages in the assembly of these complexes. In an effort to understand the various molecular factors responsible for the different steps in the heirachical assembly process. In particular we are examining the factors that control the selectivity and specificity of the process.
We have investigated a number of aspects of pigment binding - concentrating in particular on quantifying the effects of different molecular determinants on the binding affinity. At the level of the formation of hetero-dimers we are again particularly interested in the role played by the pigment in catalysing and regulating this process.
We are also actively investigating the oligomerisation that leads to ring
formation. At this level of organisation we are particularly intrigued by
the regulation of the size of the resulting ring and the control or ring
closure. Finally at the level of aggregation. and crystalisation we are interested
in describing the environemental factors that determine the solubility limit
and thus aggregation or crystalisation of these proteins.
The Tol complex is a multi-subunit protein complex that is associated with the stability and functioning of the E. coli outer membrane. This complex is composed of a number of different proteins associated variously with the cytoplasmic membrane the periplasm and the outer bacterial membrane. We are particularly interested in the interactions between the different components associated with the cytoplasmic membrane, namely TolQ, TolR and TolA. Unlike the other systems studied the Tol complex includes a polytopic membrane protein (TolQ).
An investigation of the phylogenetic distribution of this complex (Sturgis 2000) has already demonstrated the importance of this system throughout the Gram negative bacteria, while suggesting an important role in the maintenance of outer membrane integrity - particularly in free-living bacteria. On the basis of sequence comparisons we have also suggested that the members of this complex associated with the internal membrane (TolQ, R and A) form an energy coupling proton pore, which derive energy from the proton motif force and deliver this as mechanical energy to the periplasmic components of the complex. In this article we have also proposed a structural and functional model for the inner membrane associated part of this complex (Cascales et al. 2000, 2001).
Our research on this system is aimed at the study the interactions, assembly and dynamics of this integral membrane protein complexe, and is being performed in collaboration with the group of R. Lloubes. We are currently developing methods for the rapid localisation and purification of the different polypeptides of this complex using a combination of molecular biological and biochemical techniques.
We are studying mainly colicin A, a group A colicin, and Cal, the colicin A lysis protein, both proteins are produced by strains ofE. coli carrying the plasmid pColA. Colicin A is a protein synthesised and secreted as a globular soluble protein, yet it is able to insert into biological and artifical membranes to form a voltage dependant pore. Cal is a lipoprotein which is acylated within the bacterial inner membrane before being transfered to the E. coli outer membrane. The synthesis, mode of action, purification and interactions of both colicin A and Cal are currently being studied.
One of the challenges in understanding cellular physiology is transfering measurements made in-vitro to the intra-cellular environment. We have therefor initiated a project to allow the measurement of protein-protein interactions in vivo. Importantly our objective is not only to be able to measure such interactions but to be able to quantify them, determine a dissociation constant. Furthermore we would like to be able to make these measurements on membrane proteins with the added desire for a good time resolution. The general approach that we have adopted, in common with many other labs, is to use fluorescent markers, in particular autofluorescent proteins. In an initial study we have developped a single hybrid measurement based on fluorescence anisotropy that has allowed us, in collaboration with the group of Alain Filloux, to measure and examine the interactions among the different proteins involved in quorum sensing by the opportunistic pathogen Pseudomonas aeruginosa (Ventre et al. 2003, Ledgham et al. 2003).
We have now extendid this work to a 2-hybrid system which allows us to measure the interactions between different proteins by FRET and so measure, from the concentration dependence, a dissociation constant. The method has an additional advantage that we can use the same constructions for measuring the dissociation constant in-vitro and in-vivo and thus examine the role of the intracellular environment in modulating the affinity of proteins (Prima et al. in preparation).
Recently, in collaboration with the Physicists at the Institut Fresnel we have started to investigate the dynamics of these fluorescent proteins in the intracellular environment using Fluorescence Correlation Spectroscopy.
James Sturgis, May 2003