How lipids affect the energetics of co-translational alpha helical membrane protein folding

Abstract

Membrane proteins need to fold with precision in order to function correctly, with misfolding potentially leading to disease. The proteins reside within a hydrophobic lipid membrane and must insert into the membrane and fold correctly, generally whilst they are being translated by the ribosome. Favourable and unfavourable free energy contributions are present throughout each stage of insertion and folding. The unfavourable energy cost of transferring peptide bonds into the hydrophobic membrane interior is compensated for by the favourable hydrophobic effect of partitioning a hydrophobic transmembrane alpha-helix into the membrane. Native membranes are composed of many different types of lipids, but how these different lipids influence folding and the associated free energies is not well understood. Altering the lipids in the bilayer is known to affect the probability of transmembrane helix insertion into the membrane, and lipids also affect protein stability and can promote successful folding. This review will summarise the free energy contributions associated with insertion and folding of alpha helical membrane proteins, as well as how lipids can make these processes more or less favourable. We will also discuss the implications of this work for the free energy landscape during the co-translational folding of alpha helical membrane proteins.

Keywords: co-translational folding; lipids; membrane proteins; protein folding.

Figure 1.. Lipid composition determines bilayer properties.

(A) Examples of lipids commonly used in membrane protein folding studies. Lipids such as those with PC headgroups tend to form bilayers, while non-bilayer forming lipids such as DOPE and cardiolipin (CL) adopt hexagonal phases. Single-chain lipids such as Lyso PC tend to form micelles. (B) Lipid composition determines the overall electrostatic and mechanical properties of mixed lipid bilayers. Introducing non-bilayer forming lipids alters the lateral pressure profile of a bilayer.

Figure 2.. Different methods to measure the free energy landscape of folding.

(A) A computational model for the insertion of a hydrophobic peptide into a lipid bilayer [34]. An initially unstructured peptide in water partitions to the bilayer interface, where it folds to adopt an alpha-helical structure and inserts across the bilayer. ΔG values denote standard transfer free energies, with subscripts indicating states involved. Each state is described by two letters, with the first denoting environment (w, water; i, interface; t, transmembrane) and the second denoting the state of the peptide (u, unfolded; f, folded). For co-translational insertion of membrane proteins in vitro and in vivo, TM domains form alpha-helices in the ribosome exit tunnel rather than folding interfacially. (B) Measuring the folding free energy of a full-length membrane protein with equilibrium folding/unfolding. The ΔG of folding (ΔG_f) and unfolding (ΔG_u) are determined by measuring the change in structure upon addition and removal of a denaturant (shown in the red dashed box). Alternatively, a protein can be unfolded in detergent with either SDS or urea, and then refolded into a lipid bilayer.

Figure 3.. The stages involved in co-translational folding.

The steps involved in co-translational TM interfacial headgroup partitioning, bilayer insertion and folding either without (A) or with (B) the translocon. Folding begins while later TMs are being translated and inserted in the bilayer. Some of these stages, and free energies associated with each stage, are the same as those measured using in vitro methods (as illustrated in Figure 2). We use the models in this figure to highlight the similarities and differences in pathways between translocon-assisted and spontaneous TM insertion and membrane protein folding, and to show that interactions between the nascent chain and the lipid bilayer are involved in all stages.

Figure 4.. How bilayer properties affect the free energy landscape of folding.

The lipid composition is known to affect insertion and folding in studies with peptides, full-length membrane proteins, and co-translationally during cell-free expression (summarised in Table 1). Here, we propose how bilayer properties may influence each stage of co-translational insertion and folding guided by such studies.

Figure 5.. Methods to measure co-translational TM insertion in vivo.

(A) Measuring TM insertion using a glycosylation assay. Glycosylation sites are engineered into the protein flanking the TM of interest. Glycosylation cannot occur on one of the sites if the TM has inserted across the membrane. The amount of single and doubly glycosylated protein is compared with calculate the ΔG_app of membrane insertion of that particular TM helix. (B) Measuring TM insertion using force profile analysis. In these experiments an arrest peptide (AP) is engineered into the protein of interest (B, i) to study the force generated by a nascent chain during translation in vivo [12,74,75]. APs stall translation with a duration that is proportional to the force exerted on the nascent chain during translation by the ribosome. Where little force is generated on the nascent chain (e.g. L₁, where TM2 is not interacting with SecYEG, B, ii), the AP stalls translation and only arrested products are generated. Where a higher force is generated (e.g. L₂, where TM2 is integrated into the membrane by SecYEG, B, iii), the AP stalling is released and a full-length product is generated. Mutations in a TM to introduce more hydrophilic residues (B, iv) can alter the force exerted on the nascent chain (TM2, green, compared with TM2^mut, purple). A plot of the fraction of full-length product against the position of the AP in the nascent chain amino acid sequence (L) provides details into the force acting on a nascent chain during translation (B, v).

Figure 6.. Measuring co-translational TM insertion using cell-free expression.

Cell-free transcription-translation in vitro is used to express a membrane protein in liposomes composed of different lipids (A). The cell-free reaction components are separated from the liposomes on a sucrose gradient (B), and the total amount of protein in each lipid is quantified (by SDS–PAGE bands, radioactive counts or by the amount of functional protein) to measure the effect of lipid bilayer properties on insertion and folding (C). ΔG values have not yet been ascertained with this method.