Mechanisms of integral membrane protein insertion and folding

Abstract

The biogenesis, folding, and structure of α-helical membrane proteins (MPs) are important to understand because they underlie virtually all physiological processes in cells including key metabolic pathways, such as the respiratory chain and the photosystems, as well as the transport of solutes and signals across membranes. Nearly all MPs require translocons--often referred to as protein-conducting channels--for proper insertion into their target membrane. Remarkable progress toward understanding the structure and functioning of translocons has been made during the past decade. Here, we review and assess this progress critically. All available evidence indicates that MPs are equilibrium structures that achieve their final structural states by folding along thermodynamically controlled pathways. The main challenge for cells is the targeting and membrane insertion of highly hydrophobic amino acid sequences. Targeting and insertion are managed in cells principally by interactions between ribosomes and membrane-embedded translocons. Our review examines the biophysical and biological boundaries of MP insertion and the folding of polytopic MPs in vivo. A theme of the review is the under-appreciated role of basic thermodynamic principles in MP folding and assembly. Thermodynamics not only dictates the final folded structure but also is the driving force for the evolution of the ribosome-translocon system of assembly. We conclude the review with a perspective suggesting a new view of translocon-guided MP insertion.

Keywords: lipid–protein interactions; membrane protein biogenesis; membrane protein folding; transmembrane helix.

Figure 1

This schematic cartoon represents in broad terms current thinking about the insertion of multi-span proteins into membranes. Two ideas are captured in the cartoon. First, TM segments (red) emerge from the ribosome and pass into the translocon (blue). Second, the nascent segments partition into the membrane from the translocon. As a starting point for discussion, we present alternative views of the membrane protein insertion pathway in Fig. 15.

Figure 2

Alpha-helical MPs exist in their native state in highly thermally disordered lipid bilayers, as illustrated here for the SecYEG translocon [78] from Methanococcus jannaschii. The image is from a molecular dynamics simulation of the translocon (PDB code 1RHZ) executed in a phospholipid bilayer. In this view, parallel to the membrane plane, the so-called gate helices 2b and 7 (red cylinders) were exposed by cutting away the lipid bilayer. Water molecules within the translocon are shown as van der Waals spheres in red (oxygen) and white (hydrogen). Waters surrounding the bilayer are shown as H-O-H bonds in blue-gray. Lipid acyl chains are white and the phospholipid headgroups are red. Image provided courtesy of J. Alfredo Frietes and Stephen H. White.

Figure 3

Summary of the various interactions that stabilize MPs stably folded in fluid lipid bilayers (blue lines are interface boundaries, red lines represent boundaries of the lipid hydrocarbon core). “Global bilayer effects” accounts for changes in the structure and stability of the lipid bilayer when perturbed by the protein [160], emphasizing that the bilayer itself sits in a free energy minimum determined by the tendency of the system to minimize exposure of the acyl chains (grey) to water (blue) on the one hand, and the tendency to maximize the distance between headgroups on the other . Both the bilayer and protein must adjust structurally to minimize the free energy of the protein plus bilayer system. The protein shown (red helices) is bacteriorhodopsin determined to a resolution of 1.55 Å, PDB Code 1C3W [161]. Image modified from [162].

Figure 4

A four-step thermodynamic cycle for describing the energetics of the partitioning, folding, insertion, and association of an α-helix (red helices) in a lipid bilayer (grey). The process can follow an interfacial path, a water path, or a combination of the two. Studies of folding along the interfacial path are experimentally more tractable [163]. The ΔG symbols indicate standard transfer free energies. The subscript terminology indicates a specific step in the cycle. The subscript letters are defined as follows: w = water, i = interface, h = hydrocarbon core, u = unfolded, f = folded, and a = association. With these definitions, for example, the standard free energy of transfer from water to interface of an unfolded peptide would by ΔG_wiu. However, some parts of the cycle (dashed box) are generally inaccessible experimentally. Image modified from [1].

Figure 5

The energetics of inserting an α-helix into lipid bilayers (grey) is dominated by the peptide bonds, as illustrated here for the glycophorin A (GpA) TM helix. Even with the helical backbone internally H-bonded, it is costly to dehydrate the H-bonded peptide bonds (ΔG_bb(f)) upon insertion into the bilayer. For the helix to be stable, the favorable free energy of transfer of the sidechains (ΔG_sc), determined by the hydrophobic effect, must compensate for the unfavorable ΔG_bb(f). In the case GpA, the net stability of the helix ΔG_TM is −12 kcal mol⁻¹. The energetic cost of unfolding the polyglycine helix within the bilayer is immense: ΔG_bb(u) is greater than 100 kcal mol⁻¹! Modified from [162].

Figure 6

Equilibrium microsecond-scale simulations of the folding and membrane insertion of polyleucine sequences (here 10 leucines) reveal only two states [39]. The simulation shown here begins with the unfolded peptide (U) in water (W) about 10 Å from the phosphatidylcholine bilayer (grey). Within a few nanoseconds (ns) it absorbs to the membrane interface and never returns to the bulk water phase. After the next 40 ns, the peptide becomes α-helical and fluctuates between being on the surface (S) and across the membrane (TM) during the ensuing several microseconds. The trajectory of the simulation is represented as a plot of the insertion depth of the peptide's center-of-mass against its helicity. The sampled time points are connected sequentially by blue lines. Modified from [39]. This simulation reveals the importance of the membrane interface in membrane protein folding, and suggests that the interface may play a role in translocon-guided insertion of TM helices.

Figure 7

Helix-helix interactions of the glycophorin A (GpA) dimer (blue and orange helices) based upon the structure of GpA in SDS [47]. A. Glycine (or other small residues) separated by three residues allow the helices to pack tightly. B. Other amino acids in the structural vicinity of the GXXXG motif can facilitate or inhibit specific binding of complementary surfaces [120].

Figure 8

High-resolution structure of a mammalian ribosome-Sec61 complex. The structure was obtained using advanced cryo-EM methods. PTC indicates the peptidyl transferase center. Image from Voorhees et al. [164].

Figure 9

Structure of the SecYE translocon from Pyrococcus furiosus [80]. The left panel shows a view in the plane of the membrane, the right panel shows a view from the cytoplasm. The plug domain is circled, and the residues in the hydrophobic ring are shown as van der Waals spheres, indicated by *. The arrow points into the lateral gate between TMH2b and TMH7. SecE is shown in red, SecY is color coded from N-terminus (blue) to C-terminus (orange).

Figure 10

Two models for translocon-mediated insertion of a N_out-C_in orientated TMH in a single-span (type I) membrane protein. (a) The “In-out” model”. The TMH (in black) first moves all the way into the central translocon channel, and then exits sideways through the lateral gate. (b) The “sliding” model. The TMH slides along the lateral gate into the membrane, with one side exposed to lipid at all times. The leading polar segment penetrates through the lateral gate and is shielded from lipid contact. SecE is shown in red, SecY is color coded from N-terminus (blue) to C-terminus (orange). On the right-hand side a schematic of the translocon (blue) interacting with a membrane protein (green) and a membrane inserting TMH (red) are shown.

Figure 11

Structure of the YidC translocon from Bacillus halodurans [96] viewed in the plane of the membrane, looking into the hydrated cleft in the center of the protein. The left panel shows a surface representation; the right panel shows the walls of the cavity, with the rest of the protein in stick representation.

Figure 12

Amino acid ΔG_app values for Sec61-, SecYEG-, and YidC-dependent membrane insertion of a model TMH [84, 85, 102]. The SecYEG (panel a) and YidC (panel b) data are plotted against the Sec61 data. Full lines indicate linear fits to the data (with the equations given in the panels), while the broken line in panel b is the linear fit obtained when including only non-polar and weakly polar residues (slope = 0.8) [85].

Figure 13

Using APs as in vivo force sensors [106]. (a) An AP (AP; in blue) is inserted into a membrane protein with two natural TMHs (TM1, TM2; in black) and a model TMH composed of six leucines and 13 alanines (H; in red). Depending on the length L of the tether between the H-segment and the AP, the H-segment will be in different locations relative to the translocon at the time when the ribosome reaches the last codon in the AP, as shown in the cartoon. The pulling force F(L) on the nascent chain will determine the fraction of stalled vs. full-length protein produced. The fraction full-length protein can be determined by [³⁵S]-Met pulse-labeling of growing E. coli, followed by immunoprecipitation of the protein construct and analysis by SDS-PAGE, as shown on the right for a construct with L=63 residues (a control construct in which a critical proline in the AP has been mutated to alanine, preventing stalling, is also shown). (b) Fraction full-length protein as a function of L for a set of constructs designed as shown in panel a (top).

Figure 14

Folding spaces during membrane protein structure formation. During nascent chain synthesis by the ribosome (brown), helical segments (shown in blue and orange) can associate already in the ribosomal vestibule (1) [136] or within the translocon channel (2a, green) [138], where polar residues can be shielded within the helix-helix interface (2b, red spheres) and hydrophobic side-chains (orange spheres) form a membrane insertion competent surface. Helices can furthermore interact during insertion into a membrane (3, grey) whereas one segment is already inserted into the membrane and drives the insertion of a less hydrophobic segment by shielding hydrophilic residues within the interaction interface [165]. Interactions of helices within different proteins can also occur co-translationally (5) and was shown do drive membrane protein complex assembly [149].

Figure 15

An alternative view of translocon-aided insertion of multi-span membrane proteins and the secretion of soluble proteins. The idea of this alternative view is shown in a cartoon fashion in panel a (also see Fig. 10). We suggest that initial contact of the nascent chain (green, with red TMHs) is with the membrane interface in the vicinity of the translocon (blue), and that the chain does not immediately thread into the translocon. Rather, the translocon has YidC-like behavior; it provides a pathway for polar components of membrane proteins to cross the membrane. This is not to say that it cannot form a passageway through the membrane, but we suggest that it does this only for polar polypeptide segments in membrane proteins and for secreted proteins, as shown in panel b.