COMMENTARY Membrane protein folding makes the transition Paula J. Bootha,1 and Jane Clarkeb,1 aDepartment of Biochemistry, School of Medical Sciences, University of Bristol, Bristol BS8 1TD, United Kingdom; and bDepartment of Chemistry and Medical Research Council Centre for Protein Engineering, University of Cambridge, Cambridge CB2 1EW, United Kingdom he study of the folding of mem- brane proteins has lagged far T behind that of small soluble proteins—yet proteins that reside within biological membranes ac- count for approximately a third of all proteomes. The article by Huysmans et al. in this issue of PNAS (1) represents a breakthrough by reporting a compre- hensive ϕ-value analysis of the folding of a membrane protein (i.e., PagP) into a lipid bilayer. ϕ-value analysis is the most powerful tool for experimental analysis of protein folding pathways (2). It com- bines protein engineering with equilibrium and kinetic measurements to determine which regions of a protein are largely fol- ded (high ϕ-values) or largely unfolded (low ϕ-values) at the rate-limiting transition state. Fig. 1. Schematic diagrams of proposed folding models for β-barrel and α-helical membrane proteins, There are two major structural classes highlighting potential transition state structures from ϕ-value studies. Folding of a β-barrel protein oc- of integral membrane proteins: α-helical curs from a fully denatured, membrane-absorbed state in urea with a tilted, partly inserted transition bundles and β-barrels. The latter are found state as proposed by Huysmans et al. (1). In contrast, folding of an α-helical protein such as bacterio- in the outer membranes of Gram- rhodopsin occurs from a partly denatured state in SDS, which contains some helical content. The tran- sition state is proposed to have significant native helix content. Only one transmembrane helix has been negative bacteria and mitochondria, ϕ whereas helical structures are ubiquitous, analyzed by -values (3), and this suggests a largely formed helix with partial helix formation at the cytoplasmic side (shown in the bottom of this diagram, with this helix outlined in black). The degree of occurring, for example, in plasma and in- β structure in other helices is estimated from previous studies (e.g., ref 15). Unfolded structure is shown in ner membranes. PagP has a -barrel red and folded structure in blue. SDS is shown in green. PagP was folded into lipid bilayers vesicles, structure and resides in the outer mem- whereas bacteriorhodopsin was folded into mixed DMPC/CHAPS micelles (the detergent CHAPS is shown brane of Escherichia coli. The analysis here capping the edges of a DMPC disc). presented by Huysmans et al. (1) sheds light on the transition state structure for sible and protein concentration– polarized transition state with the barrel formation of this β-barrel in a bilayer. The independent. The folding and unfolding partly formed and the C-terminal part of work complements a previous ϕ-value kinetics were determined, and, perhaps the protein significantly more structured study of membrane protein folding that surprisingly, conditions could be found than the N-terminal regions. The authors was the first to map out a membrane (urea concentrations greater than approx- protein transition state, but in this case, for propose that the results are consistent with imately 8 M) in which folding could be folding of an α-helical protein, bacterio- a tilted folding-insertion mechanism (Fig. rhodopsin, into lipid-detergent mixtures characterized as a simple two-state re- 1), as has been seen in simulations for in- action. It is also of note that the choice of sertion of folded OmpA into a bilayer (7). (3). Thus detailed insight into the folding β mechanisms of the two major membrane a -barrel protein means that folding can The results also agree with previously protein structural classes is now emerging. be studied from an apparently fully un- proposed models of concerted folding and A key feature of this current study (1) is folded, denatured state. Most helical insertion of β-barrels into membranes (8); fi the examination of folding into a lipid bi- membrane proteins retain signi cant re- as an unfolded barrel protein within a bi- layer. To achieve this, Huysmans et al. (1) sidual structure in common denaturants layer is unlikely, as it would expose po- exploited a previously demonstrated trait such as urea and SDS, but the denatured tential hydrogen bonding groups (9). of β-barrel proteins: they can be reversibly state of PagP in 10 M urea, although as- Helical membrane proteins, by contrast, refolded from a urea-denatured state into sociated with the bilayer, is essentially seem to fold by a fundamentally different lipid bilayer vesicles, and the kinetics and unstructured (as judged by circular di- mechanism than barrels (9, 10). Individual thermodynamics of the folding reaction chroism and fluorescence spectra). The transmembrane helices can be stable en- can be determined (4, 5). The successful use of urea gives added importance to the tities in bilayers, as backbone hydrogen demonstration of this for PagP (6) paved PagP work, as urea apparently does not bonding is satisfied locally within the helix. the way for a ϕ-value study. Identification partition into the bilayer, unlike SDS used of conditions for reversible folding of PagP for studies of the folding of several relied on manipulation of the lipid to α-helical proteins. Author contributions: P.J.B. and J.C. wrote the paper. protein ratio. At low ratios, PagP folding Nineteen variants of PagP were inves- The authors declare no conflict of interest. was protein concentration-dependent, but tigated, at least one mutation in each See companion article on page 4099. β at the lipid-to-protein ratio used here -strand, thus giving a snapshot across the 1To whom correspondence may be addressed. E-mail: paula. (3,200:1), refolding was completely rever- entire protein. The protein folds via a [email protected] or [email protected]. www.pnas.org/cgi/doi/10.1073/pnas.0914478107 PNAS Early Edition | 1of2 Downloaded by guest on September 25, 2021 Thus partially folded states occur within a most of the helix is formed in the tran- structures. A full understanding of the in- membrane, as conceptualized by the clas- sition state, but intermediate and low fluence of the lipid bilayer will require sic two-stage folding model, whereby ϕ-values toward the cytoplasmic end of the further investigation of several proteins helices first form within a membrane (10). helix suggest only partial helix formation under contrasting lipid conditions, to- Final helix packing and tertiary structure and tertiary contacts. Intermediate ϕ-val- gether with a quantitative understanding formation then occurs as a second stage ues can also arise from multiple transition of the relevant lipid properties. (Fig. 1). Models emerging for in vivo states or folding paths, which could be In some respects, the structures of folding also propose this type of folding envisaged for packing of stable helical membrane proteins, with only two main mechanism, as transmembrane helices entities. Many intermediate ϕ-values are structural classes, are generally less com- form within the translocon complex also seen in the PagP study (1). The pos- plex than those of soluble proteins. Huys- during cotranslational folding, and this is sible contributions of the bilayer environ- mans et al. (1) suggest that there might be fi followed by nal helix packing and folding ment and intermolecular forces to such a generic tilted folding-insertion mecha- within the membrane (11, 12). Inves- intermediate values await further exami- nism for β-barrel membrane proteins. tigations of folding mechanisms by studies nation for both protein classes. Conservation of folding mechanisms in in vitro are also consistent with the Membrane proteins may reside within a protein families has been observed before overall two-stage idea, although the sit- heterogeneous and largely hydrophobic —particularly in all β-proteins in which the uation is more complex than two simple environment, which complicates exper- topology is complex (14). This lends fur- steps (9, 13). imental study, but given careful consid- ther support to the notion that there may The increased hydrophobicity of helical eration of experimental conditions, membrane proteins compared with membrane protein folding can be inves- be general folding mechanisms for the two β-barrels makes the former more chal- tigated in detail, and many of the folding classes: insertion and packing of lenging to experimental manipulation. approaches that have worked well for preformed helical elements or concerted Consequently, the folding system used for soluble proteins can be suitably trans- folding and insertion. Intriguingly, the β the bacteriorhodopsin ϕ-value study (3) is ferred. These successful applications of early events in the folding of the -barrel — more complex than the urea/bilayer one ϕ-value analysis are especially propitious, proteins might also be more complex in used by Huysmans et al. for PagP (1). A as they will reveal structures of key mem- the study by Huysmans et al., the folding partially denatured state of bacterio- brane protein intermediates and transition kinetics of PagP became much more dif- rhodopsin in SDS was used; this contains states. The membrane environment itself ficult to interpret below 7.8 M urea—sug- just more than half the native helix content directly affects protein folding (13), with gesting that early stable intermediates and seems to represent a critical helical the membrane elastic properties influenc- might be populated. core that is a prerequisite for correct ing the folding rates, yields, and pathways The importance of this study and that of folding. More extensive denaturation of of both helical and β-barrel proteins. bacteriorhodopsin is that they show it is bacteriorhodopsin requires acidic organic Matching of the hydrophobic thickness of possible to investigate the folding mecha- acids that are incompatible with refolding the bilayer to the protein is also vital, with nisms of membrane proteins in detail in detergent or lipid systems.