PerspectiveStructural Biology

(Pseudo-)Symmetrical Transport

See allHide authors and affiliations

Science  25 Jan 2013:
Vol. 339, Issue 6118, pp. 399-401
DOI: 10.1126/science.1228465

Most walls incorporate some way of getting through them, such as a gate, a revolving door, or an airlock. The airlock concept is also used in moving cargo across cell membranes, for example, for intake of nutrients or expulsion of toxins, which is often an uphill battle against a concentration gradient. Nature fights this battle using membrane-embedded proteins called transporters. These proteins can alternate between at least two preferred conformations. In one of those conformations, the cargo diffuses in and binds to its binding site. The first door then shuts and the other opens, providing unhindered access to the other side of the membrane. Recent studies suggest a key role for structural symmetry in this process.

Any given cell contains dozens of different transporters, each with its own preferred substrate, relying on various energy sources. One class of transporters uses the energy released by a chemical reaction, such as adenosine triphosphate hydrolysis. A second class, the secondary transporters, uses downhill diffusion of one molecule to power the uphill work for another. For example, nutrient/metabolite transporters enable a cell to take up a nutritious substrate against its concentration gradient and, after consuming it, allow the waste to diffuse back out.

The rules of transport in transporters, including the airlock-like alternating-access mechanism, have long been known. Only in the past decade, however, has a clear picture of the structures of these proteins been unveiled, thanks to x-ray crystallography (1). A recurring theme in these structures, particularly of secondary transporters, is the presence of internal pseudosymmetry of sets of membrane-spanning helices. Sometimes the repeats are arranged in a circular symmetry. However, most transporter architectures have a pseudo-two-fold symmetry, with the two halves of the protein related by an axis running along the plane of the membrane. The two inverted halves are usually intricately intertwined, like clasped hands with interlocking fingers.

Membrane-embedded airlocks.

(A) Inverted-topology repeats with similar structures can form channels through the membrane. (B) In transporters, the difference between the two halves creates an asymmetry in the overall protein structure, resulting in opening of the outer or inner gate. (C) In the antiparallel form of the dimeric transporter EmrE, the outside-open structure (left) has the same energy as the inside-open structure (right, same structure rotated 180°). (D) In lactose permease, the repeated elements of the protein are part of a continuous polymer chain and are intertwined. As in EmrE, the two halves appear to alternate between two distinct conformations. (E) Exchange between the two conformations of the aspartate transporter, GltPh, involves a dramatic elevator-like movement of the substrate across the membrane.

Such inverted repeat elements are not unique to secondary transporters. Another set of membrane proteins, called channels, also include these arrangements. Duplication in this case may be a conservative solution to forming a continuous pathway across the membrane, because it results in two half-channels made from similar components that line the axis of the symmetry (see the figure, panel A). However, the inverted repeat elements of secondary transporter structures differ from those of channels: In any given conformation, whether open to the outside or to the inside of the cell, the repeated elements have slightly different arrangements (see the figure, panel B). These differences lead to the structural asymmetry that is required if one door is to be open while the other is closed.

This concept has been illustrated for several different transporter architectures (27) using a modeling strategy that involves swapping the three-dimensional structures of the two repeated halves, starting with the structure of a transporter with one door open. Using homology modeling—in which protein structures are modeled based on known structures of related proteins—the structure of one half of the transporter is used as the template onto which the sequence of the other half is threaded. The result is a three-dimensional model of the same transporter but with the other door open (see the figure, panels C to E).

Different aspects of these models have been experimentally corroborated (35). An inward-facing conformation predicted for an aspartate transporter, for instance, was captured by introducing sulfhydryl groups at specific positions in the protein and reacting them with mercury ions (8). The structure of this trapped state was then determined by x-ray crystallography (see the figure, panel E).

Other informative approaches include engineering fluorescent or paramagnetic probes onto the protein surface and measuring their proximity by fluorescence resonance energy transfer (FRET) or electron spin resonance (EPR). Such data have shown that the intracellular water-exposed loops of the lactose permease transporter are closer together (9) in one of the airlock states than in the other airlock state known from x-ray crystallography (see the figure, panel D), consistent with predictions based on the inverted-topology repeats (3). Other studies measure the accessibility of putative substrate pathways as indicated by the reactivity of engineered sulfhydryl groups to chemical modification or cross-linking (2, 4). Repeating these measurements with different substrates, state-specific inhibitory compounds, or antibodies provides insights into conformational changes during transport.

The transporters mentioned so far comprise a single continuous protein chain, with different amino-acid sequences in the two repeated elements. A much simpler transporter, EmrE, has two identical chains, which according to available structural data (10, 11) are arranged in opposite directions in the membrane. Biochemical analyses suggest that the protein normally prefers to function with the two halves parallel in the membrane (12). Nevertheless, the antiparallel form provides a useful case study, because the two chains form slightly different structures in this arrangement (see the figure, panel C) (10, 11). Nuclear magnetic resonance (NMR) studies suggest that EmrE exchanges between two degenerate (that is, same-energy) states (13). By interchanging between these conformations, the transporter opens either its outer or inner door. Because both shapes are similar in energy, the protein avoids the need for additional energy input beyond that typically gained from binding.

Will all pseudosymmetric transporters function this way, and if not, how much do the amino acid sequences of the two halves have to differ before the principle of degenerate states breaks down? Different transporters sharing the same overall architecture may connect their states via different types of movements (14), although the key factors underlying these distinctions remain incompletely understood. It also remains unclear where the boundary lies between channels (with similar repeat structures) and transporters (with alternating repeat conformations). A notable case is the CLC protein family, which includes both channels and transporters for chloride. Structural and spectroscopic approaches such as EPR, NMR, and FRET should help to elucidate the distinctions between CLC transporters and CLC channels.

How did transporters evolve such complex intertwined polymeric chains? Did they start as transporters made of dimers of adjacent units like EmrE and then begin to swap segments? Hints of these evolutionary pathways have been obtained, for example, by comparison of membrane proteins sequences (15). Similar studies using more sensitive sequence comparison methods to scan ever-growing sequence databases, combined with structural analysis of the identified proteins, should provide important insights into this intriguing evolutionary process.


Navigate This Article