Structural Bioinformatics of Membrane Proteins

Although the genes that encode membrane proteins make about 30% of the sequenced genomes, the evolution of membrane proteins and their origins are still poorly understood. Here we address this topic by taking a closer look at those membrane proteins the ancestors of which were present in the Last Universal Common Ancestor, and in particular, the F/V-type rotating ATPases. Reconstruction of their evolutionary history provides hints for understanding not only the origin of membrane proteins, but also of membranes themselves. We argue that the evolution of biological membranes could occur as a process of coevolution of lipid bilayers and membrane proteins, where the increase in the ion-tightness of the membrane bilayer may have been accompanied by a transition from amphiphilic, pore-forming membrane proteins to highly hydrophobic integral membrane complexes.

We here review studies concerned with the evolutionary pathways taken for the appearance of complex transport systems. The transmembrane protein constituents of these systems generally arose by (1) intragenic duplications, (2) gene fusions, and (3) the superimposition of enzymes onto carriers. In a few instances, we have documented examples of “reverse” or “retrograde” evolution where complex carriers have apparently lost parts of their polypeptide chains to give rise to simpler channels. Some functional superfamilies of transporters that are energized by adenosine triphosphate (ATP) or phosphoenolpyruvate (PEP) include several independently evolving permease families. The ubiquitous ATP-binding cassette (ABC) superfamily couples transport to ATP hydrolysis where the ATPases are superimposed on at least three distinct, independently evolving families of permeases. The prokaryotic sugar transporting phosphotransferase system (PTS) uses homologous PEP-dependent general energy-coupling phosphoryl transfer enzymes superimposed on at least three independently arising families of permeases to give rise to complex group translocators that modify their sugar substrates during transport, releasing cytoplasmic sugar phosphates. We suggest that simple carriers evolved independently of the energizing enzymes, and that chemical energization of transport resulted from the physical and functional coupling of the enzymes to the carriers.

The first atomic-resolution structure of a membrane protein was solved in 1985. After 25 years and 213 more unique structures in the database, we learned some remarkable biophysical features that thanks to computational methods help us to model the topology of membrane proteins (White 2009). However, not all the features can be predicted with statistically relevant scores when few examples are available (Oberai et al. Protein Sci 15: 1723–1734, 2006). Too often the notion that similar functions are supported by similar structures is expanded far behind the limits of a safe sequence identity value (>50%) to select templates for modeling the membrane protein at hand. To select proper templates we introduce a strategy based on the notion that remote homologs can have a role in determining the structure of any given membrane protein provided that the two proteins are co-existing in a cluster. Sequences are clustered in a set provided that any two sequences share a sequence identity value ≥ 40% with a coverage ≥ 90% after cross-genome comparison. This procedure not only allows safe selection of a putative template but also filters out spurious assignments of templates even when they are generally considered as the structure reference to a given functional family. The strategy also can play a role in indicating which membrane protein sets still would be worthwhile a structural investigation effort. Possibly when more membrane proteins will be available, the clustering system will allow fold coverage of the membrane protein universe.

Transmembrane β-barrel (TMB) proteins are embedded in the outer membranes of mitochondria, Gram-negative bacteria, and chloroplasts. These proteins perform critical functions, including active ion-transport and passive nutrient intake. Therefore, there is a need for accurate prediction of secondary and tertiary structures of TMB proteins. A variety of methods have been developed for predicting the secondary structure and these predictions are very useful for constructing a coarse topology of TMB structure; however, they do not provide enough information to construct a low-resolution tertiary structure for a TMB protein. In addition, while the overall structural architecture is well conserved among TMB proteins, the amino acid sequences are highly divergent. Thus, traditional homology modeling methods cannot be applied to many putative TMB proteins. Here, we describe the TMBpro: a pipeline of methods for predicting TMB secondary structure, β-residue contacts, and finally tertiary structure. The tertiary prediction method relies on the specific construction rules that TMB proteins adhere to and on the predicted β-residue contacts to dramatically reduce the search space for the model building procedure.

Multiple sequence alignment remains one of the most powerful tools for assessing evolutionary sequence relationships and for identifying structurally and functionally important protein regions. Membrane-bound proteins represent a special class of proteins. The regions that insert into the cell membrane have a profoundly different hydrophobicity pattern as compared with soluble proteins. Multiple alignment techniques employing scoring schemes tailored for sequences of soluble proteins are therefore in principle not optimal to align membrane-bound proteins. In this chapter we describe some of the characteristics leading transmembrane proteins to display differences at the sequence level. We will also cover computational strategies and methods developed over the years for aligning this special class of proteins, discuss some current bottlenecks, and suggest some avenues for improvement.

The function of a membrane protein is dependent on that it is inserted into the lipid bilayer in a correct way. Intriguingly, for a small number of membrane proteins, there is growing evidence that they have flexible topologies. Some of them have a varying number of helices inserted into the membrane, or have the same number of transmembrane helices, but different membrane spanning regions. Others are inserted with opposite topologies in the membrane in an approximate 1:1 ratio, forming antiparallel homodimers. Thus, the same sequence can code for more than one topology. During the last few years, there have been increasing efforts in studying topologically flexible proteins since they might hold clues about the evolution of membrane proteins.

In multipass transmembrane proteins one face of the transmembrane helices is in contact with the aliphatic acyl chains of the phospholipids and with the polar interface region. The other face makes contacts with other helices or points into the protein interior. In larger proteins, some helices may even be buried completely. Analysis of the available three-dimensional crystal structures has shown that inwards pointing residues tend to be more conserved than outwards pointing residues. Furthermore, residues pointing outwards are generally very hydrophobic whereas inward pointing residues may have different characteristics. Based on these two findings, knowledge-based propensity scales have been derived that, when combined with analysis of residue conservation, allow predicting the exposure status of residues in the hydrophobic core region with about 80% accuracy. These tools give biologists insight in the putative topology of transmembrane helix bundles.

Membrane-spanning α-helices represent major sites of protein-protein interaction in membrane protein oligomerization and folding. As such, these interactions may be of exquisite specificity. Specificity often rests on a complex interplay of different types of residues forming the helix-helix interfaces via dense packing and different non-covalent forces, including van der Waal’s forces, hydrogen bonding, charge-charge interactions, and aromatic interactions. These interfaces often contain complex residue motifs where the contribution of constituent amino acids depends on the context of the surrounding sequence. Moreover, transmembrane helix-helix interactions are increasingly recognized as being dynamic and dependent on the functional state of a given protein.

Helix-helix contacts are an important feature of alpha-helical membrane proteins as they define their characteristic helix bundle structure. No bioinformatics approaches for the prediction of pairwise residue contacts in membrane proteins have existed until recently. In this chapter we describe novel contact prediction methods based on residue coevolution and machine learning techniques specifically geared towards membrane proteins. While contact prediction accuracies are limited to ~10% using co-evolving residues alone, machine learning methods are able to improve these accuracies significantly to more than 25% by using available membrane protein structures as a training dataset and incorporating membrane protein specific sequence features into the prediction process. Importantly, predicted residue contacts allow for identification of interacting transmembrane helices with high accuracy. As different membrane protein structures can be distinguished by their specific pattern of helix interactions, predicted residue contacts may not only serve as structural constraints in modeling experiments, but also constitute valuable information for structural classification of membrane proteins with unknown structure.

Transmembrane (TM) helical proteins are of fundamental importance in many diverse biological processes. To understand these proteins functionally, it is necessary to characterize the forces that stabilize them. What are these forces (both within the protein itself and between the protein and membrane) and how do they give rise to the multiple conformational states and complex activity of TM helical proteins? How do they act in concert to fold TM helical proteins, create their low-energy stable states, and guide their motion? These central questions have led to the description of critical natural constraints and partial answers, which we will review. We will then describe how these constraints can be tracked through homologs and proteins of similar folds in order to better understand how amino acid sequence can specify structure and guide motion. Our emphasis throughout will be on structural features of TM helix bundles themselves, but we will also sketch the membrane-related aspects of these questions.

The family of G-protein-coupled receptors (GPCRs) is by far the best-studied family among the integral membrane proteins, because it represents the largest and most important group for therapeutics. In this chapter we provide an overview of the major developments in the GPCR field since the 19th century, and we shed some light on some of the questions that are relevant now and those that need to be answered in the future regarding GPCR structure and function.