Volume 39, August 2016, Pages 39–45

Engineering and design • Membranes

Edited By Tawfik, Varadarajan, Gulbis and Waksman

Computational design of protein self-assembly

  Open Access

Highlights

The recent progress in the design of self-assembling protein complexes is reviewed.

Methodological advancements have enabled design of assemblies with atomic accuracy.

Biophysical principles that favor or complicate self-assembly design are discussed.

Design complexity has been decreased by use of oligomeric building blocks and symmetry.

Design of complex assemblies may in the future benefit from use of monomeric building blocks.


Protein self-assembly is extensively used in nature to build functional biomolecules and provides a general approach to design molecular complexes with many intriguing applications. Although computational design of protein–protein interfaces remains difficult, much progress has recently been made in de novo design of protein assemblies with cyclic, helical, cubic, internal and lattice symmetries. Here, we discuss some of the underlying biophysical principles of self-assembly that influence the design problem and highlight methodological advances that have made self-assembly design a fruitful area of protein design.


Current Opinion in Structural Biology 2016, 39:39–45

This review comes from a themed issue on Engineering and design

Edited by Dan Tawfik and Raghavan Varadarajan

http://dx.doi.org/10.1016/j.sbi.2016.04.002

0959-440/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Introduction

Much of the molecular complexity of life is formed by self-assembly of protein monomers into higher-order oligomers. Self-assembly is thus a powerful design template to create complex molecular assemblies from a limited number of building blocks. In the simplest case a single type of building block is sufficient to generate homomeric structures with complex morphologies such as rings, filaments or containers. These assemblies enable functions such as multivalent binding, ultrasensitive regulation and compartmentalization and are therefore ubiquitous in biology [1]. If controlled, the functions encoded by self-assembly in natural systems could be replicated or extended to novel applications in biotechnology, biomedicine and material science [2].

The quaternary structure of self-assembling proteins is stabilized by protein–protein interfaces. Accurate design of protein interfaces is therefore required to control self-assembly. Beyond the design of the simplest α-helical assemblies, computational methods are necessary to explore the vast space of protein interface sequences. Computational methods are also necessary to find suitable building blocks and binding geometries when designing complexes of novel components and assembly structures. Until recently, de novo protein interface design was primarily directed towards heterodimers but in the past few years tremendous progress in design of higher-order protein assemblies has been made. The advancements can be attributed to new design methodologies, but the biophysical properties of self-assembling proteins and peptides also make them particularly amenable to design.

We begin our review with a description of the biophysical properties of self-assembling homomeric systems and their implications for design. We then review the recent advances in designing homomeric self-assembly using computational methods. Finally, we describe some of the future challenges in the field.

Biophysical properties of protein self-assembly

Design of protein complexes is simplified by fundamental physical properties of self-assembling systems, which may partly explain why this has been such a fruitful direction in protein design. In this section we briefly consider properties that simplify and set unique challenges in design of protein assemblies.

The vast majority of all homomers have nearly perfect structural symmetry [1]. This simplifies modeling and design calculations as we can assume that homomeric complexes are made up of building blocks with identical structure and interactions (Figure 1a) [3]. Furthermore, symmetry also strongly limits the number of ways that protein subunits may associate in three dimensions. Because interfaces are replicated by symmetry, fewer independent residues must be designed in homomers compared to heteromers or asymmetric homomers. Limiting the number of residue changes in the protein building block is crucial, as the underling assumptions for design (e.g. rigid backbone) are more likely to fail, when more residues are changed.

Biophysical principles of protein self-assembly. Biophysical principles works ...
Figure 1. 

Biophysical principles of protein self-assembly. Biophysical principles works both for and against design of self-assembling protein complexes. (a) Symmetry simplifies the design calculations by limiting the number of subunits and interfaces to be considered in design calculations. (b) Multivalency yields stable assembly formation from weak protein–protein interfaces. The apparent disassociation constant is defined as the concentration where the building block and the capsid concentration are equal. (c) Design of oligomerization specificity is complicated by isoenergetic energy landscapes between different symmetries and oligomers. Further complicating design is that the relative stability of different oligomers is concentration dependent. Finally (d) assembly of oligomers is complicated by the possible formation of kinetic traps and aggregates during assembly.

A second property that favors design of self-assembling structures is avidity. Avidity yields stable complexes from building blocks with weak interfaces. This is for instance illustrated in homomeric icosahedral protein capsids, which can be thought of as assemblies of multivalent cyclic symmetric building blocks. Here the avidity lets the capsids assemble with an apparent stability that is six orders of magnitude higher than the affinity between the individual cyclic symmetric building blocks (Figure 1b) [6 and 7]. As it remains challenging to design high affinity interfaces by computation alone, the affinity leverage provided by avidity provides a key benefit in the design of higher-order symmetric homomers.