Evolution and Dynamics of Protein Complexes: 2

A lecture by Dr Sarah Teichmann of the MRC Laboratory of Molecular Biology, Cambridge, UK at the Institute of Structural and Molecular Biology Symposium, 21 June 2012

Continued from previous post:



Dr Teichmann’s group is harnessing the information available using their database, 3Dcomplex.org.  The 20,000 protein complexes in a cell have been represented graphically with a system of lines and dots symbolising the topology of the monomeric building blocks. 

Illustration of graphs used to represent different oligomeric complex structures, taken from 3Dcomplex.org, Levy E.D., Pereira-Leal J.B., Chothia C., Teichmann S.A., (2006) 3D complex: a structural classification of protein complexes. PLoS Comput Biol. 2(11):e155.

These graphical representations can be used to give a hierarchical classification of cellular complexes and then to discover evolutionary links between pathways which use similar mechanisms.  Since approximately 2/3 of proteins form symmetric macromolecular machines, with only 1/3 being monomeric in the native state, this gives an unprecedented insight into the evolution of cellular protein networks.

The complexes on the database are classified initially into homomers, which are comprised of repeats of a single monomer and account for approximately 90% of the complexes, and heteromers, which are complexes of different monomers.  Most are then separated into either cyclic or dihedral symmetry. 

The preponderance of homomers is explained by various selective advantages.  Proteins achieve greater stability as oligomers since the surface area at the interfaces, which would be solvent accessible in a monomer, becomes buried.  There are also a variety of functional advantages.  Oligomers such as actin and tubulin (as examined in PPS section 7) can polymerise.  Some complexes, like haemoglobin, achieve cooperativity through allosteric change (explained in PPS section 10).  Some oligomers provide a template or scaffold for higher order assemblies, of which there are many examples in cross membrane proteins (some examples are in PPS section 11).  In some cases the act of oligomerization provides functional regulation, for example caspase-9 is activated by dimerization to initiate apoptosis (programmed cell death).

Oligomerization provides major pathways in assembly and, consequently, in the evolution of cellular interactions.  Many of these are based on dimers, which represent the most stable assembly as each interaction between residue pairs manifests twice.  This carries large evolutionary currency because each mutation at the interface has a double effect.  Tetramers are often evolved in a stepwise manner, as a dimer of dimers, such that the dihedral tetramers are much more common than cyclical tetramers, and higher order dihedral complexes tend also to assemble from dimer conglomeration.  Where there are certain functional stability issues, however, such as in membrane proteins, cyclical oligomers do predominate and stepwise evolution can occur with the oligomerization of higher order cyclical oligomers.

Illustration of the most common forms of homomeric symmetry, taken from 3Dcomplex.org, Levy ED, Pereira-Leal JB, Chothia C, Teichmann SA, (2006) 3D complex: a structural classification of protein complexes. PLoS Comput Biol. 2(11):e155.

 

It has been found that in monomers which are known to oligomerise, there is often homology at the largest interface, showing that this interaction interface is conserved (Levy et al., (2008) Nature).  This conservation reflects assembly pathways of complexes since (Levy et al., (2008) Nature) showed that assembly occurs via an ordered pathway, with intermediate species corresponding to sub-complex with large interfaces.  Therefore, the preservation of the largest oligomeric interface can be thought of as a unifying principle in the evolution and assembly of homomers.  Once recognised, this principle can be used to predict interfaces based on sequence conservation across homologues and, therefore, provides crucial information in complex structure prediction.

Protein folding is selected for in evolution to be both fast and spontaneous.  We can infer from this, and from our observation of the crowded nature of the cytoplasm, that complex assembly must also be ordered and fast.  In heteromers, the order of assembly is, therefore, of vital importance, since establishing interactions in the wrong order will lead to incorrect complex structure.  Dr Teichmann is interested in establishing a basic principle that heteromer assembly occurs via spontaneous ordered pathways with fixed intermediates and this is currently under investigation.