Protein Structure Highlighted in Birkbeck Science Week

At the end of October Birkbeck College hosted its first Science Week, featuring lectures from distinguished researchers in all science departments. Two talks by members of the Department of Biological Sciences highlighted the beauty and importance of protein structure: the topic of the PPS course. Both lecturers came originally from North America; both moved to Birkbeck about twenty years ago; and both have achieved very significant honours as researchers in structural biology since.

Proteins, DNA and the Components of Life

Professor Bonnie Wallace’s research focuses on the structural analysis of proteins and peptides bound to membranes, and particularly on those involved in ion transport and cell signalling. Her talk in Science Week was entitled “Proteins, DNA and the Components of Life”. She began by describing proteins and nucleic acids as large molecules – “macromolecules” – that are essential components of all living cells. They are both polymers of smaller units; as you should know by now, proteins are polymers of amino acids and nucleic acids similarly are polymers of nucleotides. Amino acids combine as “beads on a string” in many different ways to form the vast array of known proteins; the functional – structural, metabolic or transport among many others – of each is determined by its three-dimensional structure. The techniques used for determining protein (and DNA) structure, including X-ray diffraction, which have developed from fairly crude beginnings about fifty years ago, now allow even large structures to be determined quickly. The section of Professor Wallace’s talk that introduced basic protein structure was illustrated with a physical model of a protein from the 1970s that was then compared with modern molecular graphics representations of protein structures. “Cartoon”, “ball and stick” and “space-filling” protein models, as well as stereo views that can give the illusion of three-dimensionality, are all useful for different purposes.

DNA molecules also are linear polymers, and these – particularly in eukaryotic cells – are extremely long and need to be packed very tightly if they are to fit into the tiny cell nuclei. The packing mechanism involves interactions between the DNA molecule and proteins called histones, to form a structure resembling “beads on a string” that condenses further to form fibres and eventually chromosomes.

The final part of Professor Wallace’s talk concerned the evolutionary differences between DNA in different individuals and in different species, celebrating the tenth anniversary of the publication of the first draft of the human genome. Genomics has provided a useful correction to human arrogance by revealing that many organisms have larger, if not necessarily more complex, genomes than ours. More importantly, differences between bacterial and human cells, between cancer cells and normal ones, and between dysfunctional (mutated) and normal versions of disease-linked proteins, are enabling us to design and develop drugs against a wide range of diseases. Important drugs for infectious diseases including influenza and HIV have been designed using “structure-based” techniques. Professor Wallace ended her talk by describing some of her group’s work on the structure of a protein that is embedded in a cell membrane and provides a channel for sodium ions to pass into cells. Anti-epileptic drugs such as lamotrigine work by binding to and blocking these channels, which will be discussed in detail in the section on Membrane Proteins towards the end of the PPS course.

Model for the structure of a human sodium channel

The Arms Race between Man and Pathogen

Professor Helen Saibil studies proteins using electron microscopy, which can determine the structure of very large proteins and of protein complexes embedded in cell membranes. Her talk, intriguingly titled “The Arms Race between Man and Pathogen”, focused on two pieces of her group’s research into proteins that punch holes, or pores, in membranes. These independent studies led to the discovery of an unexpected similarity between a group of bacterial proteins that cause damage during infection and one of human (and other mammalian) proteins that help protect against viral infection and blood cancers.

Pore-forming protein toxins are “weapons” that bacteria can use in order to break holes in the cells of their human or other animal hosts. They are used by a wide range of bacteria, including some very dangerous ones; one example is Streptococcus pneumoniae, which, as its name implies, causes pneumonia. Unusually, these proteins can exist in two forms, one that is water-soluble and another that is membrane-embedded. The change between these forms requires a considerable rearrangement of the protein’s shape, with one hydrophobic (“water-hating”) segment that is located deep in the core of the soluble protein uncurling to form a “beta-hairpin” structure that embeds in the membrane.

In 2005, Professor Saibil’s group used electron microscopy to determine the structures of two forms of the pore forming toxin from S. pneumoniae, pneumolysin, that together illustrate its mechanism of action. Initially, the protein associates into ring-shaped “pre-pores” of about forty monomers that sit on the membrane surface; the pore itself is formed when part of each monomer changes shape and burrows deep into the membrane. They were able to fit the more detailed X-ray crystal structure of a related protein as a soluble monomer into the shape of the pre-pore protein, and show the type of rearrangement that would be necessary for part of that protein to “collapse” into the membrane.

Another protein that has been studied by Professor Saibil and her colleagues recently is perforin, which is secreted by some lymphocytes (white blood cells that form an important part of the immune system) and which also punches holes in membranes: this time, in the membranes of infected or cancerous cells. This allows the lymphocyte to deliver toxins to the affected cell that trigger the process of apoptosis, or programmed cell death. People with a defective form of this protein suffer from severe viral infections and have a greatly increased risk of blood cancers.

There is no detectable similarity between the sequence of perforin and that of pneumolysin. However, when the first structure of a human perforin-like molecule was discovered in 2007, researchers were surprised to discover that its fold was similar to that of the bacterial protein. It became clear that human and bacterial pore-forming proteins must have evolved from a common ancestor. Protein structures are more tightly conserved than gene and protein sequences, and the structures of distantly related proteins are often much more clearly similar than their sequences.

Hoping to understand more about the mechanism of this protein, Professor Saibil’s group then used electron microscopy to solve the structure of perforin assemblies in contact with membranes. They observed pore structures that were similar, but not identical, to those of pneumolysin. In particular, the perforin structures indicated a much smaller conformational change between non-bound and membrane bound forms than the one seen with pneumolysin. Also, the perforin molecule seemed to be arranged “inside out” in the pore structure as compared to that of pneumolysin, an observation that the group was able to verify with biochemical tests. This finding was very unexpected, but turned out to be consistent with a twist between two parts of the pneumolysin molecule that does not occur in perforin. Both molecules have “arms” that extend down into the membrane to form the sides of the pore, but perforin’s arms are much longer and do not require as big a conformational change to reach down into the membrane.

View of the cryo-EM reconstruction of a perforin pore, with the front half cut away to show how modelled perforin molecules fit into the EM electron density.

The grey surface is the EM map and the subunits are coloured by domain / subregion, as follows. N-terminal region, blue; central beta sheet, red; C2 domain, yellow; modelled beta hairpins (pore lining), orange; EGF domain, green; C-terminal region, magenta. The membrane is seen as a double layer of grey electron density at the base of the pore.

This story of two related proteins, one on each side of the battle between man and pathogen, features many aspects of protein structure research that are covered during the PPS course. These include the beta-hairpin structure, the interatomic forces that determine the way proteins fold, and the structures of membrane proteins. If, however, you have been inspired to learn more about the way electron microscopy is used to study protein structure, you should aim to take our second-year MSc module, Techniques in Structural Molecular Biology.