The growth, metabolism and replication of cells, the basic building blocks of life, depend very heavily on proteins. Proteins perform a vast range of functions of both an intracellular and intercellular nature, being expressed within the cells themselves. The profile of expressed proteins within a cell depends upon the metabolic status of the cell, its age and its local environment. Understanding the role that proteins play in the status of the cell is crucial to the understanding of the diseased state and is therefore of great importance within medical and pharmaceutical studies of disease. Their behaviour at interfaces (e. g. surfaces) is of particular relevance given that many biologically important processes are interfacial in nature.

Neville Freeman: Farfield Scientific Ltd, Farfield House, Southmere Court, Elec-tra Way, Crewe Business Park, Crewe, Cheshire, CW1 6GU, UK, E-mail: [email protected]

Principles and Practice Proteins at Solid-Liquid Interfaces Philippe Dejardin (Ed.) © Springer-Verlag Berlin Heidelberg 2006

To appreciate how proteins function and the diverse roles that they are capable of playing, an understanding of the structure of proteins and structural changes occurring in real time is required. Proteins are capable of undergoing substantial conformational and structural changes in the presence of a range of stimuli such as changes in environment, for example pH, the presence of small molecules, such as candidate drug molecules, or the presence of secondary binding partners (typically, but not limited to, other proteins).

The structure of a protein is determined by its primary structure, the sequence in which the amino acids that constitute its composition are co-valently bound together. The nature of the amino acid sequences leads to regions of the molecule forming coils (alpha helices) or sheets (beta sheets), which define the secondary structure of the molecule. The molecule undergoes further organisation (folding and reordering) to define the tertiary structure of the protein. The secondary and tertiary structures are non-covalent in nature and a range of interactions contribute to the final structure, such as hydrogen bonding, electrostatic interactions and dispersion forces. These interactions occur between adjacent amino acid groups and maybe mediated or altered by solvation, the ionic strength of the solution and a range of other environmental factors. Finally, these individual pep-tides often interact with other peptides to perform specific functions or to provide structural integrity, and this final structure of the protein is known as the quaternary structure.

The dynamic nature of the secondary and tertiary structures of proteins enables them to undergo structural changes in response to stimuli that are often related to their functional roles. Structurally distinct regions within a protein are often associated with specific functions, and these structures may be conserved to undertake similar functions across a range of different proteins. The shape of the protein, especially its external surface often provides pockets or clefts, which offer specific binding sites for small molecules or other proteins. Interactions with these sites is often described as specific binding and is associated with the activation or regulation of the activity of the protein.

The analysis of protein structure at the tertiary and quaternary levels, especially in real time, is a particularly challenging task, but is extremely important to the pharmaceutical industry and to protein researchers. Since the early successes in determining the x-ray crystal structure of myoglobin and haemoglobin (Kendrew et al. 1958; Muirhead and Perutz 1963; Pe-rutz 1963), biochemists and pharmacologists have tended to rely on x-ray diffraction techniques for information on a protein's three-dimensional structure and its relationship to its function. Structure-function relationships are particularly important in order to gain a better understanding of how a particular protein operates or how drug molecules could be designed to modulate the behaviour of a protein implicated in a disease mechanism. The importance of these studies is demonstrated by the current drive in pharmaceutical research towards structurally informed drug discovery.

X-ray diffraction techniques require a diffraction-quality protein crystal. Producing such crystals is extremely challenging and is often the rate-limiting, or even terminal step in the process of determining molecular structure. In the crystalline state the protein molecule is in a condition that is far from that found in vivo. The contributions of the crystal lattice, changes in solvation state and use of modifiers in order to obtain crystals, for example, all lead to a degree of uncertainty regarding the fidelity of the crystal structure when compared to the structure adopted by the protein in its natural environment. In addition, x-ray diffraction provides no information on the mechanisms by which structural change occurs. The best that can be achieved is the crystal structure of the native protein and the crystal structure of the protein in the presence of a small molecule (i. e. a candidate drug molecule). This provides the molecular structures of the starting condition (the native protein) and the molecular structure of the end point (when the protein has bound the small molecule). Whilst this is a highly information-rich technique, the crystallisation of proteins is an extremely difficult and time-consuming process and it is not uncommon for the drug discovery process to have to proceed in the absence of critical structural information.

The drug discovery process typically starts with the isolation and identification of proteins that, it is suspected, are implicated in the diseased state. Once this has been achieved the next key step is to obtain as much structural information as possible. This usually begins with attempts to obtain x-ray diffraction data to solve crystal structures, which can be a serious bottleneck in the discovery process. Membrane-bound proteins, which are currently of great interest to the pharmaceutical industry, have been found to be particularly difficult in this respect, representing a very small fraction of the protein structures available on the Brookhaven Protein Database (Berman et al. 2000). Even when structural information is available, the value is limited as the data provided is static, the protein being "frozen" in a crystal lattice, and few insights are offered into the mechanism by which the protein functions.

As a result of the limitations of structural information and its relationship to the function of a protein obtained through x-ray crystallography, several techniques have been developed in recent years to supplement the information provided by diffraction experiments. These include high-field nuclear magnetic resonance (NMR) and neutron reflection techniques. NMR techniques are useful but very data intensive (and the structures are somewhat ambiguous), whilst neutron reflection offers more limited structural information but can provide a limited degree of temporal information.

True real-time capabilities are offered by sensor techniques such as mi-crocalorimetry, acoustic wave devices (predominantly quartz crystal microbalance) and optical techniques, including, for example, surface plas-mon resonance (SPR). As a consequence, structural determination has tended to remain the preserve of x-ray crystallography, whilst function has tended to be determined using calorimetric, acoustic wave or optical techniques. The latter, often being at best semi-quantitative in nature, have tended to rely on temporal characteristics, measuring the kinetics of protein interaction events, which are then related to function. Any relationship between temporal sensor data and structure tends to be, at best, putative in nature. This has led to the separation of the structural and functional measurements, which requires the adoption of some significant assumptions. For readers wishing to obtain more general reviews on tagless biosensor platforms are directed towards reviews on this subject (Baird et al. 2001).

Recently, techniques have been developed that attempt to link structure and function. Certain modifications to acoustic-wave devices, utilising the characteristics of the dissipation of acoustic modes (sometimes referred to as 'QCM-D') can under some conditions provide information on the dimensionality of deposited thin films which, in conjunction with the mass loading (determined from the change in resonant frequency of the acoustic device), can offer some insights in to surface structure. In order to provide a full analytical treatment of such data, several significant assumptions must be made. QCM data is of relatively low resolution when compared with analogous optical techniques, which limits its utility in the pharmaceutically relevant area of protein-small molecule interactions. Optical approaches have been described; these include optical waveguide light spectroscopy (OWLS) (Tiefenthaler and Lukosz 1989), ellipsometry (Spaeth et al. 1997), coupled plasmon waveguide resonance (CPWR) (Sala-mon et al. 2000) and dual polarisation interferometry (DPI; Cross et al. 2003). OWLS provides relatively low-layer parameter and temporal resolution, whilst ellipsometry suffers from both resolution issues and ambiguity in terms of data interpretation. CPWR has found limited use in the life science arena but does not appear to have been widely adopted. The latter technique, DPI, involves exciting a waveguide structure with two polarisations of light, which provides two independent measures of the surface condition of the waveguide itself. From this data, the thickness and density (or refractive index) of thin-layer structures on the waveguide surface can be determined with relatively few assumptions to very high precision and with good temporal resolution. The technique has found broad application in life science and surface science research.

In the present article, the application of DPI will be explored. A general background, and introduction to the theoretical treatment of DPI has been provided at the end of this article (see Appendices 1-3).

Was this article helpful?

0 0

Post a comment