Delta3D; Bench top assays for the rapid detection of protein 3D structural changes

Pharmaceuticals were once just small molecules such as aspirin and penicillin. These contain 21 and 42 atoms respectively and their structure is determined by the rigid bonds between each atom. Their purity is easily measured and any chemical changes are detected by established methods and their production is straight forward. Currently, a new generation of biopharmaceuticals is emerging which use large molecules such as those found in the body e.g. proteins, to selectively treat a range of disorders. One of the first biopharmaceuticals was insulin, from animal pancreas which since the 1920's has treated forms of diabetes caused by the patients' inability to make insulin. No small molecule can do what insulin does and now medicinal insulin is made by bacteria carrying the gene for human insulin. Insulin is a small protein but even so has 791 atoms linked not only by covalent bonds but also by weaker effects which stabilise its complex 3D structure. This exemplifies the main features of biopharmaceuticals; they offer highly selective effects of great medical value but are very complicated and difficult to make. Currently there is much discussion of drugs which treat previously untreatable conditions e.g. Herceptin and metastatic breast cancer. Herceptin is a protein called an antibody which kills specific cancer cells just as our own antibodies protect us from invading cells. Antibodies currently account for most new biopharmaceuticals and are complexes of more than 20,000 atoms. There are many other proteins used to target specific effects impossible to achieve with small molecules. These include hormones (signalling molecules like insulin), enzymes which carry out chemical reactions in the body, and antiviral proteins like interferons. Although very different, they all start off being made in living cells which also contain several thousand other proteins. Thus they need to be purified or contaminants will poison the patient. Once pure, their fragile structure needs to be stabilised by storage at low temperature in the presence of soluble and solid stabilising agents. Development of these stages of manufacture is called bio processing and relies upon sensitive techniques to measure the purity, concentration and quality of the product at every stage. We already have ways to look at the 3D structure of proteins and these are to ensure that the products are correct. However, they need expensive apparatus, large amounts of protein and years of experience to use effectively. We already use these methods but want to develop small test kits that are sensitive to structural changes so that any variation in protein quality can be rapidly detected and subjected to proper analysis. These kits are based upon slow, complex existing procedures that currently need too much protein to be useful. We will develop methods for the specific purification of target protein from complex mixtures so that more stages of the process can be analysed. We will also reduce the amounts of protein needed for the traditional assays. However, we wish to spend most time developing easy to use kits which detect changes in the 3D structure by other means. In one example we will separate small amounts of protein according to how much strongly they bind to a waxy surface. Good quality proteins should not whilst ones that have not properly formed are sticky and bind. Another example will use a chemical that lights up when it binds to parts of proteins that are not correctly formed. We intend not only to miniaturise this but also use it to test protein stability by observing at what temperature the chemical is released. To test whether the proteins are individual or aggregated we will use small chemical linkers to freeze the aggregated state for a later analysis which separates proteins according to their size. Finally, we will use small enzymes which cut floppy bits of protein to detect whether the molecules are correctly or loosely folded up.

Proteins and complex biologicals (such as viral particles) are a significant growth area in pharmaceuticals and now account for 30% of the drug pipeline and 10% of sales. Biologics present an important extra variable compared to the small molecule therapeutics that once dominated the market and that is a complex non-covalent 3D structure. Changes to this are not revealed by normal analytical processes but can adversely affect solubility, stability and function. In recent years industry has adopted a series of biophysical techniques to measure the 3D structural integrity of proteins. These include fluorescence, circular dichroism spectroscopy, analytical ultracentrifugation, NMR, X-ray crystallography, light scattering and gel permeation chromatography. Whilst powerful, these methods are expensive, require specialist analytical knowledge and often require large amounts of protein. We are avid users of biophysical methods but also wish that non-specialists may be able to detect changes to the soft 3D structure of a known protein. The methods need not define the exact alteration as this can be done with the existing methods once the problem has been identified. Thus we hope to improve the early detection of structural changes or structural heterogeneity in samples. Furthermore, we hope to extend the analysis from the pure protein stage towards the fermentation and formulation stages. The methods which include spectroscopy of protein probe complexes, small scale hydrophobic interaction chromatography, cross-linking and limited proteolysis are not new but their design for robust generic analysis of protein structure by non-specialists has not been realised. We hope to develop the foundations for some commercialisable kits which will become commonly used in industry and academia.