Summer Intern Project 2009: Nathan Topping



Project: Realistic calculation of protein infrared spectra



Infrared spectroscopy is a tool which is very useful when trying to explore the structure of molecules. The infrared region is just below the red end of the spectrum of visible light, and is given off by all hot objects; if an object absorbs the infrared radiation then it warms up. However, the particular wavelengths of infrared radiation that a molecule can absorb varies from molecule to molecule; each molecule can therefore be characterised by its specific pattern of absorption, which can be visualised as a series of peaks on a graph, showing at what wavelengths the radiation is absorbed.

Each peak in a molecule’s infrared spectrum corresponds to a particular vibration of the molecule; and the number of possible vibrations of a molecule increase in proportion to the molecule’s size. The infrared spectrum of a very large molecule such as a protein, then, is made up of many peaks, many of which overlap each other. Because the spectrum is so complicated it is difficult to extract any useful information from it. Infrared spectroscopy is therefore most useful for studying changes in a protein structure; the spectra of the protein in one state can be measured, and used as the baseline against which the spectra of the protein in a second state can be measured. Any peaks and troughs in a difference spectrum will correspond to peaks which are present in one of the protein states, but are in a different position, or have disappeared entirely, in the other state.

The building blocks of proteins are amino acids. Many of the peaks in a protein’s infrared spectrum correspond to vibrations which only involve one of the amino acids. If we can assign the peaks and troughs in the difference spectrum to vibrations belonging to particular amino acids then this tells us that something about that amino acid must be changing between the two different forms of the protein. And this in turn gives us clues as to how a protein is working. Understanding the infrared spectrum of amino acids, therefore, can be very useful when trying to discover how a protein, such as an enzyme, is working.

The main aim of my project was to use software run on the UCL supercomputer, Legion, to calculate the infrared spectra of the amino acids of most interest to us in enzymes. To simulate the amino acids within a protein they were modelled in short amino acid chains, with the amino acid in the centre. I had to start by building approximate versions of the amino acid structures. Then Gaussian, computational chemistry software which runs on Legion, was used to turn my approximate structure into a more realistic structure, as it would be likely to be found in reality. Then Gaussian used this structure to calculate the expected infrared spectrum of the molecule.

From the data that this generated I then needed to pick out the important peaks in the spectra of the amino acids, and, where we had experimental data available, compare how well the simulations were matching reality; generally we found that we were achieving reasonable match to the experimental data. We then pulled together all the information we were gathering to get our end product: a library of the expected infrared spectra of amino acids which can now be used to interpret the difference spectra of proteins.

This was completed within the time available, and so I also had time to begin to explore some extensions to the project; in particular, calculating the infrared spectrum of a segment which was cut from a real protein, the structure of which had previously been determined by x-ray crystallography. My project barely scratched the surface of what could be done, though; among other things, it would be interesting to explore the effect that isotope changes and binding of metals to amino acids has on the infrared spectrum.

The project pulled together diverse aspects of science from across the range of disciplines; physics, chemistry and biochemistry. As a Natural Sciences student studying both physics and biology, this fitted in very well with my degree programme and helped consolidate what I have been learning over the past couple of years. It also gave me an opportunity to gain experience in new areas; in computing, in working in an academic environment and in presenting my work to others. And, finally, it not only gave me invaluable experience, but has given me more idea as to what I would like to go on to do when I have finished my degree.

Page last modified on 15 sep 09 21:26