|
Laboratory of Molecular Biophysics
|
Previous: Bacterial sialidase., Up: Damage and Neuraminidases., Return to: Contents.
James Murray in collaboration with Dr. Geoff Grime, Department of Materials, Oxford University.
|
The PIXE technique for element indentification involves bombarding dry samples with a 2-3 MeV beam of finely focussed protons which then induce X-ray emission at characteristic wavelengths (energies) from the elements present in the sample. The Department of Materials in Oxford has a state-of-the-art proton microprobe, which routinely provides a 3MeV proton beam with a diameter of 1 µm. By scanning this beam across the sample in X and Y and sorting the characteristic X-ray signals as a function of beam position, elemental maps of the sample can be collected (see Figure 7). Over the last 12 years, we have developed the use of this technique to identify unknown elements in both liquid and crystalline protein samples (1). X-rays from elements lighter than neon do not have enough energy to penetrate the X-ray detector window and so can not be observed.
For proteins, an enormous advantage of this technique is that the sulphur in the cysteine and methionine residues provides an X-ray signal which can be used as an internal calibration for the number atoms of the element of interest per protein molecule. Absolute meausurements are unecessary and an accuracy of +/-6-10% in the elemental composition has been obtained.
We are currently concentrating on improving the sensitivity and reliability of the measurements. Experimentally this has involved two developments. Firstly, we are fabricating a new pure carbon Faraday cup for collection of the main proton beam downstream of the sample. The existing Faraday cup has become contaminated over time and significant X-ray emission from these contaminants (in particular sulphur and copper) is visible in the X-ray detector unless the cup was very carefully aligned prior to each run. This has caused us problems during the last year since we require a clean sulphur signal both to confirm the presence of protein and quantitatively to provide the internal calibration of samples.
Secondly, we are investigating alternative mounting film for our samples. We currently use 2 µm thick mylar (`spectro-film') stretched across a 1cm diameter hole in an aluminium holder which in turn is screwed on to the target ladder to be inserted into the microprobe vacuum chamber. The spectro-film contains parts per million levels of calcium, which thus increases the lower detectable limit for calcium in proteins. Since calcium is a common cation and thus an undesirable source of background signal, we are seeking different mounting films but have not yet found a better one.
This year LMB has been the beta-test site for the Dan32 software package, written by Geoff Grime (2) of the Department of Materials in Oxford, who runs the microprobe facility. Dan32 is a PC compatible and extended version of the original GUPIX analysis software. It combines PIXE and RBS (Rutherford Back Scattered proton) analyses in one package, and also implements the `Q-factor' method (3) of RBS normalisation. After significant teething problems, the software is now running relatively reliably and we are able to analyse our X-ray and RBS spectra in-house, rather than at the µprobe facility sited in Particle and Astrophysics in Keble Road.
|
We have carried out analyses of a number of proteins in liquid or crystal form during the last year, for collaborators both in Oxford and in Europe. Among these have been samples from Dr. Susan Lea (LMB), Dr. Rick Lewis (LMB), Professor Stuart Ferguson and Michael Cartron (Biochemistry Department), Professor Maria Romao and Dr. Hans Raaijmakers (Lisbon), Dr. Ehmke Pohl (EMBL, Hamburg) and Drs. Norica Nichita and Mark Wormald (Glycobiology Institute). From these and previous work, enough measurements have now been accumulated to obtain more realsitic limits for the senstitivity of the microPIXE technique for proteins in liquid form. This in turn puts lower limits on the protein concentration for which a measurement is likely to give an unambiguous result and be worth the time and effort involved, although this also depends on the question to be answered (e.g. unambiguous identification of an element versus quantitative measurement of its concentration per protein molecule). Figure 8 shows a plot of the number of cysteines and methionines (i.e. sulphur containing amino acids) divided by the total number of amino acids against the protein concentration. The dotted line demarks the region above which sulphur can be reliably detected. In crystal form the approximate concentration is 800mg/ml (calculated for Salmonella typhimurium neuraminidase crystals which have 42% solvent and a 42kD protein molecular weight).
In addition, the microPIXE technique and Oxford facility is now being used by a few protein crystallographers to analyse samples independently of us (e.g. reference 4).
References:
1. Leaving no element of doubt: analysis of proteins using microPIXE. Elspeth Garman, Structure (1999) 7, R291-299.Previous: Bacterial sialidase., Up: Damage and Neuraminidases., Return to: Contents.