In animal cells, the centrosome nucleates microtubule polymerization, anchors those microtubules to create arrays that are capable of work, and duplicates exactly once per cell cycle. The centrosome is unique because it is has a discrete size and structure, even though it lacks a membrane to separate it from the rest of the cytoplasm. Since the mid-1980s, approximately 60 proteins associated with the centrosome have been described. Most of these are defined as such only by immunofluorescence microscopy localization of the protein to the centrosome, and few of them have known functions. It is clear that there are many more proteins in the centrosome, but how can we identify them?
The problem with identifying centrosome proteins is that this small organelle is difficult to purify and comprises only a tiny fraction of total cell protein. Mass spectrometry has the sensitivity to overcome this problem, but there is still the problem of purity. Andersen et al. used a straightforward variation of standard mass spectrometry approaches, which they call “protein correlation profiling”, allowing many contaminating proteins to be eliminated at the analysis stage. The basis for this is intuitively obvious to anyone who has purified a protein on the basis of activity: genuine centrosome proteins should co-fractionate in a purification step, whereas contaminating proteins will not usually co-fractionate with the centrosome proteins. The authors analysed the relative abundance of the mass spectrometry peptides in the peak centrosome fraction and in surrounding fractions from the last step of centrosome purification; they focused on those proteins corresponding to peptides that co-fractionated with known centrosomal proteins.
This is a preview of subscription content, access via your institution
