Conical tomography : a simple method to study proteins in cells at high resolution

Despite advances in molecular biology and genetics, the location of thousand of proteins in cells remains undetermined. The principal problems are their small dimensions and their capacity to form large assemblies by associating either with themselves or other proteins. We took advantage of the tendency to form aggregates and developed a simple method that describe the three-dimensional structure of these assemblies in cells at high resolution (2-3nm) and in three-dimensions. As a proof of principle, we studied the alphaA-crystalline, a small chaperone that plays an important role in lens transparency and cataract formation. To identify the assemblies containing the chaperone, lens tissues labeled with primary anti-alphaA-crystalline and probed with both 2nm and 5nm diameter gold particle conjugates were reconstructed by conical tomography. First, we determined the location of all gold particles contained within the reconstructed volume. From maps of their 3D-distribution, we determined that gold particles formed files that repeated at 6-7nm center-to-center apart and bent at angles measuring ~90° or ~120°. Second, we identified the tethers that linked each gold particle to the assemblies containing the chaperone. Independent of the diameter of the gold particle, tethers formed by the association of primary and secondary antibodies measured ~14nm in length. Finally, by applying the constraints represented by the repeat period, the angles and the structure of the assemblies, we identified the chaperone in unlabeled tissues as small globules spaced 6-7nm apart decorating thin filaments of the cytoskeleton. In conclusion, the high resolution in three-dimensions, the reliance on geometric constraints instead of exogenous probes and the technical simplicity are unrivaled properties of our method for studying the contribution that proteins made to normal cell homeostasis and pathological conditions. Supported by NIH EY-04110