A cell contains many large macromolecular assemblies, composed of multiple proteins in precise arrangements, which carry out integral biological functions. To really understand the functions of these large structures, biologists want to know what they look like, down to the atomic level. But there is not just one technology that now allows this, and researchers are beginning to take an integrated approach to tackle these large structures.

An integrated structural model for the nuclear pore complex. Credit: Nature

To determine what a macromolecular assembly of interest is made up of, proteomics tools, especially mass spectrometry–based approaches, are being used to reveal composition and stoichiometry. To obtain a portrait of the overall architecture of the assembly, cryo-electron microscopy is invaluable but has not yet yielded atomic-level structures. For this, atomic structures of the individual protein components must be modeled or determined by X-ray crystallography or NMR spectroscopy.

Fitting the pieces together like a jigsaw puzzle is perhaps the most challenging aspect. Here again, proteomics tools, including chemical cross-linking and the emerging technique of native mass spectrometry, can provide valuable information about how the components specifically interact. Rigorous computational methods are being developed to bring the data together and create structural models of macromolecular assemblies.

These structures, however, are not static entities. NMR can provide dynamic information about how the assembly functions. Cryo-electron tomography is also emerging as a noninvasive tool for analyzing large structures in situ, as they would occur in 'real life' in the cell.

In the near future, we can expect to see a growing number of examples of hybrid proteomic and structural approaches for determining high-resolution models of large macromolecular assemblies. Recently, a detailed structure of the nuclear pore complex was obtained using such an integrated approach (Nature 450, 683–694; 2007 and Nature 450, 695–701; 2007), which should surely serve as a model for success.