Scientific Report 2005

Cell Biology

Automated Molecular Imaging

B. Carragher, C.S. Potter, A. Cheng, D. Fellmann, F. Guerra, G. Lander, S. Mallick, P. Mercurio, J. Pulokas, J. Quispe, S. Stagg, C. Suloway, C. Yoshioka, Y. Zhu

Elucidating the structure and mechanism of action of molecular machines is an emerging frontier in understanding how the information in the genome is transformed into cellular activities. Molecular machines are associations of individual components (e.g., proteins, nucleic acids, lipids) in the form of large complexes; examples include ribosomes, transcription complexes, track-motor complexes, and membrane-embedded pumps and channels. These machines are large and may also be conformationally and compositionally dynamic or present in comparatively low numbers, factors that make them extremely challenging (or impossible) objects for study by x-ray crystallography and nuclear magnetic resonance methods. Molecular microscopy, however, holds great promise for routinely and efficiently providing structural information at a resolution sufficient to resolve the secondary structure in these large molecular machines. This method could then be used in conjunction with high-resolution x-ray structures of individual proteins to interpret very large complexes to near-atomic resolution.

Unfortunately, the methods generally used in molecular microscopy are both time-consuming and labor intensive. These include the preparation of suitable specimens, the acquisition of the required very large numbers of electron micrographs, and the supervision of the sometimes-complex software needed for analysis and reconstruction of the 3-dimensional electron density maps.

The challenge then is to transform structure determination via electron microscopy into a high-throughput method. Success in this endeavor will not only facilitate the process of molecular microscopy but also expand the scope of accessible problems and make possible investigations that currently are deemed too high risk because of the inordinate effort involved. To this end, we are developing technologies to address automation for specimen handling, image acquisition, data processing, and integration of data information. We have created an integrated software system, called Leginon, that automatically collects electron micrographs of macromolecular structures (Fig. 1). This system has been integrated with automated particle-selection algorithms and analysis and processing software.

Fig. 1. Multiscale image collection from a transmission electron microscope is controlled by using Leginon, an automated data collection system. Data are managed by using a relational database and can be visualized by using a Web browser.

A major focus of our activities is the National Resource for Automated Molecular Microscopy (NRAMM), a biotechnology resource center funded by the National Center for Research Resources, National Institutes of Health. The overall mission is to develop, test, and apply technology to completely automate the processes involved in using electron cryomicroscopy to solve macromolecular structures. The current focus of NRAMM is the development of new approaches for specimen handling, automated acquisition, automated processing, and information handling.

The activities of the NRAMM are closely coupled to a number of collaborative and service projects in which fundamental biological goals are incentives for developing the new technology. During 2004, more than 20 of these projects were actively pursued. Specific examples include structural studies of coatomer complex II–coated vesicles, which are responsible for transport of proteins from the endoplasmic reticulum to the Golgi apparatus; characterization of viruslike particles manufactured in recombinant expression systems; structural studies of the crustacean clotting protein; structural studies of coronaviruses; and the structural characterization of the chloroplast ribosome. All of these collaborative projects guided the development of new approaches in the 4 core technologies while simultaneously providing new structural information relevant to specific biological problems.

An additional project, sponsored by the National Science Foundation, is the development of automated data collection techniques for imaging serial sections obtained by using an electron microscope. Understanding the fine structure of cells and cellular components contributes to a more profound understanding of cellular function and intracellular or intercellular interactions. In order to visualize these large, complex structures in 3 dimensions at resolutions sufficient to observe structure on the nanoscale, the cells must be cut into sections and then examined by using a transmission electron microscope. Acquiring high-magnification images of a long series of sections is difficult and extremely labor intensive. The region of interest in each section must be tracked across sections and across grids, a process that requires examining the sections at a variety of scales before acquiring high-magnification images of interesting areas. Multiscale imaging of this sort is not straightforward because the image formed by an electron microscope shifts and rotates as the magnification is changed. The overall task of reconstructing a 3-dimensional volume from a set of serial sections is challenging and time-consuming, and the number of large-scale reconstructions has been limited to a few spectacular examples. Our objectives are to design, develop, and implement a software application to automate the task of acquiring high-magnification images of specific regions of the cell across tens to hundreds of serial sections.

Publications

Dang, T.X., Farah, S.J., Gast, A., Robertson, C., Carragher, B., Egelman, E., Wilson-Kubalek, E.M. Helical crystallization on lipid nanotubes: streptavidin as a model protein. J. Struct. Biol. 150:90, 2005.

Mallick, S.P., Carragher, B., Potter, C.S., Kriegman, D.J. ACE: automated CTF estimation. Ultramicroscopy 104:8, 2005.

Mallick, S.P., Zhu, Y., Kriegman, D. Detecting particles in cryo-EM micrographs using learned features. J. Struct. Biol. 145:52, 2004.

O’Keefe, M.A., Turner, J.H., Musante, J.A., Hetherington, C.J.D., Cullis, A.G., Carragher, B., Jenkins, R., Milgrim, J., Milligan, R.A., Potter, C.S., Allard, L.F., Blom, D.A., Degenhardt, L., Sides, W.H. Laboratory design for high-performance electron microscopy. Microsc. Today 12:8, 2004.

Suloway, C., Pulokas, J., Fellmann, D., Cheng, A., Guerra, F., Quispe, J., Stagg, S., Potter, C.S., Carragher, B. Automated molecular microscopy: the new Leginon system. J. Struct. Biol., in press.