Automated Microscopy: Automated Acquisition of Transmission Electron Micrographs

Sponsor: NSF DBI-9730056 5/1/98-4/30/01

PIs: Bridget Carragher and Clinton S. Potter

Summary:

The goal of this project is to develop a portable and extensible set of tools and applications for control and data acquisition from a remote transmission electron microscope (TEM).  This will be used to develop a system for automatically acquiring large numbers of high quality images from a TEM in a routine, robust and efficient manner.  In particular we will develop applications that allow for data collection from specimens preserved in vitreous ice.

Electron microscopy is an essential technique for structure determination of macromolecular complexes in the important resolution range from 0.3-3.0nm.  In this resolution range, TEM complements X-ray crystallography as it allows examination of large complexes which will not form three dimensional crystals.  It is also often the only technique capable of providing information on complexes under defined ionic conditions or in various conformational states.

There are several classes of biological structures which are amenable to three dimensional reconstruction based on electron micrographs.  These include two dimensional crystals, helical filaments, icosahedral viruses and single particles.  Biological structures are extremely sensitive to damage from the electron beam so that images must be acquired using very low doses of radiation.  As a result, the signal to noise ratio of individual images is very low and can only be improved by averaging together large numbers of images.

The immediate benefits of automating the collection of large numbers of electron micrographs are clear.  The data collection task is repetitive and labor intensive and is an inefficient use of time spent by a researcher.   The increase in efficiency will free up time to be spent on other critical tasks as well as improve specimen throughput and thus increase the numbers of structures which can be studied

The long term benefits carry implications for how the technique of biological structure analysis using TEM will develop in the future.  In a recent analysis, Henderson has calculated that to solve structures to atomic resolution using TEM will require the collection of tens of thousands of electron micrographs.  The promise of using these electron imaging techniques as a routine method for analysis cannot be realized using the current manual data collection procedures as the number of images involved is prohibitively high.  Furthermore, making the technique more efficient and accessible should help to make it available to more researchers and thus contribute to the overall strength of the field of three dimensional electron microscopy.  Thus the long term vision for automation is to help work towards the goal of routine atomic resolution structural analysis using EM.

The foundation of the software infrastructure that we will develop is a device independent and extensible set of applications and tools for control and data acquisition from an electron microscope over a network.  These tools will allow a researcher to design, oversee and manage protocols for the collection of large numbers of low-dose electron micrographs.  Specific applications to be developed will include a system for acquiring a systematic overview of the entire area of the specimen grid at low magnification and a complete system for acquiring final images, while subjecting the specimen to a negligible electron dose.  The distributed design of the software means that the electron microscope can be controlled and managed either locally or over a network.

By the judicious combination of automated algorithms and input from an experienced TEM operator the proposed system of software tools will provide a reliable method for collecting large numbers of high quality electron micrographs using a minimum of manual labor.

Reports:

Year 2 Annual Report