This award provides continued support for a project whose ultimate aim is the development of new instrumentation for determination of the three-dimensional atomic structure of protein and protein complexes that cannot be crystallized. Preliminary work has been supported through a prior NSF Small Grant for Exploratory Research (SGER) to the PIs. The proposed instrumentation will produce a beam of frozen hydrated molecules that traverses a synchrotron X-ray beam to produce an X-ray scattering pattern to be used for structure determination. The molecular beam will consist of droplets that contain one protein molecule or macromolecular complex; each droplet will freeze as it enters a vacuum chamber, forming a thin vitreous coating of ice around the molecule. Droplets will be aligned with a polarized infrared laser before passing through the X-ray beam. The resulting scattering pattern for one orientation will be collected continuously until sufficient data is acquired, then the direction of laser polarization will be changed, and a new pattern acquired. This process will be repeated until there is sufficient data from multiple orientations to create a tomographic image of the protein's charge-density distribution. The use of an elliptically polarized laser permits alignment of the direction (but not the sense) of all molecular axes. The phase problem will be solved by iterative methods based on the Gerchberg-Saxton-Feinup algorithm. Initial efforts will use tobacco mosaic virus as a model system, then progress to molecules of decreasing size at lower temperatures and higher laser power. Challenges include issues such as the accuracy of droplet alignment, damping times and recoil. The project addresses a problem of recognized importance using instrumentation and methodology to be developed by an interdisciplinary team of researchers. The intellectual merit of the proposal lies in its highly original approach to a problem of the great importance in modern biology. If successful, the project will enable improved understanding of the mechanisms and energetic landscape for the folding of classes of proteins whose structure has been refractory to existing structural approaches. Furthermore, by avoiding the need for crystallization, it may substantially increase the rate of structure determination for proteins of all types.
