MOTORS FROM BACTERIAL FLAGELLA: STRUCTURAL ANALYSES AND A STRUCTURE-BASED MODEL OF A BIOLOGICAL ROTARY MOTOR

David Gene Morgan (1), Dennis Thomas (2), Noreen Francis (1), Tanvir Shaikh (1) and David J. DeRosier (1)
(1) Rosenstiel Basic Medical Sciences Research Center, Brandeis University, Waltham, MA, USA
(2) Structural Biology Program, European Molecular Biology Laboratory, Heidelberg, Germany

Most bacteria move using motility organelles called flagella. Although different species of bacteria have different numbers and distributions of flagella, all bacterial flagella are organized in a similar fashion. Each flagellum contains a rigid helical filament which is rotated by a motor located in the bacterial envelope. The flexible connection between the filament and the motor is called the hook. The energy for rotation comes from a proton or other ion gradient across the cytoplasmic membrane: the flow of ions across the membrane and through the motor propels the bacterium by causing parts of the motor and the attached hook and filament to rotate. Thus this motor can be thought of as the smallest rotary motor powered by the movement of electrical charge.

Biochemical and genetic analyses have identified ~50 genes involved in the regulation, structure and function of the bacterial flagellum. About 20 of these genes encode proteins which are found in the flagellum itself. Our laboratory uses electron microscopy and image analysis to examine the structure of the flagellum from Salmonella typhimurium. The motor, also referred to as the basal body complex, is composed of the rod and a series of ring-like structures. For technical reasons, all previous work on basal bodies has assumed cylindrical symmetry, meaning that these sub-structures are not divided azimuthally into subunits. We have developed new image processing methods which allow us to determine azimuthal orientations of the images we analyze. This has allowed us to determine the first true three-dimensional reconstructions of the motor containing the rod and the L-, P- and MS-rings.

From this three-dimensional structure, we have confirmed that the rod is a helical assembly similar to the hook and filament. The rotational symmetries of the L-, P- and MS-rings are not yet clearly defined, but our preliminary results are consistent with the biochemical data indicating that their symmetry is 25- to 30-fold. Simultaneous work on the three-dimensional structure of the C-ring has revealed that its rotational symmetry is higher: 33- or 34-fold.

We have also analyzed basal body complexes from motile S. typhimurium strains which contain several protein fusion mutations in basal body proteins. In particular, a fusion between the major component of the MS-ring and FliG, a protein involved in switching motor direction and in torque generation, shows striking but unexpected effects on the structure of the C-ring. The FliG protein is associated with the cytoplasmic surface of the MS-ring, but this recent work suggests that it also reaches towards and influences the C-ring. Among the defects of the C-rings associated with this fusion protein is a reduction in subunit number (rotational symmetry) from 33 or 34 to about 30.

Combining our structural information with results from physical studies of motor function has led us to propse a model for the motor in which both the C-ring and the MS-ring rotate in the same direction but at different rates. This model is consistent with the observed facts, helps to explain several observations of motor function and is specific enough to suggest further experiments to confirm or disprove the model.