MODELLING THE AXONEME

Michael E.J. Holwill (1,2), Helen C. Taylor (1), Ernestina Guevara (1) and Peter Satir (2)
(1) Physics Department, King's College London, UK
(2) Department of Anatomy and Structural Biology, Albert Einstein College of Medicine, NY, USA

To gain insight into the molecular interactions which are responsible for generating and regulating ciliary motion, we use computer modelling and associated experimental studies. The computer modelling consists of three interrelated and interdependent elements: 1. Structural modelling; 2. Computer animation; 3. Finite element analysis.

1: Our 3-dimensional structural computer model was developed initially in 1991 and is based on electron microscope images. The model can be viewed from any angle for comparison with electron micrographs and can be amended readily as new information becomes available. 2: Computer simulation studies predict satisfactorily the quantitative behavior of microtubules transported across a field of dynein molecules, assuming either stochastic or cooperative molecular activity. A kinetic analysis of the stochastic process indicates that the force-generating phase of an arm occupies about 1% of its cycle time. With this information, and five distinct structural states of the arm revealed by electron microscopy, we used interpolation to construct an animated computer model of the dynein arm. The moving model shows the arm interacting with the neighboring microtubule and demonstrates that the active arm must be more compact than the bouquet form seen in isolation. 3: To investigate the function of the axonemal components, it is necessary to assign mechanical parameters, such as forces and elastic moduli, to the components. For this investigation, we are using finite element (FE) analysis, in which the forces generated by the dynein arms, together with the mechanical properties of the axonemal components, can be represented on a mesh which will distort with time as the forces are applied. Deformations predicted by FE analysis can be mapped onto the structural model so that problems associated with molecular overlap and interaction can be identified; appropriate modifications and refinements to the modelling can then be made.