Measuring Nanometer Scale Gradients in Spindle Microtubule Dynamics Using Model Convolution Microscopy

Chad G. Pearson,* ${dagger}$ Melissa K. Gardner, ${dagger}$ ${ddagger}$ Leocadia V. Paliulis, ${sect}$ E. D. Salmon, ${sect}$ David J. Odde, ${ddagger}$ and Kerry Bloom ${sect}$

^*Department of Molecular, Cellular, and Developmental Biology, University of Colorado at Boulder, Boulder, CO 80309-0347; Department of Biomedical Engineering, University of Minnesota, Minneapolis, MN 55455; Department of Biology, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599-3280

Monitoring Editor: Orna Cohen-Fix

A computational model forthe budding yeast mitotic spindle predicts a spatial gradientin tubulin turnover that is produced by kinetochore-attachedmicrotubule (kMT) plus-end polymerization and depolymerizationdynamics. However, kMTs in yeast are often much shorter thanthe resolution limit of the light microscope, making visualizationof this gradient difficult. To overcome this limitation, wecombined digital imaging of fluorescence redistribution afterphotobleaching (FRAP) with model convolution methods to comparecomputer simulations at nanometer scale resolution to microscopicdata. We measured a gradient in microtubule dynamics in yeastspindles at $~$ 65 nm spatial intervals. Tubulin turnover is greatestnear kinetochores and lowest near the spindle poles. A ${beta}$ -tubulinmutant with decreased plus-end dynamics preserves the spatialgradient in tubulin turnover at a slower time scale, increasesaverage kinetochore microtubule length $~$ 14% percent, and decreasestension at kinetochores. The ${beta}$ -tubulin mutant cells have an increasedfrequency of chromosome loss, suggesting that the accuracy ofchromosome segregation is linked to robust kMT plus-end dynamics.

These authors contributed equally to this work.

Address correspondence to: Kerry Bloom (kbloom{at}email.unc.edu)