During prometaphase in mitotic cell division, chromosomes attach to th
e walls of microtubules and subsequently move to microtubule ends, whe
re they stay throughout mitosis(1,2). This end-attachment seems to be
essential for correct chromosome segregating. However, the mechanism b
y which kinetochores, the multiprotein complexes that link chromosomes
to the microtubules of the mitotic spindle(3,4), recognize and stay a
ttached to microtubule ends is not understood. One due comes from the
hydrolysis of GTP that occurs during microtubule polymerization. Altho
ugh tubulin dimers must contain GTP to polymerize, this GTP is rapidly
hydrolysed following the addition of dimers to a growing polymer. Thi
s creates a microtubule consisting largely of GDP-tubulin, with a smal
l cap of GTP-tubulin at the ends. It is possible that kinetochores dis
tinguish the different structural states of a GTP-versus a GDP-microtu
bule lattice, We have examined this question in vitro using reconstitu
ted kinetochores from the yeast Saccharomyces cerevisiae. We found tha
t kinetochores in vitro bind preferentially to GTP-rather than GDP-mic
rotubules, and to the plus-end preferentially over the lattice. Our re
sults could explain how kinetochores stay at microtubule ends and thus
segregate chromosomes correctly during mitosis in vivo. This result d
emonstrates that proteins exist that can distinguish the GTP conformat
ion of the microtubule lattice.