Examining chromosome-microtubule attachment.

Fig. 8

Chromosome-microtubule attachments during mitosis.

Fig. 9

In the presence of Eg5 inhibitors (monastrol or HR22C16), bipolar mitotic spindles do not form in mammalian cells, as spindle poles do not separate. The mono-astral spindles have most of their chromosomes in the syntelic configuration, with both sister kinetchores attached to the unseparated spindle pole.

Fig. 10

Formation of syntelic attachments during mammalian mitosis. In a cell recovering from monastrol arrest, microtubules form near the chromosome (not directly connected to the centrosome), grow, are captured at their minus-ends, and transported to spindle poles. If the transport of the chromosome-associated microtubule is to the spindle pole to which the other sister chromosome is attached, a syntelic attachment error is formed.

Fig. 11

An assay to examine the correction of improper chromosome-microtubule attachments. In the presence of Eg5 inhibitors, mono-astral spindles form that have many chromosomes with syntelic attachments. Removal of these inhibitors allows spindle bipolarization and error correction. Aurora kinases function is required for this correction process. Reversible inhibition of these kinases allows the correction process to be directly observed and analyzed.

During cell division, chromosomes need to attach to microtubules such that each sister kinetochore is attached to opposite poles of a bipolar spindle (amphitelic attachment, Fig. 8). An intermediate in the attachment process is called monotelic attacment, where only one sister chromosome is attached to spindle microtubules. Other attachments described by cytologists are syntelic attachments, where both sister chromosomes are attached to one spindle pole, and merotelic attachments, where one sister kinetochore is attached to two spindle poles (Fig. 8). If these improper chromosome-microtubule attachments are not corrected, whole chromosomes will be lost during cell division.

Our research has focused on syntelic microtubule attachments. We have examined how these errors can form and how they are corrected during mammalian mitosis (REF)?

Powerful tools in examining this chromosome-microtubule attachment have been reversible small molecule inhibitors of the mitotic kinesin Eg5. In the presence of Eg5 inhibitors, spindle poles do not separate and as many as 70% of the chromosomes in a cell have syntelic attachments (Fig. 9). Using high resolution microscopy assays we have directly observed one pathway by which these attachments errors can arise (Fig 10).

Eg5 inhibitors like monastrol are reversible. Therefore, we can use these inhibitors to activate Eg5 function in mammalian mitosis with exquisite temporal control. Monastrol treatment accumulates syntelic attachment errors, and by washing-out monastrol to allow spindle bipolarization, we can show that mechanisms exist to rapidly correct syntelic errors. This conditional inhibition and activation of a key mitotic motor protein is the basis of a simple assay we use to examine how syntelic errors are detected and corrected.

Aurora kinases play an important role in regulating chromosome-microtubule attachment during cell division. Using different inhibitors for Aurora kinases, high-resolution microscopy, and our assay for syntelic chromosome attachment in mammalian cells, we have shown that Aurora kinases are required for correcting syntelic attachment. Using Aurora kinase inhibitors that are reversible, and after determining the kinetics of kinase activation upon wash-out, we have examined how syntelic attachments are corrected. From live cell analysis of individual chromosome-microtubule attachments we find that Aurora kinases activate the selective disassembly of microtubules at syntelic chromosomes, while amphitelic (correct) attachments remain stable (Fig 11, Movie 6). (REF). Further details of the correction of syntelic attachments, through electron microscopy analysis, and the roles of Aurora kinase substrates in this process are being analyzed.