page 949, Green and Kaplan show that this is more than coincidence—APC truncations lead to defective mitotic spindles resembling those of colorectal tumor cells.
The faulty spindles are a result of defective capture of microtubule plus ends. The authors expressed the truncated APC protein in noncancerous cell lines that also had normal APC and found that, in these cells, spindle microtubles could no longer grab hold of the kinetochores. This caused chromosome misalignment and segregation defects. Astral microtubules were lost, presumably because they were not stabilized by interactions with the cell cortex. Without complete asters, spindles were free floating and often mispositioned.
Similar spindle problems were seen in human colorectal tumor cell lines that have high rates of chromosome instability. Patients are more susceptible to colorectal cancer even when they have only one APC truncation allele. As the truncated product dominantly interfered with plus-end attachments, Kaplan suggests that these individuals might have mitotic abnormalities in intestinal epithelial tissues. He plans to examine spindle morphology in a heterozygous APC mouse model.
It is not yet clear how APC makes microtubule plus ends sticky. APC associates with the plus-end binding protein EB1, whose loss phenocopies the APC mutant, so possibly this interaction is necessary for microtubule search and capture. ▪