Biology Reference
In-Depth Information
separase, but does not require sister chromatid disjunction (Tsou et al.
2009
).
Separase-dependent centriole disengagement during mitosis is required for timely
centriole replication in the following cell cycle. Both cells lacking separase and cells
expressing a catalytically inactive mutant show a variety of centrosome duplication
defects that range from asynchronous centriole replication to a complete failure to
replicate centrioles (Tsou et al.
2009
). Some centriole disengagement and replica-
tion is observed in cells lacking separase activity, because Plk1 also contributes to
centriole disengagement. Downregulation of either separase or Plk1 causes a similar
phenotype, while their co-depletion completely blocks both centriole disengage-
ment during mitosis and centriole replication in the subsequent cell cycle (Tsou et al.
2009
). While it was not initially clear whether the centriolar substrate(s) of separase
could be the same as those involved in sister chromatid cohesion, recent studies have
shown that cohesin subunits localize to centrosomes and likely do regulate centriole
engagement. Scc1/Rad21 localizes to centrosomes in a manner regulated by Aurora
B and Plk1, and its presence at centrosomes prevents premature centriole disen-
gagement (Nakamura et al.
2009
). Interestingly, centrosomal localization of Rad21
requires its cleavage by separase (Gimenez-Abian et al.
2010
), and while the nature
of this requirement is not clear, it demonstrates that Scc1/Rad21 is a separase
substrate that is relevant for the regulation of centriole engagement. Depleting any of
the cohesin subunits Scc1/Rad21, Smc1, or Smc3 results in multipolar mitosis as a
result of premature centriole disengagement, and the defects in centriole disen-
gagement are direct rather than representing some indirect consequence of defects in
sister chromatid cohesion (Beauchene et al.
2010
; Diaz-Martinez et al.
2010
).
However, while the role of separase in centriole engagement may involve the same
substrates as sister chromatid cohesion, it does not appear that cohesin regulates
centriole disengagement in the same way that it regulates sister chromatid dis-
junction, because Scc1/Rad21 does not localize to centrosomes until after its
cleavage by separase (Simmons Kovacs and Haase
2010
).
There are many other examples of non-proteasomal protease activities that
regulate centrosome biology, and inhibition of non-proteasomal proteases leads to
aberrant centriole elongation (Korzeniewski et al.
2010
). Although the main
compound used in that study, Z-L
3
VS, is a proteasome inhibitor, other proteasome
inhibitors could not induce centriole elongation as potently. Moreover, the ability
of Z-L
3
VS to induce centriole elongation correlated with its ability to inhibit
cleavage of casein, which is not a proteasome substrate and was not greatly
affected by other proteasome inhibitors. Although neither the specific proteases
whose inhibition promotes centriole elongation nor any explicit protease substrates
were identified, a directed RNAi screen identified several centriole assembly
factors that enhanced or suppressed the effect of Z-L
3
VS. This screen not only
validated the recently described roles of CPAP/CENPJ and CP110 in centriole
elongation (Tang et al.
2009
; Kohlmaier et al.
2009
), but also uncovered several
additional proteins involved in the control of centriole length. Although it is not
clear whether all of these proteins represent substrates of non-proteasomal pro-
teolysis, two of these proteins, FOP and CAP350 were stabilized in Z-L
3
VS treated
cells (Korzeniewski et al.
2010
).
