Research Article
137
Anomalous centriole configurations are detected in
Drosophila wing disc cells upon Cdk1 inactivation
Smruti J. Vidwans1, Mei Lie Wong2 and Patrick H. O’Farrell1,*
1Department
of Biochemistry and Biophysics, University of California – San Francisco, 513 Parnassus Avenue, Box 0448, San Francisco,
CA 94143, USA
2Howard Hughes Medical Institute, Department of Biochemistry and Biophysics, University of California – San Francisco, 513 Parnassus Avenue,
Box 0448, San Francisco, CA 94143, USA
*Author for correspondence (e-mail: [email protected])
Accepted 2 October 2002
Journal of Cell Science 116, 137-143 © 2003 The Company of Biologists Ltd
doi:10.1242/jcs.00204
Summary
The centriole, organizer of the centrosome, duplicates by
assembling a unique daughter identical to itself in overall
organization and length. The centriole is a cylindrical
structure composed of nine sets of microtubules and is thus
predicted to have nine-fold symmetry. During duplication,
a daughter lacking discrete microtubular organization first
appears off the wall of the mother centriole. It increases in
length perpendicularly away from the mother and
terminates growth when it matches the length of the
mother. How a unique daughter of the correct length and
overall organization is assembled is unknown. Here, we
describe three types of unusual centriole configurations
observed in wing imaginal discs of Drosophila following
inactivation of Cdk1. First, we observed centriole triplets
consisting of one mother and two daughters, which
suggested that centrioles have more than one potential site
for the assembly of daughters. Second, we observed
centriole triplets comprising a grandmother, mother and
daughter, which suggested that subsequent centriole
duplication cycles do not require separation of mother and
daughter centrioles. Finally, we observed centriole pairs in
which the daughter is longer than its mother. These
findings suggest that regulatory events rather than rigid
structural constraints dictate features of the stereotyped
duplication program of centrioles.
Introduction
Successful duplication and segregation of cellular organelles,
especially of those that exist in low copy number, is of
paramount importance for cell duplication. The centrosome
nucleates microtubules and organizes each of the two mitotic
spindle poles. Centrosome number is dictated by duplication
of its organizer, the centriole (Alberts et al., 1994). The
centriole, and hence the centrosome, duplicates precisely once
in each cell cycle, much like DNA. While the structure of DNA
provided immediate insight into the mechanisms of its
duplication, the physical processes underlying centriole
duplication are unknown.
A centriole is an open-ended cylinder composed of nine
sets of microtubules. The Drosophila melanogaster centriole
is about 100 nm in diameter and length (Vidwans et al.,
1999). During duplication, a daughter assembles
perpendicular to and off the side of the mother centriole
producing an asymmetric arrangement wherein the ends of
the mother centriole are free while one end of the daughter
abuts the mother (Alberts et al., 1994). The daughter is
initially shorter and lacks the distinct microtubular
appearance of a mature centriole and hence is referred to as
immature. It matures upon attainment of the microtubular
arrangement and length of the mother centriole. Finally,
mother and daughter separate (referred to as disengagement)
and each can undergo another round of duplication. These
observed ‘rules’ of daughter centriole assembly ensure that
the length of the centriole is maintained through generations
of duplication.
Centriolar structure and that of the mother-daughter pair
raises issues about the mode of duplication. Specifically, how
is the structure of the daughter centriole specified? Is centriole
duplication replicative, like that of DNA (i.e. does information
present in the mother play an instructive role in assembling the
daughter?). Unlike the case of DNA, it is difficult to imagine
how information might be passed from a mother centriole to
its daughter because the orthogonal relationship of motherdaughter centriole pairs would seem to preclude the
juxtapositioning of substructures that might allow templating.
The process is more easily thought of as a nucleated selfassembly in which the final structure of the centriole daughter
is dictated by interactions inherent to the assembly process.
Even so, the assembly of a centriole appears to depend on preexisting centrioles (Maniotis and Schliwa, 1991; Sluder et al.,
1989) and appears to be regulated in the context of the cell
cycle (Callaini et al., 1997; Vidwans et al., 1999). These
regulatory inputs may impinge directly on the duplication
process, for example, by regulating activity of sites of assembly
on a mother centriole or limiting the time frame for growth of
a daughter. At present, we have very little experimental
evidence to resolve these types of issues and define the nature
of the centriole duplication process.
We addressed some of these questions by manipulating a key
cell cycle regulator in the wing imaginal discs of Drosophila.
Key words: Centriole, Cdk1, Duplication, Nucleation, Cell cycle
138
Journal of Cell Science 116 (1)
The mitotic kinase, Cdk1, is required for mitosis and for a
checkpoint that normally prevents another S phase prior to
mitosis (Hayashi, 1996; Hayles et al., 1994; Knoblich et al.,
1994; Weigmann et al., 1997). Cdk1 inactivation in the
normally diploid cells of wing discs prevents further mitosis
but allows multiple rounds of S phases (Weigmann et al.,
1997). Examination of centrioles by electron microscopy
revealed centrioles that were longer than normal, daughter
centrioles that were longer than their mothers and triplet
centrioles. The presence of these centriole configurations
indicates that, at least under these circumstances: (1) more than
one site on a centriole can participate in daughter assembly;
(2) separation of a mother-daughter pair is not essential for new
centriole synthesis; and (3) unlike DNA duplication,
templating by the mother does not specify the length of the
daughter. Because these outcomes followed inactivation of a
cell cycle regulator, Cdk1, rather than a disruption of
centrosomal components, the results suggest that, normally,
regulatory controls, rather than structural constraints dictate
aspects of the stereotyped centriolar duplication program.
Materials and Methods
Inactivation of Cdk1 in wing imaginal discs
Sevelen (wild-type) or Cdk1ts flies were maintained at 18°C. Embryos
were collected and allowed to hatch and larvae allowed to develop at
18°C. Late second instar larvae were picked on a grape-juice plate
and incubated at 30°C by transferring the grape juice plate to a humid
30°C incubator. Larvae were fixed 2 days after incubation at 30°C.
Larval heads with imaginal discs attached to them were fixed in 7%
formaldehyde (in PBS) for 20 minutes (for visualization under the
light microscope) or as described below for EM.
Preparation of wing imaginal discs for EM
Dissected larvae heads were fixed in 5% glutaraldehyde (diluted in 0.1
M cacodylate buffer) (Vidwans et al., 1999) for 1-2 hours at room
temperature. They were then washed overnight in cacodylate buffer at
4°C. This was followed by post-fixation in 2% osmium tetroxide
(diluted in 0.1 M cacodylate buffer) at room temperature for 2 hours.
The discs were then dehydrated by incubation for a few minutes each
in a series of ethanol washes increasing in strength (10%, 30%, 50%
and 70%). We typically dissected away the wing imaginal discs from
the larvae heads in 70% ethanol. Then the imaginal discs were
dehydrated and treated as described for embryos (Vidwans et al., 1999).
Results
Inactivation of Cdk1 leads to aberrant centriole
duplication in wing discs
We first examined centrioles from wild-type third instar wing
discs (wt). EM analysis revealed that 20% of centrioles in
interphase cells were singlets, whereas the rest were pairs with
immature daughters (n=31, data not shown). Wing disc cells
double in about 8 hours and spend approximately equal time
in G1, S and G2 phases (Milan et al., 1996). It follows that
about one-third of interphase cells from wt wing discs are in
G1. In Drosophila embryos, G1 centrioles exist as singlets,
daughter initiation accompanies S phase (Callaini et al., 1997)
and maturation of daughter centrioles occurs in mitosis
(Vidwans et al., 1999). If the timing of steps in the centriole
cycle were conserved between Drosophila embryos and wing
imaginal discs, then about two-thirds of interphase centriole
Fig. 1. Wing discs from sevelen (wild-type) larvae or Cdk1ts larvae.
Eggs were collected and aged to the late second instar larval stage at
18°C. Larvae were then shifted to 30°C for 2 days, after which they
were fixed. Discs were dissected and stained for DNA with the dye,
bis-benzidine (Hoechst). Wing discs in which Cdk1 has been
inactivated are, typically, smaller in size compared with similarly
treated sevelen discs and have fewer cells that are bigger than
normal. Bar, 25 µm.
pairs should contain immature daughters (corresponding to
cells in S and G2) and the rest should be singlets. The observed
preponderance of centriole pairs with immature daughters
agrees with this expectation.
It has been previously reported that inactivation of Cdk1
converts mitotic wing disc cells into endoreplicating cells
(Weigmann et al., 1997). Consistent with this, wing imaginal
discs in which Cdk1 had been inactivated (ts discs) were
composed of fewer cells that were bigger and more brightly
stained with a DNA stain (presumably due to suppression of
mitosis and polyploidy, respectively) (Fig. 1). In contrast to
naturally endoreplicating cells in which centriole duplication
is arrested and the centrioles retain normal morphology
(Mahowald et al., 1979), endoreplicating cells of ts discs
exhibited a variety of centriole forms that suggested continued
and anomalous duplication cycles. Triplet forms provide
insights into the control of new daughter assembly (see below)
while centriole pairs with unusually long daughters provide
insights into the processes that terminate replicative growth of
daughter centrioles.
Measurement of the mother centriole in control discs defines
the average length of the mature centriole as 115 nm (Fig.
2A,B; Fig. 3). The daughter centriole of a pair is generally
shorter than (70 nm on the average) or occasionally comparable
in length to its mother (Fig. 3; Fig. 2A,B). This result is
consistent with observations made in the embryos where
daughter centrioles reach their mature length only upon entry
into mitosis whereupon they soon separate from their mothers.
In contrast, 15% of the centriole pairs from ts discs were longer
than their mothers (Table 1). The average daughter centriole
length was observed to be 114 nm – a 63% increase over
normal – while mother centriole length averaged 134 nm – a
17% increase (Figs 2, 3). These data suggest that the observed
aberrant centriole pairs is attributable to excess growth of the
daughter and not shrinkage of the mother.
Mother centrioles with two daughters
About 12.5% (n=64) of centrioles in ts discs were triplets
containing three centrioles in one of two different
Anomalous centrioles upon Cdk1 inactivation
139
Fig. 2. Centriole length is misregulated upon
inactivation of Cdk1. In all cases, the mother centriole
is on the left and the daughter on the right.
Furthermore, we have oriented images so that mothers
in longitudinal section have their long axes running
top-bottom and daughters in longitudinal section have
their long axes facing left-right. Since the section is
thinner than the centriole, a longitudinal section
through a centriole appears as a pair of parallel lines
(representing the walls of the centriole). A centriole in
cross-section appears as a circle (Vidwans et al.,
1999). The pairs of pictures in A, D and E are adjacent
sections from a series. (A,B) Control centrioles from
sevelen wing discs incubated at 30°C for 2 days.
(C,D,E) Centrioles from Cdk1-inactivated wing disc
cells. (A) Both the mother and the daughter in this
centriole pair are in longitudinal section. Note that the
daughter is shorter than its mother (compare the
mother in the left section to daughter in the right).
(B) Both the mother and the daughter centrioles are in
longitudinal section, with the daughter shorter than its
mother. (C) A longitudinal mother-daughter centriole
pair with a long mother. Compare the length of this
mother to those in A and B. (D) A mother in crosssection with a long daughter in longitudinal section.
Compare the length of this mother to that in A and B.
(E) A longitudinal centriole pair with long mother and
daughter. The daughter is longer than its mother. Bar,
100 nm.
Table 1. Distribution of centrioles from Cdk1ts and sevelen
wing discs incubated at 30°C for 2 days
Triplets*
Centriole pairs l(m)>l(d)
Centriole pairs l(m)=l(d)
Centriole pairs l(m)<l(d)
Cdk1-inactivated
wing disc centrioles
Sevelen (wild-type)
wing disc centrioles
12.5% (n=64)
75% (n=20)
10%
15%
0 (n=47)
97% (n=35)
0%
3%
*Of 64 centrioles observed in Cdk1-inactivated wing discs, 12.5% were
triplet, whereas none of the wild-type centrioles was triplet. Of 20 centriole
pairs from the Cdk1-inactivated discs in which lengths of the mother and
daughter were compared, 10% consisted of daughters equal in length to their
mothers and 15% of pairs consisted of daughters longer than mothers.
configurations, neither of which was observed in wild-type
(n=31). In one configuration, two daughter centrioles appeared
to associate with a single mother. The relationship between the
centrioles in a triplet configuration was inferred based on the
asymmetric nature of centriole duplication (see Introduction).
In the examples shown in Fig. 4, one centriole was identified
as mother because neither of its ends abuts another centriole.
The other two centrioles extend perpendicular to this mother
centriole in the orientation typical of daughter centrioles.
Based on this observation, we conclude that, at least under
these circumstances, a daughter centriole can be assembled at
more than one position on a mother centriole.
It is noteworthy that in each triplet, one of the daughters is
full length (or excessively long) and exhibits the morphology
Fig. 3. Distribution of
lengths of mother and
daughter centrioles from
sevelen and Cdk1ts wing
discs incubated at 30°C for
2 days. The x-axis is
centriole length (measured
in nm) and the y-axis is the
number of centrioles at any
given length. Note that the
distribution of daughter
centrioles from wild-type
discs varies between 46 and
93 nm, while the
corresponding distribution
from ts discs varies
between 46 and 162 nm.
The wider distribution of
daughter centriole lengths
in ts discs is reflected in the
higher average: 114 nm
compared with 70 nm in
wild-type. The distribution
of mother centrioles from ts
discs is also broader than
wild-type (105 to 184 nm
in ts discs compared with 93 to 139 nm in wild-type). The average
mother centriole length in ts discs is 134 nm compared with 115 nm
in wild-type.
140
Journal of Cell Science 116 (1)
Fig. 4. Mother centrioles with two daughters. Three examples of mother centrioles with two daughters. Each triplet is shown in two consecutive
serial sections with a schematic to orient the reader. (A) A mother centriole (in the center) that is slightly angled with two longitudinal
daughters, one to the left and the other to the right. (B) A mother in cross-section with two daughters, both in longitudinal section, one to the
left and one (seen only in the section at the right) at the top. (C) A mother in cross-section with two longitudinal daughters, one to the right and
the other to the left. Note that one of the daughters in each triplet is always longer than the other, suggesting that the two daughters were
assembled at different times. Furthermore, note that the differences in the angle between the two daughter centrioles suggest that the initiation
of a second daughter can occur at a variety of positions with respect to the first daughter. Bar, 100 nm.
of a mature centriole. The second daughter is short and has the
indistinct morphology typical of an immature centriole.
Mother-daughter-granddaughter triplets
The second triplet configuration appeared to have a motherdaughter-granddaughter relationship as illustrated in Fig. 5
(inferred, again, based on the asymmetric nature of centriole
duplication). The mother centriole, whose ends are free, has a
single mature daughter arranged in the typical position
orthogonal to it. The third centriole is immature and is arranged
perpendicular to the daughter. Thus, in contrast to the triplets
described earlier, a new generation is nucleated off the side of
a daughter within a mother-daughter centriole pair. While in
some of these triplets, the mother and the daughter were not
quite orthogonal (Fig. 2B,C) in other triplets, the mother and
daughter were tightly paired and oriented (Fig. 2D,E). The
simplest explanation for the precise relationship of this latter
set of centriole pairs is that, in this experimental setting,
movement of the daughter centriole away from the mother is
not necessary to initiate assembly of a new second-generation
daughter.
Discussion
Centrioles are unique cytoskeletal organelles that duplicate
precisely once in each cell cycle to generate daughters that are
identical to themselves. Unlike the other replicative structure,
DNA, the structure of the centriole does not provide insight
into the mechanism of duplication. Rather, the structural
relationships between mother and daughter centrioles highlight
the mysterious features of centriole duplication. For example,
while a centriole is purported to have ninefold symmetry along
its long axis, it assembles only one daughter at a time.
Additionally, the perpendicular arrangement of mother and
daughter centrioles minimizes opportunities for structural
templating; yet at the end of the duplication process the
daughter is as long as its mother. And finally, a daughter
centriole fails to initiate a new centriole until it has dissociated
from its mother, even thought the associated mother does not
occlude the future site of centriole initiation on the daughter.
The absence of an obvious structural basis for these observed
‘rules’ of centriole duplication suggests that these behaviors
are guided by regulatory interactions rather than rigid structural
constraints. If this were the case, one would expect that
disruptions in the regulatory circuit might result in violations
of the normal ‘rules’ of centriole duplication. We gained
insight into these issues when we observed aberrant centriole
structures in Drosophila wing imaginal discs lacking Cdk1.
Daughter centriole length is regulated
One of the observed aberrant centriole morphologies consisted
of centriole pairs in which the daughter centriole was longer
than its mother. Our observations differ from previous
descriptions of centriole elongation in that the extra growth
appears to have occurred predominantly in the daughter
centrioles. In contrast, the dramatic (20×) elongation of
centrioles in Drosophila spermatogonial cells influences both
mother and daughter centrioles (A. D. Tates,
Anomalous centrioles upon Cdk1 inactivation
141
Fig. 5. Examples of mother, daughter and granddaughter
triplet centrioles. Each triplet is shown in two consecutive
serial sections with a schematic on the right to orient the
reader. All the centriole triplets are oriented with the
mother on the left, the daughter to the right and the
granddaughter to the top or bottom. To accomplish this
orientation, adjustments were made in Adobe PhotoShop,
some of which appear as discontinuities within the images
(D,E). Note that mother and daughter in B are not quite
orthogonal, while the arrangement in D and E appears to be
orthogonal. Note that if mother daughter orientation is
fixed, the granddaughter can emerge from different faces of
the daughter, suggesting that the asymmetry is not
programmed by the preceding generation of centriole. Bar,
100 nm.
not a direct function of the length of the parent
centriole.
Cytodifferentiation during spermatogenesis in Drosophila
melanogaster: an electron microscope study, PhD thesis,
Rijksuniversiteit, Leiden, 1971). Furthermore the production of
a pseudocillia, which occurs in several species, involves
elongation of the mother centriole. The growth of the mother
centrioles by themselves or concomitantly with daughter
centrioles in these cases indicates that the excess elongation is
not connected with duplication.
While the inequity in length between mother and daughter
centrioles upon inactivation of Cdk1 suggests that misspecification of centriole length occurred during duplication, it
should be noted that mother centrioles were also unusually
long. We cannot exclude the possibility that extra centriole
growth outside the context of duplication produced these
longer mothers. However, measurements of centriole length
showed that daughters are more dramatically affected than
mothers. This is consistent with the possibility that the long
mother centrioles are secondary and represent long daughters
seen in a subsequent round of duplication.
If daughter centriole length were structurally constrained by
the mother centriole, daughter length would be at most that of
the mother. We emphasize the observation that daughter
centrioles within pairs can exceed the length of their mothers.
Consequently, we propose that maximum centriole length is
A centriole has multiple sites for daughter
centriole assembly
A second type of aberrance we observed consisted of
a mother centriole sporting two daughters
simultaneously. While it is possible that these triplet
centrioles are the result of reassociation of previously
dissociated centrioles, it appears implausible that any
random mechanism would produce the highly
stereotyped and close association that is observed
between the centrioles of such triplets. Therefore, we
suggest that these triplet centrioles result from two
rounds of centriole initiation. We emphasize that this
observation demonstrates that more than one site on a
mother centriole can be used for daughter centriole
assembly in this experimental setting.
The angle between the long axes of the two
daughter centrioles varied among the observed
triplets. Although our data are not sufficient to define the
number of possible angles between the two daughters, we note
that this observation is compatible with a model in which there
are nine potential sites (corresponding to the nine sets of
microtubules that form the wall of the centriole) and that a site
used for a round of initiation is selected randomly with respect
to the site of centriole assembly in the previous round. Given
that at least two sites on a mother centriole can be used for new
centriole assembly, the normal limit of one daughter centriole
per cycle is apparently not due to a structural asymmetry in the
centriole that specifies a unique assembly site. Rather, our
results suggest that regulatory interactions limit initiation of
new daughter assembly to one per cycle.
The difference in the level of maturation of the two daughter
centrioles suggests that they did not initiate, grow, and mature
simultaneously, but instead represent successive generations of
centriole assembly. Hence, even during defective duplication
in the absence of Cdk1, it appears that there is still a temporally
regulated cycle, and in an individual cycle, a single daughter
centriole is initiated but this daughter fails to separate from its
mother prior to initiation of a second cycle. In the second cycle,
a second daughter assembles at a site distinct from the site of
assembly of the first. Apparently, triplets arise because of a
failure or defect in separation of mother and daughter
142
Journal of Cell Science 116 (1)
centrioles, or a failure in coordination of a new round of
centriole duplication with completion of the first. Note that
even if transition to the second cycle of duplication is
accompanied by a microscopically undetectable change in
mother-daughter association, the precise structural association
observed between the members of the triplet suggests that
positions of the daughters mark their birth sites and hence
demonstrate that the mother can use more than one site to
initiate daughters.
Mother-daughter centriole pair separation is not
essential for a new round of assembly on the daughter
centriole
What prevents a daughter centriole from inappropriately
nucleating a granddaughter? Prospective assembly sites on
the daughter centriole are unobstructed while in association
with its mother. Triplets consisting of mother-daughtergranddaughter centrioles indicate that, under our
experimental conditions, the transition from a daughter to a
mother does not require centriole pair separation. What
conditions, if any, have to be met for a daughter in a motherdaughter pair to nucleate its own daughter? Ordinarily, the
daughter centriole does not mature until entry into mitosis
(Callaini et al., 1997; Vorobjev and Chentsov, 1982), hence
the lack of maturity might limit a new generation of daughter
centriole assembly during G2. The daughters we observed
that had nucleated granddaughters were at least the length of
a normal mother. We consequently suggest that a daughter
may have to attain mature length as a prerequisite for
unveiling sites for daughter assembly.
The impact of cell cycle regulators on centriole
duplication
Drosophila centrioles at the developmental stages we studied
are unique: they are comprised of singlet microtubules and lack
some of the sophisticated appendages normally associated with
more ‘mature’ centrioles. It is possible that ‘immature’
Drosophila centrioles may be able to assemble non-canonical
replicative structures unlike ‘mature’ centrioles that may be
more constrained (Garreau De Loubresse et al., 2001). Thus
these observations might be unique to ‘immature centrioles’.
However, cancerous cells are often associated with aberrant
centriole morphologies, which suggests that our observations
might be generalized to other systems and that such anomalous
centriole duplication might be one of the defects underlying
the mitotic instability of cancer cells. The possible involvement
in genome destabilization emphasizes the importance of
understanding the basis of the regulation of daughter centriole
assembly.
The steps between inactivation of Cdk1 and the final
anomalies are unknown and could involve a complex cascade
and multiple defects. Nonetheless, our observations are
consistent with alteration of the centriole cycle at a single step.
As described above, the two daughters in a triplet
configuration differ in maturity suggesting that they arise as a
result of two duplication cycles without intervening
disengagement of the products of the first cycle. Centriole
pairs with excessively long daughters could also be due to
continued (or renewed) replicative growth without
disengagement. The simplest explanation for the observed
anomalies is that coordination between disengagement and
progression of duplication is lost because disengagement is
defective or retarded. Alternatively, centriole maturation and
duplication might lack normal constraints and occasionally get
a step ahead of disengagement. In either case, it appears that
disengagement is not totally blocked as centriole pairs
continue to predominate.
Cell cycle regulators have previously been implicated in
coordinating centriole duplication with cell cycle progression.
In S. cerevisiae, mutation of Y19 of CDC28 (Cdk1) to E
(cdc28-E19), a change that is presumed to mimic inhibitory
phosphorylation at this position, or overproduction of the
kinase phosphorylating this site blocked duplication of the
centrosome analog, the spindle pole body (SPB), at the stage
of mother-daughter separation (Lim et al., 1996). Importantly,
the block was specific in that other aspects of the cycle
continued. These results led the authors to conclude that Cdk1
dephosphorylation is required for SPB duplication. In
Drosophila, maturation of daughter centrioles requires
Cdc25string, a phosphatase that dephosphorylates Y-15 of Cdk1
(analogous to Y19 of Cdc28). Dephosphorylation of Cdk1
might be an activating change or Cdk1-YPO4 might impose a
constraint on centriole duplication that is removed upon
dephosphorylation. Two types of observation favor the
interpretation that phosphorylated Cdk1 inhibits progress of
the centriole cycle. First, in S. cerevisiae the cdc28-E19 mutant
has high kinase activity that is sufficient to support cell cycle
events other than SPB duplication, hence the block to SPB
duplication is not likely to be due to a lack of activity, but could
be due to the ability of Cdc28-E19 kinase to mimic a
phosphorylated form that can inhibit SPB duplication (Lim et
al., 1996). Second, findings in S. cerevisiae and S. pombe, as
well as those reported here for Drosophila, suggest that loss
of function of Cdk1 releases constraints on centriole/SPB
duplication (Haase et al., 2001) (S. Uzawa and W. Z. Cande,
personal communication), arguing against a stimulatory role of
cyclin/Cdk1. A plausible synthesis of these results suggests
that cyclin/Cdk1-YPO4 inhibits progress of the SPB/centriole
duplication cycle and its elimination leads to miscoordination
of the duplication program. However, because the results are
based on genetic tests, the connection between Cdk1 and the
centriole cycle might be indirect and complex, and diverse
interactions might contribute to the results in the different
systems. More studies examining the role of the cyclin/Cdk1
complex in regulating progress of the centriole duplication
cycle are needed.
In summary, we provide evidence that regulatory rather than
structural constraints limit the number of daughters assembled
by a centriole, and argue that the ‘yardstick’ that defines
centriole length is independent of the mother centriole. Like
the cell biologists of a century ago, we remain confounded by
the duplication program of the centriole, but hope that
identification of regulatory contributions to this duplication
will shed light on the process.
We thank T. Su, M. Winey, R. Feldman, A. Echard, E. Foley, P.
DiGregorio, A. Shermoen, F. Banuett and D. Parry for comments on
the manuscript. This work was supported by a Howard Hughes
Medical Institute predoctoral fellowship to S.J.V. and NIH grant
GM37193 to P.H.O’F.
Anomalous centrioles upon Cdk1 inactivation
References
Alberts, B., Bray, D., Lewis, J., Raff, M., Roberts, K. and Watson, J.
(1994). Molecular Biology of the Cell. New York: Garland Publishing.
Callaini, G., Whitfield, W. G. and Riparbelli, M. G. (1997). Centriole and
centrosome dynamics during the embryonic cell cycles that follow the
formation of the cellular blastoderm in Drosophila. Exp. Cell Res. 234, 183190.
Garreau de Loubresse, N., Ruiz, F., Beisson, J. and Klotz, C. (2001). Role
of delta-tubulin and the C-tubule in assembly of Paramecium basal bodies.
BMC Cell Biol. 2, 4.
Haase, S. B., Winey, M. and Reed, S. I. (2001). Multi-step control of spindle
pole body duplication by cyclin-dependent kinase. Nat. Cell Biol. 3, 38-42.
Hayashi, S. (1996). A Cdc2 dependent checkpoint maintains diploidy in
Drosophila. Development 122, 1051-1058.
Hayles, J., Fisher, D., Woollard, A. and Nurse, P. (1994). Temporal order of
S phase and mitosis in fission yeast is determined by the state of the
p34cdc2-mitotic B cyclin complex. Cell 78, 813-822.
Knoblich, J. A., Sauer, K., Jones, L., Richardson, H., Saint, R. and Lehner,
C. F. (1994). Cyclin E controls S phase progression and its down-regulation
during Drosophila embryogenesis is required for the arrest of cell
proliferation. Cell 77, 107-120.
Lim, H. H., Goh, P. Y. and Surana, U. (1996). Spindle pole body separation
143
in Saccharomyces cerevisiae requires dephosphorylation of the tyrosine 19
residue of Cdc28. Mol. Cell Biol. 16, 6385-6397.
Mahowald, A. P., Caulton, J. H., Edwards, M. K. and Floyd, A. D. (1979).
Loss of centrioles and polyploidization in follicle cells of Drosophila
melanogaster. Exp. Cell Res. 118, 404-410.
Maniotis, A. and Schliwa, M. (1991). Microsurgical removal of centrosomes
blocks cell reproduction and centriole generation in BSC-1 cells. Cell 67,
495-504.
Milan, M., Campuzano, S. and Garcia-Bellido, A. (1996). Cell cycling and
patterned cell proliferation in the wing primordium of Drosophila. Proc.
Natl. Acad. Sci. USA 93, 640-645.
Sluder, G., Miller, F. J. and Rieder, C. L. (1989). Reproductive capacity of sea
urchin centrosomes without centrioles. Cell Motil. Cytoskeleton 13, 264-273.
Vidwans, S. J., Wong, M. L. and O’Farrell, P. H. (1999). Mitotic regulators
govern progress through steps in the centrosome duplication cycle. J. Cell
Biol. 147, 1371-1378.
Vorobjev, I. A. and Chentsov, Yu. S. (1982). Centrioles in the cell cycle. I.
Epithelial cells. J. Cell Biol. 93, 938-949.
Weigmann, K., Cohen, S. M. and Lehner, C. F. (1997). Cell cycle
progression, growth and patterning in imaginal discs despite inhibition of
cell division after inactivation of Drosophila Cdc2 kinase. Development 124,
3555-3563.